Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Double Balanced Mixer - Very Linear ?
1,125 views
Skip to first unread message
RobertMacy
Apr 2, 2014, 3:18:50 PM4/2/14
to
Anybody have experience/recommend an extremely linear double balanced
mixer?
DRIVE around 0+ dBm:
prefer unbal single drive, but can live with bal drive.
0.1MHz < fc < 100MHz
MODULATION 0+dBm
prefer unbal modulation, but can live with bal drive.
also can live with low end of 10kHz instead of DC
DC < fm < 100kHz
Need to create a lower SSB system WITH suppressed carrier, so linearity is
important.
Of course need the reverse, 'receiver' function, too, so component should
not just be labeled "upconverter", or such.
So far have identified Linear, Analog Devices, and TI as potential
suppliers, I historically shy away from Maxim.
But after sitting an hour at those websites have given up doing any
effective searches there.
Also, there might be some serious contenders who have real 'knock your
socks off' niche components for this application.
mixer?
DRIVE around 0+ dBm:
prefer unbal single drive, but can live with bal drive.
0.1MHz < fc < 100MHz
MODULATION 0+dBm
prefer unbal modulation, but can live with bal drive.
also can live with low end of 10kHz instead of DC
DC < fm < 100kHz
Need to create a lower SSB system WITH suppressed carrier, so linearity is
important.
Of course need the reverse, 'receiver' function, too, so component should
not just be labeled "upconverter", or such.
So far have identified Linear, Analog Devices, and TI as potential
suppliers, I historically shy away from Maxim.
But after sitting an hour at those websites have given up doing any
effective searches there.
Also, there might be some serious contenders who have real 'knock your
socks off' niche components for this application.
Tauno Voipio
Apr 2, 2014, 3:31:48 PM4/2/14
to
very linear and wide dynamic range mixers for years,
you should start your search there.
Google for 'Tayloe mixer'.
It seems that you need linearity only for the signal path,
and not for the carrier path, so that the mixer can be
a switching type run at the carrier rate.
Your required baseband width is a serious challenge, if
you want SSB. Please remember that a balanced mixer generates
double-sideband suppressed-carrier AM, not SSB.
--
Tauno Voipio
Jon Elson
Apr 2, 2014, 5:55:03 PM4/2/14
to
Tauno Voipio wrote:
> It seems that you need linearity only for the signal path,
> and not for the carrier path, so that the mixer can be
> a switching type run at the carrier rate.
This is the KEY, of course. A mixer HAS to be non-linear,
> It seems that you need linearity only for the signal path,
> and not for the carrier path, so that the mixer can be
> a switching type run at the carrier rate.
at least in one way, to function at all. It can be tricky
to achieve that required nonlinearity without introducing
other nonlinearities that are non desired.
Jon
Maynard A. Philbrook Jr.
Apr 2, 2014, 7:12:16 PM4/2/14
to
In article <op.xdpelpvu2cx0wh@ajm>, robert...@gmail.com says...
I used to use the LM1496 chip! :)
Jamie
Jamie
Maynard A. Philbrook Jr.
Apr 2, 2014, 7:16:52 PM4/2/14
to
nicely..
heres a video..
http://www.youtube.com/watch?v=v7h0x2Xzfns
http://www.youtube.com/watch?v=ubO4V1ehsB4
Jamie
Tim Wescott
Apr 2, 2014, 6:36:36 PM4/2/14
to
that by 'drive' you mean 'signal' and by 'modulation' you mean local
oscillator), then MiniCircuits has some diode ring mixers that might make
you very happy.
They're passive devices, so their "operating power" comes from the local
oscillator drive.
Figure on over 10dBm of local oscillator drive, and (probably)
significantly less than 0dBm of signal to get linearity that this ex-Ham
operator/receiver builder considers "very linear".
--
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com
Phil Hobbs
Apr 2, 2014, 6:54:48 PM4/2/14
to
On 4/2/2014 5:55 PM, Jon Elson wrote:
> Tauno Voipio wrote:
>
>
>> It seems that you need linearity only for the signal path, and not
>> for the carrier path, so that the mixer can be a switching type run
>> at the carrier rate.
> This is the KEY, of course. A mixer HAS to be non-linear, at least
> in one way, to function at all.
No, it doesn't. An ideal mixer is bilinear, which means that it
> Tauno Voipio wrote:
>
>
>> It seems that you need linearity only for the signal path, and not
>> for the carrier path, so that the mixer can be a switching type run
>> at the carrier rate.
> This is the KEY, of course. A mixer HAS to be non-linear, at least
> in one way, to function at all.
satisfies linear superposition at both inputs.
V_IF = k V_RF V_LO
If you double either one, the output doubles, and is otherwise
unchanged. The amplitudes of mixing products of order +-N go as the Nth
power of the input amplitudes. (*)
It can be tricky
> to achieve that required nonlinearity without introducing other
> nonlinearities that are non desired.
drive can do a very good job.
One of these times I'll build a bridge made out of E-pHEMTs such as the
ATF58143 and drive it with PECL. Should be pretty good. (Joerg or JL
may already have done this.)
I've been using quite a few BFP640FESD SiGe bipolars lately. They're a
lot easier to use than you'd think, considering that their fT is like 46
GHz. Their Early voltages are effectively infinite, which is
astonishing for such a fast transistor.
Cheers
Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics
160 North State Road #203
Briarcliff Manor NY 10510
hobbs at electrooptical dot net
http://electrooptical.net
RobertMacy
Apr 2, 2014, 8:00:50 PM4/2/14
to
On Wed, 02 Apr 2014 12:31:48 -0700, Tauno Voipio
<tauno....@notused.fi.invalid> wrote:
>> ...snip...
help. As I said, searches and I don't get along well, got something like
5,000+ useless hits form google and after sitting over 20 minutes at
Linear's website gave up.
Yes, I understand communication basics. Besides NEEDING that suppressed
carrier aspect, I plan on using two mixers and judiciously combining one
mixer's output which comes from cos fc and cos fm [add or subtract
depending on whether you want the upper, or lower sideband] with the
output from another mixer which comes from sin fc and sin fm voila! SSB, a
SINGLE tone, except for the ODD related upper harmonic tones that seem to
get in there! I need suppressed carrier because I need those upper
harmonics way down. a 5 pole Tchebyshev barely makes it to kill that third
harmonic related tone.
Is the bandwidth too much for the baseband, or for the carrier?
<tauno....@notused.fi.invalid> wrote:
>> ...snip...
>
> The ham radio technology developers have been exploring
> very linear and wide dynamic range mixers for years,
> you should start your search there.
>
> Google for 'Tayloe mixer'.
>
> It seems that you need linearity only for the signal path,
> and not for the carrier path, so that the mixer can be
> a switching type run at the carrier rate.
>
> Your required baseband width is a serious challenge, if
> you want SSB. Please remember that a balanced mixer generates
> double-sideband suppressed-carrier AM, not SSB.
>
thanks for the suggestion. If someone has a URL, or part number would
> The ham radio technology developers have been exploring
> very linear and wide dynamic range mixers for years,
> you should start your search there.
>
> Google for 'Tayloe mixer'.
>
> It seems that you need linearity only for the signal path,
> and not for the carrier path, so that the mixer can be
> a switching type run at the carrier rate.
>
> Your required baseband width is a serious challenge, if
> you want SSB. Please remember that a balanced mixer generates
> double-sideband suppressed-carrier AM, not SSB.
>
help. As I said, searches and I don't get along well, got something like
5,000+ useless hits form google and after sitting over 20 minutes at
Linear's website gave up.
Yes, I understand communication basics. Besides NEEDING that suppressed
carrier aspect, I plan on using two mixers and judiciously combining one
mixer's output which comes from cos fc and cos fm [add or subtract
depending on whether you want the upper, or lower sideband] with the
output from another mixer which comes from sin fc and sin fm voila! SSB, a
SINGLE tone, except for the ODD related upper harmonic tones that seem to
get in there! I need suppressed carrier because I need those upper
harmonics way down. a 5 pole Tchebyshev barely makes it to kill that third
harmonic related tone.
Is the bandwidth too much for the baseband, or for the carrier?
RobertMacy
Apr 2, 2014, 8:13:16 PM4/2/14
to
On Wed, 02 Apr 2014 16:12:16 -0700, Maynard A. Philbrook Jr.
<jamie_...@charter.net> wrote:
>> ...snip....
to 10MHz, hard to believe very good at 100MHz.
Thanks to Jim Thompson who kindly sent me NPN CA3046 'chip sized' models,
I've characterized the simplified MC1496 [as it appears in the data
sheet/app note]. I now have a complete LTspice model now.
However, the model predicts incredible carrier suppression [well, duh!]
but it also probably more accurately predicts the third related harmonic
to rear its ugly head at a pretty energetic level. It will be difficult to
filter that tone and get it below levels that interfere.
By the way, I've developed what I call .tranoise for use in LTspice [and
PSpice] simulations. It enables .tran analysis AND .noise analysis during
the same run. Which means you get 'realistic' noise performance through a
non-linear circuit, like a mixer. The resulting waveforms are like
watching a real scope trace with fuzz AND the spectrum has the realistic
noise floor one would expect, inlucding [if you used an OpAmp with pesky
1/f noise] you get appropriate 'bumps' in your resulting spectrum, too.
It's like a 'closer to reality' simulation. The zipped model for the
MC1496 is over 35MB.
The reason I posted is that the spec for carrier suppression is like 60
dB, not the unrealistic -120dB I now get in simulation, so thought I
should look for a better mixer...if there is one.
<jamie_...@charter.net> wrote:
>> ...snip....
>
> I used to use the LM1496 chip! :)
>
> Jamie
>
Not sure that mixer is specced much above 1MHz, probably will funciton out
> I used to use the LM1496 chip! :)
>
> Jamie
>
to 10MHz, hard to believe very good at 100MHz.
Thanks to Jim Thompson who kindly sent me NPN CA3046 'chip sized' models,
I've characterized the simplified MC1496 [as it appears in the data
sheet/app note]. I now have a complete LTspice model now.
However, the model predicts incredible carrier suppression [well, duh!]
but it also probably more accurately predicts the third related harmonic
to rear its ugly head at a pretty energetic level. It will be difficult to
filter that tone and get it below levels that interfere.
By the way, I've developed what I call .tranoise for use in LTspice [and
PSpice] simulations. It enables .tran analysis AND .noise analysis during
the same run. Which means you get 'realistic' noise performance through a
non-linear circuit, like a mixer. The resulting waveforms are like
watching a real scope trace with fuzz AND the spectrum has the realistic
noise floor one would expect, inlucding [if you used an OpAmp with pesky
1/f noise] you get appropriate 'bumps' in your resulting spectrum, too.
It's like a 'closer to reality' simulation. The zipped model for the
MC1496 is over 35MB.
The reason I posted is that the spec for carrier suppression is like 60
dB, not the unrealistic -120dB I now get in simulation, so thought I
should look for a better mixer...if there is one.
RobertMacy
Apr 2, 2014, 8:18:50 PM4/2/14
to
On Wed, 02 Apr 2014 16:16:52 -0700, Maynard A. Philbrook Jr.
<jamie_...@charter.net> wrote:
>> ...snip...
can't view videos at youtube, do you have a part number URL?
<jamie_...@charter.net> wrote:
>> ...snip...
> Any way, getting back on what I said before, a gilbert cell would work
> nicely..
> heres a video..
>
> http://www.youtube.com/watch?v=v7h0x2Xzfns
> http://www.youtube.com/watch?v=ubO4V1ehsB4
>
> Jamie
what's a gilbert cell, anything like the MC1496?
> nicely..
> heres a video..
>
> http://www.youtube.com/watch?v=v7h0x2Xzfns
> http://www.youtube.com/watch?v=ubO4V1ehsB4
>
> Jamie
can't view videos at youtube, do you have a part number URL?
RobertMacy
Apr 2, 2014, 8:26:21 PM4/2/14
to
On Wed, 02 Apr 2014 15:36:36 -0700, Tim Wescott <t...@seemywebsite.please>
wrote:
>> ...snip...
thanks.
Not sure I followed your definitions regarding signal etc, but I
understand the principles behind using diode H bridges as a mixer.
Actually designed one from discrete components that went into autopilot
electronics modules years back.
The diode mixer might have better specs than the MC1496 topology, I'm not
skilled enough in chip design to be able to answer that one. My gut
feeling is that it will be hard to beat a well laid out MC1496 structure.
*IF* someone makes one.
wrote:
>> ...snip...
>
> If you're willing to bend on your drive and modulation levels (I assume
> that by 'drive' you mean 'signal' and by 'modulation' you mean local
> oscillator), then MiniCircuits has some diode ring mixers that might make
> you very happy.
>
> They're passive devices, so their "operating power" comes from the local
> oscillator drive.
>
> Figure on over 10dBm of local oscillator drive, and (probably)
> significantly less than 0dBm of signal to get linearity that this ex-Ham
> operator/receiver builder considers "very linear".
>
Forgot all about minicircuits! they make great stuff. I'll take a look,
> If you're willing to bend on your drive and modulation levels (I assume
> that by 'drive' you mean 'signal' and by 'modulation' you mean local
> oscillator), then MiniCircuits has some diode ring mixers that might make
> you very happy.
>
> They're passive devices, so their "operating power" comes from the local
> oscillator drive.
>
> Figure on over 10dBm of local oscillator drive, and (probably)
> significantly less than 0dBm of signal to get linearity that this ex-Ham
> operator/receiver builder considers "very linear".
>
thanks.
Not sure I followed your definitions regarding signal etc, but I
understand the principles behind using diode H bridges as a mixer.
Actually designed one from discrete components that went into autopilot
electronics modules years back.
The diode mixer might have better specs than the MC1496 topology, I'm not
skilled enough in chip design to be able to answer that one. My gut
feeling is that it will be hard to beat a well laid out MC1496 structure.
*IF* someone makes one.
bloggs.fred...@gmail.com
Apr 2, 2014, 8:46:10 PM4/2/14
to
On Wednesday, April 2, 2014 6:54:48 PM UTC-4, Phil Hobbs wrote:
>
> That's true, of course, but good switching mixers with square-wave LO
>
> drive can do a very good job.
>
Dunno what kind of mixers you have in mind, but square wave switching of diode mixers has always been a big no-no, raises the noise floor by 6dB or something like that.
>
> That's true, of course, but good switching mixers with square-wave LO
>
> drive can do a very good job.
>
Phil Hobbs
Apr 2, 2014, 8:48:37 PM4/2/14
to
reduces the intermodulation distortion that occurs when the switches are
partially on. This is especially true with diode mixers, but also
applies to transistor mixers.
Bill Sloman
Apr 2, 2014, 8:52:59 PM4/2/14
to
On Thursday, 3 April 2014 11:18:50 UTC+11, Robert Macy wrote:
> On Wed, 02 Apr 2014 16:16:52 -0700, Maynard A. Philbrook Jr.
> <jamie_...@charter.net> wrote:
> >>
> >> ...snip...
> >
> > Any way, getting back on what I said before, a gilbert cell would work
> > nicely..
> > heres a video..
> >
> > http://www.youtube.com/watch?v=v7h0x2Xzfns
>
> > http://www.youtube.com/watch?v=ubO4V1ehsB4
>
> On Wed, 02 Apr 2014 16:16:52 -0700, Maynard A. Philbrook Jr.
> <jamie_...@charter.net> wrote:
> >>
> >> ...snip...
> >
> > Any way, getting back on what I said before, a gilbert cell would work
> > nicely..
> > heres a video..
> >
> > http://www.youtube.com/watch?v=v7h0x2Xzfns
>
> > http://www.youtube.com/watch?v=ubO4V1ehsB4
>
> what's a gilbert cell, anything like the MC1496?
Search on Barry Gilbert, who now works for Analog Devices, and has exploited the idea in a number of devices, which you can buy from Analog Devices.
Here's one of them
http://www.analog.com/static/imported-files/data_sheets/AD834.pdf
there are lots of others, mostly more specialised in the direction the OP wants
http://www.analog.com/en/rfif-components/mixersmultipliers/products/index.html
<snip>
--
Bill Sloman, Sydney
Maynard A. Philbrook Jr.
Apr 2, 2014, 10:29:49 PM4/2/14
to
In article <op.xdpr8ecq2cx0wh@ajm>, robert...@gmail.com says...
Interesting work. If memory serves, there were various versions of the
1496 that could perform well but the highest I've ever seen it used was
up in around 30Mhz where a complete HF QRP transceiver was constructed
using that as a mixer.
Anyway, I was at a meeting tonight and one of the guys brought up a
subject matter of (Lt)spice having a seminar close to my area in JUNE
13, I may attend that. Should be interesting meet Mike if I can work
that in my schedule.
Jamie
1496 that could perform well but the highest I've ever seen it used was
up in around 30Mhz where a complete HF QRP transceiver was constructed
using that as a mixer.
Anyway, I was at a meeting tonight and one of the guys brought up a
subject matter of (Lt)spice having a seminar close to my area in JUNE
13, I may attend that. Should be interesting meet Mike if I can work
that in my schedule.
Jamie
Maynard A. Philbrook Jr.
Apr 2, 2014, 10:35:19 PM4/2/14
to
In article <op.xdpshoek2cx0wh@ajm>, robert...@gmail.com says...
yes, that's basically it, but you can construct one using discretes
to your requirements.
Also, a ring diode mixer comes to mind.. :)
Jamie
to your requirements.
Also, a ring diode mixer comes to mind.. :)
Jamie
bloggs.fred...@gmail.com
Apr 2, 2014, 9:25:42 PM4/2/14
to
On Wednesday, April 2, 2014 8:48:37 PM UTC-4, Phil Hobbs wrote:
> On 4/2/2014 8:46 PM, bloggs.fred...@gmail.com wrote:
>
> > On Wednesday, April 2, 2014 6:54:48 PM UTC-4, Phil Hobbs wrote:
>
> >
>
> >>
>
> >> That's true, of course, but good switching mixers with square-wave
>
> >> LO
>
> >>
>
> >> drive can do a very good job.
>
> >>
>
> >
>
> >
>
> > Dunno what kind of mixers you have in mind, but square wave switching
>
> > of diode mixers has always been a big no-no, raises the noise floor
>
> > by 6dB or something like that.
>
> >
>
>
>
> Square wave switching is usually a big win, actually, because it greatly
>
> reduces the intermodulation distortion that occurs when the switches are
>
> partially on. This is especially true with diode mixers, but also
>
> applies to transistor mixers.
Really? Then you know something the career mixer designers at the big microwave houses don't.
> On 4/2/2014 8:46 PM, bloggs.fred...@gmail.com wrote:
>
> > On Wednesday, April 2, 2014 6:54:48 PM UTC-4, Phil Hobbs wrote:
>
> >
>
> >>
>
> >> That's true, of course, but good switching mixers with square-wave
>
> >> LO
>
> >>
>
> >> drive can do a very good job.
>
> >>
>
> >
>
> >
>
> > Dunno what kind of mixers you have in mind, but square wave switching
>
> > of diode mixers has always been a big no-no, raises the noise floor
>
> > by 6dB or something like that.
>
> >
>
>
>
> Square wave switching is usually a big win, actually, because it greatly
>
> reduces the intermodulation distortion that occurs when the switches are
>
> partially on. This is especially true with diode mixers, but also
>
> applies to transistor mixers.
bloggs.fred...@gmail.com
Apr 2, 2014, 9:28:24 PM4/2/14
to
Phil Hobbs
Apr 2, 2014, 9:33:23 PM4/2/14
to
On 4/2/2014 9:28 PM, bloggs.fred...@gmail.com wrote:
> On Wednesday, April 2, 2014 8:48:37 PM UTC-4, Phil Hobbs wrote:
>> On 4/2/2014 8:46 PM, bloggs.fred...@gmail.com wrote:
>>
>>> On Wednesday, April 2, 2014 6:54:48 PM UTC-4, Phil Hobbs wrote:
>>
>>>
>>
>>>>
>>
>>>> That's true, of course, but good switching mixers with
>>>> square-wave
>>
>>>> LO
>>
>>>>
>>
>>>> drive can do a very good job.
>>
>>>>
>>
>>>
>>
>>>
>>
>>> Dunno what kind of mixers you have in mind, but square wave
>>> switching
>>
>>> of diode mixers has always been a big no-no, raises the noise
>>> floor
>>
>>> by 6dB or something like that.
>>
>>>
>>
>>
>>
>> Square wave switching is usually a big win, actually, because it
>> greatly
>>
>> reduces the intermodulation distortion that occurs when the
>> switches are
>>
>> partially on. This is especially true with diode mixers, but also
>>
>> applies to transistor mixers.
>>
>>
>>
>> Cheers
>>
>>
>>
>> Phil Hobbs
>>
>
> On Wednesday, April 2, 2014 8:48:37 PM UTC-4, Phil Hobbs wrote:
>> On 4/2/2014 8:46 PM, bloggs.fred...@gmail.com wrote:
>>
>>> On Wednesday, April 2, 2014 6:54:48 PM UTC-4, Phil Hobbs wrote:
>>
>>>
>>
>>>>
>>
>>>> That's true, of course, but good switching mixers with
>>>> square-wave
>>
>>>> LO
>>
>>>>
>>
>>>> drive can do a very good job.
>>
>>>>
>>
>>>
>>
>>>
>>
>>> Dunno what kind of mixers you have in mind, but square wave
>>> switching
>>
>>> of diode mixers has always been a big no-no, raises the noise
>>> floor
>>
>>> by 6dB or something like that.
>>
>>>
>>
>>
>>
>> Square wave switching is usually a big win, actually, because it
>> greatly
>>
>> reduces the intermodulation distortion that occurs when the
>> switches are
>>
>> partially on. This is especially true with diode mixers, but also
>>
>> applies to transistor mixers.
>>
>>
>>
>> Cheers
>>
>>
>>
>> Phil Hobbs
>>
>
> BTW, IMD results from component mismatch and not "partially on"
> operation. The double balanced has theoretical infinite suppression
> of the biggest 3rdIM product.
Nope.
> operation. The double balanced has theoretical infinite suppression
> of the biggest 3rdIM product.
If you have a bridge made of perfectly matched nonlinear devices, the
even-order products will cancel, but all of the odd-order products will
make it to the output.
Jeff Liebermann
Apr 2, 2014, 11:27:01 PM4/2/14
to
On Wed, 02 Apr 2014 12:18:50 -0700, RobertMacy
<robert...@gmail.com> wrote:
>Anybody have experience/recommend an extremely linear double balanced
>mixer?
Yep.
>Anybody have experience/recommend an extremely linear double balanced
>mixer?
>DRIVE around 0+ dBm:
>Forget it at 0dBm. Not en
>prefer unbal single drive, but can live with bal drive.
>0.1MHz < fc < 100MHz
>
>MODULATION 0+dBm
>prefer unbal modulation, but can live with bal drive.
>also can live with low end of 10kHz instead of DC
>DC < fm < 100kHz
>
>Need to create a lower SSB system WITH suppressed carrier, so linearity is
>important.
>
>Of course need the reverse, 'receiver' function, too, so component should
>not just be labeled "upconverter", or such.
Whenever I have a problem like that, I like to look at existing
>0.1MHz < fc < 100MHz
>
>MODULATION 0+dBm
>prefer unbal modulation, but can live with bal drive.
>also can live with low end of 10kHz instead of DC
>DC < fm < 100kHz
>
>Need to create a lower SSB system WITH suppressed carrier, so linearity is
>important.
>
>Of course need the reverse, 'receiver' function, too, so component should
>not just be labeled "upconverter", or such.
equipment to see what the industry leaders are doing. I suggest you
take a close look at the Elecraft K3 radio. Schematic at:
<http://www.elecraft.com/manual/K3_Schematics_Jun_2010.pdf> (7.6MB)
See Page 8 for the mixer. Among the available ham radios, it has one
of the better IP3 (3rd order intermod crossover point) of 29dBm @
20Khz and 28dBm @ 2KHz[1], partly because if this mixer. I'll make a
guess the what you really want is not linearity, but rather dynamic
range, which is the difference between the noise floor and where the
gain compresses by 1dB (and goes non-linear). Google for "high
performance receiver" for more articles.
You ability to match and balance the FET switches determines your
carrier leakage. I suggest you use an integrated package, such as the
CBT33257AD
<http://www.nxp.com/products/logic/bus_switches/CBT3257AD.html>
You can assume that the devices on a single chip are fairly well
matched. If you see any carrier leakage, your problem is the PCB
layout.
There are various ways of producing SSB out of DSB. It's usually done
with a crystal filter at a specific IF frequency. Trying to do it
broadband or at baseband with phase shifters is not going to yield
good results. If you must cover a range of frequencies for both the
input and the LO, then you'll probably find yourself using two mixers
and two LO's, just so that you can use a single frequency crystal
filter. You ability to suppress the unwanted carrier will initially
be a function of the crystal filter shape factor, but eventually
become a PCB layout and shielding problem.
[1] <http://www.sherweng.com/table.html>
--
Jeff Liebermann je...@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
Jeff Liebermann
Apr 3, 2014, 12:11:34 AM4/3/14
to
with a square wave results in the diodes or active devices operating
like a switch. They're either on or off with nothing in between. If
there were some added noise in between, such as if the rise and fall
times on the square waves were slow, there might be some noise added
during the transition by an active device. To prevent this, the FET
and xsistor versions of the double balance mixer are driven hard to
make sure they're totally on, or totally off.
However, if you were to drive the LO with a sine wave, there would be
a longer time period when the active devices are operating as
amplifiers instead of switches. That's where some noise MIGHT creep
into the system. Even so, the most noise that can slither into the
system is that produced by the LO itself, times whatever gain the
active devices can deliver. With a 50 ohm load on the drain, there's
not much gain there. Since the LO and its even harmonics are nulled
out in the mixer, so will most of any added noise produced by the LO
be nulled out. The final NF is also the average of these twice per LO
cycle bursts of noise, thus again reducing the noise levels. However,
if the LO is clean and low noise, none of these effects are going to
be seen, and certainly not to the point of doubling the noise level
(+6dB).
upsid...@downunder.com
Apr 3, 2014, 1:54:46 AM4/3/14
to
>Anybody have experience/recommend an extremely linear double balanced
>mixer?
>
>DRIVE around 0+ dBm:
>prefer unbal single drive, but can live with bal drive.
>0.1MHz < fc < 100MHz
>
>MODULATION 0+dBm
>prefer unbal modulation, but can live with bal drive.
>also can live with low end of 10kHz instead of DC
>DC < fm < 100kHz
>
>Need to create a lower SSB system WITH suppressed carrier, so linearity is
>important.
SSB without carrier can be created by an image canceling I/Q mixer
>mixer?
>
>DRIVE around 0+ dBm:
>prefer unbal single drive, but can live with bal drive.
>0.1MHz < fc < 100MHz
>
>MODULATION 0+dBm
>prefer unbal modulation, but can live with bal drive.
>also can live with low end of 10kHz instead of DC
>DC < fm < 100kHz
>
>Need to create a lower SSB system WITH suppressed carrier, so linearity is
>important.
pair.
DSB with/without carrier can be generated by a single mixer by adding
some DC component on one input.
I do not see how the I/Q could generate some carrier, perhaps by
simply bypassing the carrier around the I/Q mixer.
upsid...@downunder.com
Apr 3, 2014, 2:14:28 AM4/3/14
to
On Wed, 02 Apr 2014 20:48:37 -0400, Phil Hobbs
<ho...@electrooptical.net> wrote:
>On 4/2/2014 8:46 PM, bloggs.fred...@gmail.com wrote:
>> On Wednesday, April 2, 2014 6:54:48 PM UTC-4, Phil Hobbs wrote:
>>
>>>
>>> That's true, of course, but good switching mixers with square-wave
>>> LO
>>>
>>> drive can do a very good job.
>>>
>>
>>
>> Dunno what kind of mixers you have in mind, but square wave switching
>> of diode mixers has always been a big no-no, raises the noise floor
>> by 6dB or something like that.
A passive diode ring mixer has a loss of about 6 dB, thus the noise
<ho...@electrooptical.net> wrote:
>On 4/2/2014 8:46 PM, bloggs.fred...@gmail.com wrote:
>> On Wednesday, April 2, 2014 6:54:48 PM UTC-4, Phil Hobbs wrote:
>>
>>>
>>> That's true, of course, but good switching mixers with square-wave
>>> LO
>>>
>>> drive can do a very good job.
>>>
>>
>>
>> Dunno what kind of mixers you have in mind, but square wave switching
>> of diode mixers has always been a big no-no, raises the noise floor
>> by 6dB or something like that.
figure goes up. Any decent RF amplifier in front of that can easily
take care of that loss.
>Square wave switching is usually a big win, actually, because it greatly
>reduces the intermodulation distortion that occurs when the switches are
>partially on. This is especially true with diode mixers, but also
>applies to transistor mixers.
products with the 3rd and 5th harmonic of the LO. The offending input
frequencies must be filtered out prior to the mixer that might mix
with the LO harmonics and produce the IF.
In an up converting receiver, this is easy and is handled in practice
with the same LPF that takes care of the image frequency, also removes
the offending frequencies.
Tauno Voipio
Apr 3, 2014, 2:16:07 AM4/3/14
to
On 3.4.14 03:00, RobertMacy wrote:
> On Wed, 02 Apr 2014 12:31:48 -0700, Tauno Voipio
> <tauno....@notused.fi.invalid> wrote:
>
>>
>> Your required baseband width is a serious challenge, if
>> you want SSB. Please remember that a balanced mixer generates
>> double-sideband suppressed-carrier AM, not SSB.
>
> On Wed, 02 Apr 2014 12:31:48 -0700, Tauno Voipio
> <tauno....@notused.fi.invalid> wrote:
>
>>
>> Your required baseband width is a serious challenge, if
>> you want SSB. Please remember that a balanced mixer generates
>> double-sideband suppressed-carrier AM, not SSB.
>
> Yes, I understand communication basics. Besides NEEDING that suppressed
> carrier aspect, I plan on using two mixers and judiciously combining one
> mixer's output which comes from cos fc and cos fm [add or subtract
> depending on whether you want the upper, or lower sideband] with the
> output from another mixer which comes from sin fc and sin fm voila! SSB,
> a SINGLE tone, except for the ODD related upper harmonic tones that seem
> to get in there! I need suppressed carrier because I need those upper
> harmonics way down. a 5 pole Tchebyshev barely makes it to kill that
> third harmonic related tone.
>
> Is the bandwidth too much for the baseband, or for the carrier?
Both, but for different reasons:
> carrier aspect, I plan on using two mixers and judiciously combining one
> mixer's output which comes from cos fc and cos fm [add or subtract
> depending on whether you want the upper, or lower sideband] with the
> output from another mixer which comes from sin fc and sin fm voila! SSB,
> a SINGLE tone, except for the ODD related upper harmonic tones that seem
> to get in there! I need suppressed carrier because I need those upper
> harmonics way down. a 5 pole Tchebyshev barely makes it to kill that
> third harmonic related tone.
>
> Is the bandwidth too much for the baseband, or for the carrier?
- Creating the analytic signal (two copies at 90 degrees difference)
makes the requirement of limited baseband. It must be limited at
the low end to make the phase shifting possible, 0 Hz does not go.
The minimum delay for the filters is a quarter wave at he lowest
frequency.
- All non-ideal mixers create undesired products at the output
besides the customary sum and difference frequencies. These
must be filtered out, which is a challenge at wide frequency
ranges. An ideal mixer is a perfect analog multiplier, which
remains to be invented to your specifications.
--
-TV
Jan Panteltje
Apr 3, 2014, 2:15:21 AM4/3/14
to
On a sunny day (Thu, 03 Apr 2014 08:54:46 +0300) it happened
upsid...@downunder.com wrote in
<fmtpj9trjd9hu1cee...@4ax.com>:
I have an
Atmel U2790B
quadrature modulator chip somewhere, cannot remember the exact specs,
anyways pdf says 50 Ohm, 100MHz to 1GHz, 5V 150mW single input, -1dbm output level,
Had it in circuit foe a few moments, replaced it with a digital solution,
its expensive (17 Euro or there about), but sometimes used by radio hams.
upsid...@downunder.com wrote in
<fmtpj9trjd9hu1cee...@4ax.com>:
Atmel U2790B
quadrature modulator chip somewhere, cannot remember the exact specs,
anyways pdf says 50 Ohm, 100MHz to 1GHz, 5V 150mW single input, -1dbm output level,
Had it in circuit foe a few moments, replaced it with a digital solution,
its expensive (17 Euro or there about), but sometimes used by radio hams.
upsid...@downunder.com
Apr 3, 2014, 2:42:56 AM4/3/14
to
On Wed, 2 Apr 2014 18:12:16 -0500, "Maynard A. Philbrook Jr."
<jamie_...@charter.net> wrote:
>In article <op.xdpelpvu2cx0wh@ajm>, robert...@gmail.com says...
>>
>> Anybody have experience/recommend an extremely linear double balanced
>> mixer?
>>
>> DRIVE around 0+ dBm:
>> prefer unbal single drive, but can live with bal drive.
>> 0.1MHz < fc < 100MHz
>>
>> MODULATION 0+dBm
>> prefer unbal modulation, but can live with bal drive.
>> also can live with low end of 10kHz instead of DC
>> DC < fm < 100kHz
>>
>> Need to create a lower SSB system WITH suppressed carrier, so linearity is
>> important.
>>
>> Of course need the reverse, 'receiver' function, too, so component should
>> not just be labeled "upconverter", or such.
So the receiver would also be tunable in the 0.1 to 100 MHz range ?
<jamie_...@charter.net> wrote:
>In article <op.xdpelpvu2cx0wh@ajm>, robert...@gmail.com says...
>>
>> Anybody have experience/recommend an extremely linear double balanced
>> mixer?
>>
>> DRIVE around 0+ dBm:
>> prefer unbal single drive, but can live with bal drive.
>> 0.1MHz < fc < 100MHz
>>
>> MODULATION 0+dBm
>> prefer unbal modulation, but can live with bal drive.
>> also can live with low end of 10kHz instead of DC
>> DC < fm < 100kHz
>>
>> Need to create a lower SSB system WITH suppressed carrier, so linearity is
>> important.
>>
>> Of course need the reverse, 'receiver' function, too, so component should
>> not just be labeled "upconverter", or such.
>I used to use the LM1496 chip! :)
waves, i.e. have an antenna connected directly to the 1496, you are
going to be disappointed, with overloading and spurious signals all
over the place.
You need a frequency selective front end ahead of the 1496.
Jan Panteltje
Apr 3, 2014, 5:23:24 AM4/3/14
to
On a sunny day (Thu, 03 Apr 2014 06:15:21 GMT) it happened Jan Panteltje
<pNaonSt...@yahoo.com> wrote in <lhiuds$r3c$1...@news.albasani.net>:
I forgot to add, I think for frequencies below 100 MHz most goes digital,
it is hard to make the RF 90 degrees phase shift on chip over 100 kHz to 100 MHz I'd think,
maybe Jimmy knows how to do that.
It is much easier to set up some flipflops, and, if you must, use analog switches to generate QAM.
I havedunnit for 4.43 MHz PAL modulator that way (UV).
<pNaonSt...@yahoo.com> wrote in <lhiuds$r3c$1...@news.albasani.net>:
it is hard to make the RF 90 degrees phase shift on chip over 100 kHz to 100 MHz I'd think,
maybe Jimmy knows how to do that.
It is much easier to set up some flipflops, and, if you must, use analog switches to generate QAM.
I havedunnit for 4.43 MHz PAL modulator that way (UV).
Phil Hobbs
Apr 3, 2014, 7:49:24 AM4/3/14
to
On 4/3/2014 2:14 AM, upsid...@downunder.com wrote:
> On Wed, 02 Apr 2014 20:48:37 -0400, Phil Hobbs
> <ho...@electrooptical.net> wrote:
>
>> On 4/2/2014 8:46 PM, bloggs.fred...@gmail.com wrote:
>>> On Wednesday, April 2, 2014 6:54:48 PM UTC-4, Phil Hobbs wrote:
>>>
>>>>
>>>> That's true, of course, but good switching mixers with square-wave
>>>> LO
>>>>
>>>> drive can do a very good job.
>>>>
>>>
>>>
>>> Dunno what kind of mixers you have in mind, but square wave switching
>>> of diode mixers has always been a big no-no, raises the noise floor
>>> by 6dB or something like that.
>
> A passive diode ring mixer has a loss of about 6 dB, thus the noise
> figure goes up. Any decent RF amplifier in front of that can easily
> take care of that loss.
>
>> Square wave switching is usually a big win, actually, because it greatly
>> reduces the intermodulation distortion that occurs when the switches are
>> partially on. This is especially true with diode mixers, but also
>> applies to transistor mixers.
>
> With square wave LO, you must also consider the spurious mixing
> products with the 3rd and 5th harmonic of the LO. The offending input
> frequencies must be filtered out prior to the mixer that might mix
> with the LO harmonics and produce the IF.
You get that anyway, because you have to turn the diodes on hard to get
> On Wed, 02 Apr 2014 20:48:37 -0400, Phil Hobbs
> <ho...@electrooptical.net> wrote:
>
>> On 4/2/2014 8:46 PM, bloggs.fred...@gmail.com wrote:
>>> On Wednesday, April 2, 2014 6:54:48 PM UTC-4, Phil Hobbs wrote:
>>>
>>>>
>>>> That's true, of course, but good switching mixers with square-wave
>>>> LO
>>>>
>>>> drive can do a very good job.
>>>>
>>>
>>>
>>> Dunno what kind of mixers you have in mind, but square wave switching
>>> of diode mixers has always been a big no-no, raises the noise floor
>>> by 6dB or something like that.
>
> A passive diode ring mixer has a loss of about 6 dB, thus the noise
> figure goes up. Any decent RF amplifier in front of that can easily
> take care of that loss.
>
>> Square wave switching is usually a big win, actually, because it greatly
>> reduces the intermodulation distortion that occurs when the switches are
>> partially on. This is especially true with diode mixers, but also
>> applies to transistor mixers.
>
> With square wave LO, you must also consider the spurious mixing
> products with the 3rd and 5th harmonic of the LO. The offending input
> frequencies must be filtered out prior to the mixer that might mix
> with the LO harmonics and produce the IF.
good behaviour. Mini-Circuits publishes boatloads of data on their
mixers, so you can see it easily. If you don't drive the LO hard
enough, you get a bunch of intermod between different signals on the RF
port, which is much worse than merely finding copies of the IF offset by
multiples of the LO.
>
> In an up converting receiver, this is easy and is handled in practice
> with the same LPF that takes care of the image frequency, also removes
> the offending frequencies.
>
problems, is a big part of receiver design.
Chris Jones
Apr 3, 2014, 8:50:39 AM4/3/14
to
driven hard by square wave LO signals. This is my main objection to the
N2PK VNA and the VNWA3 design also. I expect the mixer gain would be
much more stable with temperature if they were driven with square wave LO.
If I really did want to fairly thoroughly get rid of the response at the
3rd, 5th etc. harmonic of the LO, (e.g. to build a direct conversion
receiver to cover several octaves of RF input frequency, like 100kHz to
1MHz for example), then I would like to try the following:
* Low-pass filter the 100kHz-1MHz RF input with say a 5MHz roofing /
anti-aliasing filter
* In a FPGA, generate a fast-ish (>100Mbit/s) digital bitstream from a
sigma-delta modulator which has a numerically generated sine wave as its
input, so that the frequency-domain representation of the bit stream is
a strong tone at the desired LO frequency (in the range 100kHz to 1MHz)
and some noise-shaped dithering at >30MHz, then externally re-clock that
bitstream with a flip-flop clocked from a low-jitter clock
* Drive the LO port of the mixer hard with the bitstream, and put the
low-pass filtered input signal into the linear port of the mixer
Some (very linear) gain ahead of the mixer might be needed to overwhelm
any extra thermal noise from the input and low-pass filter that would
get mixed into the IF band due to the shaped noise components of the LO
signal, however it may also be feasible to reduce that noise by making
sure that the mixer input is presented by the low-pass filter with an
impedance having a very small real part (for low Johnson noise voltage)
or extremely high impedance (for low Johnson noise current) at those
(out of band) frequencies where the shaped noise is strong in the LO
bitstream.
I wonder if this is an established technique.
Chris
Phil Hobbs
Apr 3, 2014, 9:13:50 AM4/3/14
to
math of noise-shaping, because I don't really know how it works.
(Interesting things that I don't understand bug me like having a pebble
in my shoe, so I eventually get round to figuring them out.)
George Herold
Apr 3, 2014, 9:32:28 AM4/3/14
to
On Wednesday, April 2, 2014 3:18:50 PM UTC-4, Robert Macy wrote:
> Anybody have experience/recommend an extremely linear double balanced
>
> mixer?
>
>
>
> DRIVE around 0+ dBm:
>
> prefer unbal single drive, but can live with bal drive.
>
> 0.1MHz < fc < 100MHz
>
>
>
> MODULATION 0+dBm
> prefer unbal modulation, but can live with bal drive.
> also can live with low end of 10kHz instead of DC
> DC < fm < 100kHz
>
>
>
> Need to create a lower SSB system WITH suppressed carrier, so linearity is
>
> important.
>
>
>
> Anybody have experience/recommend an extremely linear double balanced
>
> mixer?
>
>
>
> DRIVE around 0+ dBm:
>
> prefer unbal single drive, but can live with bal drive.
>
> 0.1MHz < fc < 100MHz
>
>
>
> MODULATION 0+dBm
> prefer unbal modulation, but can live with bal drive.
> also can live with low end of 10kHz instead of DC
> DC < fm < 100kHz
>
>
>
> Need to create a lower SSB system WITH suppressed carrier, so linearity is
>
> important.
>
>
>
> Of course need the reverse, 'receiver' function, too, so component should
>
> not just be labeled "upconverter", or such.
>
>
>
> not just be labeled "upconverter", or such.
>
>
> So far have identified Linear, Analog Devices, and TI as potential
> suppliers, I historically shy away from Maxim.
> But after sitting an hour at those websites have given up doing any
> effective searches there.
>
> Also, there might be some serious contenders who have real 'knock your
> socks off' niche components for this application.
Fun thread, thanks for starting it.
> suppliers, I historically shy away from Maxim.
> But after sitting an hour at those websites have given up doing any
> effective searches there.
>
> Also, there might be some serious contenders who have real 'knock your
> socks off' niche components for this application.
I don't know much about mixers.
We do use the SBL-3 from minicircuits which would match your freq. specs
http://www.minicircuits.com/pdfs/SBL-3+.pdf
I think they also make some with FET switches rather than diodes.
I wonder if some more knowledgeable RF person would comment on the
differences in using a multiplier (like the AD834 that Bill S. mentioned)
Versus a switching mixer.
The only "problem" I found with "real" multipliers is they like to be driven by higher voltage levels..
(hmm this doesn't seem to be the case with the AD834 +/- 1V FS.)
http://www.analog.com/static/imported-files/data_sheets/AD834.pdf
Well there is the price.
George H.
RobertMacy
Apr 3, 2014, 10:30:57 AM4/3/14
to
On Wed, 02 Apr 2014 17:52:59 -0700, Bill Sloman <bill....@gmail.com>
wrote:
>> ...snip...
that background mode is searching, or finds something, and downloads
several MB's of pictures/videos [also hanging PC] etc etc. yet only have
26 cookies and a cleared out cache still takes forever, and now that
google has gone super commercial and no one bothers to purge dead links...
Thanks for the URLs from the 'index' i downloaded 831 too.
everything looks like 'overkill', set up for communications, and that
translates to relatively poor [for my requirements] carrier suppression,
linearity, etc.
wrote:
>> ...snip...
> Search on Barry Gilbert, who now works for Analog Devices, and has
> exploited the idea in a number of devices, which you can buy from Analog
> Devices.
>
> Here's one of them
>
> http://www.analog.com/static/imported-files/data_sheets/AD834.pdf
>
> there are lots of others, mostly more specialised in the direction the
> OP wants
>
> http://www.analog.com/en/rfif-components/mixersmultipliers/products/index.html
>
> <snip>
>
'search' just does not go well anymore. nothing happens, hangs PC while
> exploited the idea in a number of devices, which you can buy from Analog
> Devices.
>
> Here's one of them
>
> http://www.analog.com/static/imported-files/data_sheets/AD834.pdf
>
> there are lots of others, mostly more specialised in the direction the
> OP wants
>
> http://www.analog.com/en/rfif-components/mixersmultipliers/products/index.html
>
> <snip>
>
that background mode is searching, or finds something, and downloads
several MB's of pictures/videos [also hanging PC] etc etc. yet only have
26 cookies and a cleared out cache still takes forever, and now that
google has gone super commercial and no one bothers to purge dead links...
Thanks for the URLs from the 'index' i downloaded 831 too.
everything looks like 'overkill', set up for communications, and that
translates to relatively poor [for my requirements] carrier suppression,
linearity, etc.
Jeroen Belleman
Apr 3, 2014, 10:41:29 AM4/3/14
to
On 2014-04-03 14:50, Chris Jones wrote:
>
> Yes, although many people don't believe it, mixers work much better when
> driven hard by square wave LO signals. This is my main objection to the
> N2PK VNA and the VNWA3 design also. I expect the mixer gain would be
> much more stable with temperature if they were driven with square wave LO.
>
> If I really did want to fairly thoroughly get rid of the response at the
> 3rd, 5th etc. harmonic of the LO, (e.g. to build a direct conversion
> receiver to cover several octaves of RF input frequency, like 100kHz to
> 1MHz for example), then I would like to try the following:
> * Low-pass filter the 100kHz-1MHz RF input with say a 5MHz roofing /
> anti-aliasing filter
> * In a FPGA, generate a fast-ish (>100Mbit/s) digital bitstream from a
> sigma-delta modulator which has a numerically generated sine wave as its
> input, so that the frequency-domain representation of the bit stream is
> a strong tone at the desired LO frequency (in the range 100kHz to 1MHz)
> and some noise-shaped dithering at >30MHz, then externally re-clock that
> bitstream with a flip-flop clocked from a low-jitter clock
> * Drive the LO port of the mixer hard with the bitstream, and put the
> low-pass filtered input signal into the linear port of the mixer
> [...]
>
> Yes, although many people don't believe it, mixers work much better when
> driven hard by square wave LO signals. This is my main objection to the
> N2PK VNA and the VNWA3 design also. I expect the mixer gain would be
> much more stable with temperature if they were driven with square wave LO.
>
> If I really did want to fairly thoroughly get rid of the response at the
> 3rd, 5th etc. harmonic of the LO, (e.g. to build a direct conversion
> receiver to cover several octaves of RF input frequency, like 100kHz to
> 1MHz for example), then I would like to try the following:
> * Low-pass filter the 100kHz-1MHz RF input with say a 5MHz roofing /
> anti-aliasing filter
> * In a FPGA, generate a fast-ish (>100Mbit/s) digital bitstream from a
> sigma-delta modulator which has a numerically generated sine wave as its
> input, so that the frequency-domain representation of the bit stream is
> a strong tone at the desired LO frequency (in the range 100kHz to 1MHz)
> and some noise-shaped dithering at >30MHz, then externally re-clock that
> bitstream with a flip-flop clocked from a low-jitter clock
> * Drive the LO port of the mixer hard with the bitstream, and put the
> low-pass filtered input signal into the linear port of the mixer
That's a very interesting idea! To the best of my recollection,
I've never seen this before. If you don't mind, I'll mention it
to some of the local RF experts, to see what they think of it.
Jeroen Belleman
RobertMacy
Apr 3, 2014, 10:54:29 AM4/3/14
to
On Wed, 02 Apr 2014 21:11:34 -0700, Jeff Liebermann <je...@cruzio.com>
wrote:
>> ...snip...to keep Aioe happy
the MC1496 containing Jim Thompson's NPN 'chip size' CA3046 models show
that at CARRIER crossover, zero drive, the spectral noise output is at its
maximum of 94nV/rtHz [fairly independent of the modulation level], then at
the peak drive of 60mV, this drops to the minimum of around 45nV/rtHz.
Simulations confirm your statement that the noise from a mixer will be
less if it is abruptly driven ON/OFF.
wrote:
>> ...snip...to keep Aioe happy
> However, if you were to drive the LO with a sine wave, there would be
> a longer time period when the active devices are operating as
> amplifiers instead of switches. That's where some noise MIGHT creep
> into the system. Even so, the most noise that can slither into the
> system is that produced by the LO itself, times whatever gain the
> active devices can deliver. With a 50 ohm load on the drain, there's
> not much gain there. Since the LO and its even harmonics are nulled
> out in the mixer, so will most of any added noise produced by the LO
> be nulled out. The final NF is also the average of these twice per LO
> cycle bursts of noise, thus again reducing the noise levels. However,
> if the LO is clean and low noise, none of these effects are going to
> be seen, and certainly not to the point of doubling the noise level
> (+6dB).
FWIW, LTspice analyses of the circuit shown in Figure of spec sheet using
> a longer time period when the active devices are operating as
> amplifiers instead of switches. That's where some noise MIGHT creep
> into the system. Even so, the most noise that can slither into the
> system is that produced by the LO itself, times whatever gain the
> active devices can deliver. With a 50 ohm load on the drain, there's
> not much gain there. Since the LO and its even harmonics are nulled
> out in the mixer, so will most of any added noise produced by the LO
> be nulled out. The final NF is also the average of these twice per LO
> cycle bursts of noise, thus again reducing the noise levels. However,
> if the LO is clean and low noise, none of these effects are going to
> be seen, and certainly not to the point of doubling the noise level
> (+6dB).
the MC1496 containing Jim Thompson's NPN 'chip size' CA3046 models show
that at CARRIER crossover, zero drive, the spectral noise output is at its
maximum of 94nV/rtHz [fairly independent of the modulation level], then at
the peak drive of 60mV, this drops to the minimum of around 45nV/rtHz.
Simulations confirm your statement that the noise from a mixer will be
less if it is abruptly driven ON/OFF.
RobertMacy
Apr 3, 2014, 11:20:56 AM4/3/14
to
On Wed, 02 Apr 2014 23:15:21 -0700, Jan Panteltje
<pNaonSt...@yahoo.com> wrote:
> Atmel U2790B
This is why searches don't work well. Googled Atmel U2790B, got a first
page showing two ads, one from Atmel, one from mouser, followed by a list
of 'me-toos' trying to provide datasheets, and their own stuff.
Clicking on the ad for Atmel, did nothing for 20 minutes, blank screen! ok
paste in Ateml's home address, that worked.
then on the home page for Atmel did a search for the U2790B product line,
info, something and got...
Your search - U2790B - did not match any documents.
No pages were found containing "U2790B".
Suggestions:
Make sure all words are spelled correctly.
Try different keywords.
Try more general keywords.
I told you my searches just don't seem to work.
<pNaonSt...@yahoo.com> wrote:
> Atmel U2790B
This is why searches don't work well. Googled Atmel U2790B, got a first
page showing two ads, one from Atmel, one from mouser, followed by a list
of 'me-toos' trying to provide datasheets, and their own stuff.
Clicking on the ad for Atmel, did nothing for 20 minutes, blank screen! ok
paste in Ateml's home address, that worked.
then on the home page for Atmel did a search for the U2790B product line,
info, something and got...
Your search - U2790B - did not match any documents.
No pages were found containing "U2790B".
Suggestions:
Make sure all words are spelled correctly.
Try different keywords.
Try more general keywords.
I told you my searches just don't seem to work.
RobertMacy
Apr 3, 2014, 11:31:45 AM4/3/14
to
On Thu, 03 Apr 2014 02:23:24 -0700, Jan Panteltje
<pNaonSt...@yahoo.com> wrote:
>> ...snip...
simply pick off the 'cosine' 'sine' carrier. Albeit sq wave. which most of
these mixers seem to prefer. 1MHz to 100MHz is not too difficult a range
for generating 'clean' LO's
I already have covered the cos/sin for the modulation bands.
But am concerned with that poor carrier suppression spec. Concerned, only
as thought experiment. Having 0.2% of unwanted may not be catastrophic,
don't know for certain. Just suspect, based upon previous system where
0.01% did cause problems that only suspect something as high as 0.2% will
also cause problems.
<pNaonSt...@yahoo.com> wrote:
>> ...snip...
> I forgot to add, I think for frequencies below 100 MHz most goes digital,
> it is hard to make the RF 90 degrees phase shift on chip over 100 kHz to
> 100 MHz I'd think,
> maybe Jimmy knows how to do that.
> It is much easier to set up some flipflops, and, if you must, use analog
> switches to generate QAM.
> I havedunnit for 4.43 MHz PAL modulator that way (UV).
LO's should be simple. I envisioned a four quadrant digital waveform and
> it is hard to make the RF 90 degrees phase shift on chip over 100 kHz to
> 100 MHz I'd think,
> maybe Jimmy knows how to do that.
> It is much easier to set up some flipflops, and, if you must, use analog
> switches to generate QAM.
> I havedunnit for 4.43 MHz PAL modulator that way (UV).
simply pick off the 'cosine' 'sine' carrier. Albeit sq wave. which most of
these mixers seem to prefer. 1MHz to 100MHz is not too difficult a range
for generating 'clean' LO's
I already have covered the cos/sin for the modulation bands.
But am concerned with that poor carrier suppression spec. Concerned, only
as thought experiment. Having 0.2% of unwanted may not be catastrophic,
don't know for certain. Just suspect, based upon previous system where
0.01% did cause problems that only suspect something as high as 0.2% will
also cause problems.
RobertMacy
Apr 3, 2014, 11:39:56 AM4/3/14
to
On Thu, 03 Apr 2014 05:50:39 -0700, Chris Jones <lugn...@spam.yahoo.com>
wrote:
> ...snip to keep Aioe happy
Have you simulated using LTspice like and/or Matlab like tools?
I've added a feature to our implementation of LTspice, called .tranoise,
which combines .tran and .noise analysis into a single operation. I've
been using it to determine the best ways to use the MC1496 mixer and
really find what happens to all that johnson noise and junction noise and
cross products and 1/f from the OpAmps, etc etc
wrote:
> ...snip to keep Aioe happy
> Some (very linear) gain ahead of the mixer might be needed to overwhelm
> any extra thermal noise from the input and low-pass filter that would
> get mixed into the IF band due to the shaped noise components of the LO
> signal, however it may also be feasible to reduce that noise by making
> sure that the mixer input is presented by the low-pass filter with an
> impedance having a very small real part (for low Johnson noise voltage)
> or extremely high impedance (for low Johnson noise current) at those
> (out of band) frequencies where the shaped noise is strong in the LO
> bitstream.
>
> I wonder if this is an established technique.
>
> Chris
>
>
Do you work in this area?
> any extra thermal noise from the input and low-pass filter that would
> get mixed into the IF band due to the shaped noise components of the LO
> signal, however it may also be feasible to reduce that noise by making
> sure that the mixer input is presented by the low-pass filter with an
> impedance having a very small real part (for low Johnson noise voltage)
> or extremely high impedance (for low Johnson noise current) at those
> (out of band) frequencies where the shaped noise is strong in the LO
> bitstream.
>
> I wonder if this is an established technique.
>
> Chris
>
>
Have you simulated using LTspice like and/or Matlab like tools?
I've added a feature to our implementation of LTspice, called .tranoise,
which combines .tran and .noise analysis into a single operation. I've
been using it to determine the best ways to use the MC1496 mixer and
really find what happens to all that johnson noise and junction noise and
cross products and 1/f from the OpAmps, etc etc
Jan Panteltje
Apr 3, 2014, 11:45:58 AM4/3/14
to
On a sunny day (Thu, 03 Apr 2014 08:20:56 -0700) it happened RobertMacy
<robert...@gmail.com> wrote in <op.xdqx86kg2cx0wh@ajm>:
I have the datasheet, took 10 seconds to find it,
dunno from where.
<robert...@gmail.com> wrote in <op.xdqx86kg2cx0wh@ajm>:
dunno from where.
Phil Hobbs
Apr 3, 2014, 11:50:25 AM4/3/14
to
RobertMacy
Apr 3, 2014, 11:56:48 AM4/3/14
to
On Thu, 03 Apr 2014 08:45:58 -0700, Jan Panteltje
<pNaonSt...@yahoo.com> wrote:
>> ...snip...
<pNaonSt...@yahoo.com> wrote:
>> ...snip...
>
> I have the datasheet, took 10 seconds to find it,
> dunno from where.
would you send a copy [assuming pdf] to me?
> I have the datasheet, took 10 seconds to find it,
> dunno from where.
Jan Panteltje
Apr 3, 2014, 12:33:18 PM4/3/14
to
On a sunny day (Thu, 03 Apr 2014 08:56:48 -0700) it happened RobertMacy
<robert...@gmail.com> wrote in <op.xdqzwyts2cx0wh@ajm>:
Send to the email in the header.
if you do not have it by now, let me know.
<robert...@gmail.com> wrote in <op.xdqzwyts2cx0wh@ajm>:
if you do not have it by now, let me know.
Tim Wescott
Apr 3, 2014, 2:21:45 PM4/3/14
to
On Wed, 02 Apr 2014 17:26:21 -0700, RobertMacy wrote:
> On Wed, 02 Apr 2014 15:36:36 -0700, Tim Wescott
> <t...@seemywebsite.please>
> wrote:
>
>>> ...snip...
>>
>> If you're willing to bend on your drive and modulation levels (I assume
>> that by 'drive' you mean 'signal' and by 'modulation' you mean local
>> oscillator), then MiniCircuits has some diode ring mixers that might
>> make you very happy.
>>
>> They're passive devices, so their "operating power" comes from the
>> local oscillator drive.
>>
>> Figure on over 10dBm of local oscillator drive, and (probably)
>> significantly less than 0dBm of signal to get linearity that this
>> ex-Ham operator/receiver builder considers "very linear".
>>
>>
> Forgot all about minicircuits! they make great stuff. I'll take a look,
> thanks.
>
> Not sure I followed your definitions regarding signal etc, but I
> understand the principles behind using diode H bridges as a mixer.
> Actually designed one from discrete components that went into autopilot
> electronics modules years back.
>
> The diode mixer might have better specs than the MC1496 topology, I'm
> not skilled enough in chip design to be able to answer that one. My gut
> feeling is that it will be hard to beat a well laid out MC1496
> structure.
> *IF* someone makes one.
For "very linear" the diode-ring mixers (and, if you really want to work
at it, FET-ring mixers) have been recognized to be superior to Gilbert
cell mixers for years and years. I doubt that has changed, although
knowing whether a diode ring is superior to a FET ring is beyond my pay
scale at the moment.
Gilbert cell mixers are better if you need to build an inexpensive small
circuit that doesn't consume much power, though.
--
Tim Wescott
Wescott Design Services
http://www.wescottdesign.com
> On Wed, 02 Apr 2014 15:36:36 -0700, Tim Wescott
> <t...@seemywebsite.please>
> wrote:
>
>>> ...snip...
>>
>> If you're willing to bend on your drive and modulation levels (I assume
>> that by 'drive' you mean 'signal' and by 'modulation' you mean local
>> oscillator), then MiniCircuits has some diode ring mixers that might
>> make you very happy.
>>
>> They're passive devices, so their "operating power" comes from the
>> local oscillator drive.
>>
>> Figure on over 10dBm of local oscillator drive, and (probably)
>> significantly less than 0dBm of signal to get linearity that this
>> ex-Ham operator/receiver builder considers "very linear".
>>
>>
> Forgot all about minicircuits! they make great stuff. I'll take a look,
> thanks.
>
> Not sure I followed your definitions regarding signal etc, but I
> understand the principles behind using diode H bridges as a mixer.
> Actually designed one from discrete components that went into autopilot
> electronics modules years back.
>
> The diode mixer might have better specs than the MC1496 topology, I'm
> not skilled enough in chip design to be able to answer that one. My gut
> feeling is that it will be hard to beat a well laid out MC1496
> structure.
> *IF* someone makes one.
For "very linear" the diode-ring mixers (and, if you really want to work
at it, FET-ring mixers) have been recognized to be superior to Gilbert
cell mixers for years and years. I doubt that has changed, although
knowing whether a diode ring is superior to a FET ring is beyond my pay
scale at the moment.
Gilbert cell mixers are better if you need to build an inexpensive small
circuit that doesn't consume much power, though.
--
Tim Wescott
Wescott Design Services
http://www.wescottdesign.com
Phil Hobbs
Apr 3, 2014, 6:21:33 PM4/3/14
to
Phil Hobbs
Apr 3, 2014, 6:21:40 PM4/3/14
to
On 4/3/2014 11:39 AM, RobertMacy wrote:
Clifford Heath
Apr 3, 2014, 8:17:48 PM4/3/14
to
On 03/04/14 17:16, Tauno Voipio wrote:
> On 3.4.14 03:00, RobertMacy wrote:
>> Is the bandwidth too much for the baseband, or for the carrier?
>
> Both, but for different reasons:
>
> - Creating the analytic signal (two copies at 90 degrees difference)
> makes the requirement of limited baseband. It must be limited at
> the low end to make the phase shifting possible, 0 Hz does not go.
> The minimum delay for the filters is a quarter wave at he lowest
> frequency.
You can get pretty accurate 90 degree phase shifts across a decade of
> On 3.4.14 03:00, RobertMacy wrote:
>> Is the bandwidth too much for the baseband, or for the carrier?
>
> Both, but for different reasons:
>
> - Creating the analytic signal (two copies at 90 degrees difference)
> makes the requirement of limited baseband. It must be limited at
> the low end to make the phase shifting possible, 0 Hz does not go.
> The minimum delay for the filters is a quarter wave at he lowest
> frequency.
frequency using a fourth-order quadrature hybrid, but I really wonder if
it wouldn't be easier to generate 5.1-15MHz SSB and downconvert it to
0.1-10MHz.
I'm playing with a first-order QH for antenna phasing in a DF system,
which the simulations say will cancel well from 2.5-5MHz, with a += 20
degree error (that can be compensated because it's a fixed function of
frequency). A fourth-order version should work similarly from 3-30MHz
using the same antennae, as long as the RF levels are matched (the
current design problem).
Robert, I'm making a separate post about an oscillator simulation
problem I'm having with LTSpice and the Intersil CA3046 model. If you
have a better model you can send to clifford dot heath at gmail, I think
that might solve my problem. otherwise I'd appreciate your comments on
that thread (about to post it).
Clifford Heath.
Clifford Heath
Apr 3, 2014, 8:29:11 PM4/3/14
to
On 03/04/14 11:26, RobertMacy wrote:
> On Wed, 02 Apr 2014 15:36:36 -0700, Tim Wescott
> On Wed, 02 Apr 2014 15:36:36 -0700, Tim Wescott
>> Figure on over 10dBm of local oscillator drive, and (probably)
>> significantly less than 0dBm of signal to get linearity that this ex-Ham
>> operator/receiver builder considers "very linear".
>> significantly less than 0dBm of signal to get linearity that this ex-Ham
>> operator/receiver builder considers "very linear".
> Not sure I followed your definitions regarding signal etc, but I
> understand the principles behind using diode H bridges as a mixer.
Robert,
> understand the principles behind using diode H bridges as a mixer.
The 10dBm drive means that the diodes go through the partly-on
(non-linear) stage quickly, so they spend most of the time either fully
on or fully off.
The <0dBm requirement for the signal is because you don't ever want the
signal to turn on (even partly) one of the "off" diodes, nor to affect
the timing of its transition through the non-linear region.
Clifford Heath.
Clifford Heath
Apr 3, 2014, 8:36:36 PM4/3/14
to
On 03/04/14 11:13, RobertMacy wrote:
> Thanks to Jim Thompson who kindly sent me NPN CA3046 'chip sized'
> models, I've characterized the simplified MC1496 [as it appears in the
> data sheet/app note]. I now have a complete LTspice model now.
Trouble with the *3046 is you don't get enough transistors for a Gilbert
cell in one device, so you may have matching problems. If you already
have an input transformer, you can make is centre-tapped and feed the
emitters of two pairs, but then you're relying on transformer matching
not silicon matching again.
Effectively the same silicon (but bonded out differently) as the LM3046
(like CA3046) is available as a pre-wired Gilbert cell in the LM3101. If
you don't mind paying $5 you can also get the rocket-ship HFA3101, which
uses 6GHz transistors and is designed for 900MHz mixers.
The mixer in the SA601A/SA612A would probably also suit your purposes,
works to VHF and is <$1 in singles.
Clifford Heath
> Thanks to Jim Thompson who kindly sent me NPN CA3046 'chip sized'
> models, I've characterized the simplified MC1496 [as it appears in the
> data sheet/app note]. I now have a complete LTspice model now.
Trouble with the *3046 is you don't get enough transistors for a Gilbert
cell in one device, so you may have matching problems. If you already
have an input transformer, you can make is centre-tapped and feed the
emitters of two pairs, but then you're relying on transformer matching
not silicon matching again.
Effectively the same silicon (but bonded out differently) as the LM3046
(like CA3046) is available as a pre-wired Gilbert cell in the LM3101. If
you don't mind paying $5 you can also get the rocket-ship HFA3101, which
uses 6GHz transistors and is designed for 900MHz mixers.
The mixer in the SA601A/SA612A would probably also suit your purposes,
works to VHF and is <$1 in singles.
Clifford Heath
Bill Sloman
Apr 3, 2014, 9:04:19 PM4/3/14
to
On Friday, 4 April 2014 11:36:36 UTC+11, Clifford Heath wrote:
> On 03/04/14 11:13, RobertMacy wrote:
>
> > Thanks to Jim Thompson who kindly sent me NPN CA3046 'chip sized'
> > models, I've characterized the simplified MC1496 [as it appears in the
> > data sheet/app note]. I now have a complete LTspice model now.
>
> Trouble with the *3046 is you don't get enough transistors for a Gilbert
> cell in one device, so you may have matching problems. If you already
> have an input transformer, you can make is centre-tapped and feed the
> emitters of two pairs, but then you're relying on transformer matching
> not silicon matching again.
Transformer matching can be spectacularly good. Bifilar windings (and transmission line transformers wound with coax are a particularly fine example of the breed) offer matching to one part in a billion. Ratio transformers (where you have more than two wires in the wound bundle) can be good to one part in ten million.
> On 03/04/14 11:13, RobertMacy wrote:
>
> > Thanks to Jim Thompson who kindly sent me NPN CA3046 'chip sized'
> > models, I've characterized the simplified MC1496 [as it appears in the
> > data sheet/app note]. I now have a complete LTspice model now.
>
> Trouble with the *3046 is you don't get enough transistors for a Gilbert
> cell in one device, so you may have matching problems. If you already
> have an input transformer, you can make is centre-tapped and feed the
> emitters of two pairs, but then you're relying on transformer matching
> not silicon matching again.
http://www.amazon.com/Coaxial-AC-Bridges-B-Kibble/dp/0852743890/ref=sr_1_1?s=books&ie=UTF8&qid=1396573343&sr=1-1&keywords=Coaxial+AC+Bridges
<snipped good stuff>
--
Bill Sloman, Sydney
Clifford Heath
Apr 3, 2014, 9:38:59 PM4/3/14
to
On 04/04/14 02:20, RobertMacy wrote:
>> Atmel U2790B
> This is why searches don't work well. Googled Atmel U2790B
Use octopart.com, not google. Google doesn't know how to sub on/off
>> Atmel U2790B
> This is why searches don't work well. Googled Atmel U2790B
trailing letters on chip names.
In Chrome, after having used octopart, I just get a new tab, type
oct<tab>CHIP<enter> into the URL box. For any chip that's still
available, you get prices and almost always a datasheet link in one go.
Chris Jones
Apr 3, 2014, 10:11:36 PM4/3/14
to
On 04/04/2014 11:36, Clifford Heath wrote:
> On 03/04/14 11:13, RobertMacy wrote:
>> Thanks to Jim Thompson who kindly sent me NPN CA3046 'chip sized'
>> models, I've characterized the simplified MC1496 [as it appears in the
>> data sheet/app note]. I now have a complete LTspice model now.
>
> Trouble with the *3046 is you don't get enough transistors for a Gilbert
> cell in one device, so you may have matching problems.
[snip]
> On 03/04/14 11:13, RobertMacy wrote:
>> Thanks to Jim Thompson who kindly sent me NPN CA3046 'chip sized'
>> models, I've characterized the simplified MC1496 [as it appears in the
>> data sheet/app note]. I now have a complete LTspice model now.
>
> Trouble with the *3046 is you don't get enough transistors for a Gilbert
> cell in one device, so you may have matching problems.
You might want to look at the Intersil HFA3101
http://www.intersil.com/content/dam/Intersil/documents/hfa3/hfa3101.pdf
Also similar (but with no complete schematic), LT5560 and AD8343.
Chris
Chris Jones
Apr 3, 2014, 10:12:20 PM4/3/14
to
(perhaps about as much as I can) about stuff I worked on here:
www.chrisj.org
I have simulated mixers generally, a lot, using ADICE which is by far
the nicest simulator I have ever used, but which I sadly don't have
access to any longer. LTSpice is now the best thing I have access to and
is very useful.
I haven't simulated that scheme I discussed above though.
From what you say it sounds like you are a developer of LTSpice which
would mean you work at LT, is that right?
Chris
RobertMacy
Apr 3, 2014, 10:47:16 PM4/3/14
to
On Thu, 03 Apr 2014 09:33:18 -0700, Jan Panteltje
<pNaonSt...@yahoo.com> wrote:
>> ...snip...
<pNaonSt...@yahoo.com> wrote:
>> ...snip...
> Send to the email in the header.
> if you do not have it by now, let me know.
>
got it thanks, and I replied from that email address, too
> if you do not have it by now, let me know.
>
RobertMacy
Apr 3, 2014, 10:55:37 PM4/3/14
to
On Thu, 03 Apr 2014 15:21:33 -0700, Phil Hobbs <ho...@electrooptical.net>
wrote:
>> ...snip....
slow things down a bit. around 5 to 10 times longer than the 'normal'
speed of .tran
wrote:
>> ...snip....
>
> How does it work?
>
> Cheers
>
> Phil Hobbs
>
>
very well. I can't get 'inside' the LTspice machine so .tranoise tends to
> How does it work?
>
> Cheers
>
> Phil Hobbs
>
>
slow things down a bit. around 5 to 10 times longer than the 'normal'
speed of .tran
josephkk
Apr 4, 2014, 2:40:51 AM4/4/14
to
On Wed, 02 Apr 2014 12:18:50 -0700, RobertMacy <robert...@gmail.com>
wrote:
>Anybody have experience/recommend an extremely linear double balanced
>mixer?
>
>DRIVE around 0+ dBm:
>prefer unbal single drive, but can live with bal drive.
>0.1MHz < fc < 100MHz
>
>MODULATION 0+dBm
>prefer unbal modulation, but can live with bal drive.
>also can live with low end of 10kHz instead of DC
>DC < fm < 100kHz
>
>Need to create a lower SSB system WITH suppressed carrier, so linearity is
>important.
>
wrote:
>Anybody have experience/recommend an extremely linear double balanced
>mixer?
>
>DRIVE around 0+ dBm:
>prefer unbal single drive, but can live with bal drive.
>0.1MHz < fc < 100MHz
>
>MODULATION 0+dBm
>prefer unbal modulation, but can live with bal drive.
>also can live with low end of 10kHz instead of DC
>DC < fm < 100kHz
>
>Need to create a lower SSB system WITH suppressed carrier, so linearity is
>important.
>
>Of course need the reverse, 'receiver' function, too, so component should
>not just be labeled "upconverter", or such.
>
>
>So far have identified Linear, Analog Devices, and TI as potential
>suppliers, I historically shy away from Maxim.
>But after sitting an hour at those websites have given up doing any
>effective searches there.
>
>Also, there might be some serious contenders who have real 'knock your
>socks off' niche components for this application.
>
Well most double balanced mixers are very similar to Gilbert cell
>not just be labeled "upconverter", or such.
>
>
>So far have identified Linear, Analog Devices, and TI as potential
>suppliers, I historically shy away from Maxim.
>But after sitting an hour at those websites have given up doing any
>effective searches there.
>
>Also, there might be some serious contenders who have real 'knock your
>socks off' niche components for this application.
>
multipliers. The very wide bandwidth makes phasing type twin mixers
impossible for conversion to SSB. The remaining choice it to mix to a
much higher frequency (1 GHz) and use a SAW filter or similar to split the
sidebands and mix down again with heavy filtering to do the image
rejection.
You need some more slack in the bandwidths to make this reasonable.
?-)
Tauno Voipio
Apr 4, 2014, 2:19:00 AM4/4/14
to
On 4.4.14 03:17, Clifford Heath wrote:
> On 03/04/14 17:16, Tauno Voipio wrote:
>> On 3.4.14 03:00, RobertMacy wrote:
>>> Is the bandwidth too much for the baseband, or for the carrier?
>>
>> Both, but for different reasons:
>>
>> - Creating the analytic signal (two copies at 90 degrees difference)
>> makes the requirement of limited baseband. It must be limited at
>> the low end to make the phase shifting possible, 0 Hz does not go.
>> The minimum delay for the filters is a quarter wave at he lowest
>> frequency.
>
> You can get pretty accurate 90 degree phase shifts across a decade of
> frequency using a fourth-order quadrature hybrid, but I really wonder if
> it wouldn't be easier to generate 5.1-15MHz SSB and downconvert it to
> 0.1-10MHz.
We still have the problem of low frequencies. The OP wanted a band
> On 03/04/14 17:16, Tauno Voipio wrote:
>> On 3.4.14 03:00, RobertMacy wrote:
>>> Is the bandwidth too much for the baseband, or for the carrier?
>>
>> Both, but for different reasons:
>>
>> - Creating the analytic signal (two copies at 90 degrees difference)
>> makes the requirement of limited baseband. It must be limited at
>> the low end to make the phase shifting possible, 0 Hz does not go.
>> The minimum delay for the filters is a quarter wave at he lowest
>> frequency.
>
> You can get pretty accurate 90 degree phase shifts across a decade of
> frequency using a fourth-order quadrature hybrid, but I really wonder if
> it wouldn't be easier to generate 5.1-15MHz SSB and downconvert it to
> 0.1-10MHz.
starting from DC, which is IMHO impossible. Even with a low limit
of 10 Hz, the required bandwidth exceeds by a wide margin those
which have been realized. I have the feeling that it is in practice
impossible to get analog components that are matched well enough for
the entire baseband width. Digitally, it may be possible if
the band is limited at the low end, but the filters will be very
long to provide the necessary delay difference at the low end
of the range with the sampling rate needed for the high end.
There are other methods with e.g. matched FIR filters, polyphase RC
nets and mixing in the Weaver method.
For the output band, the standard method is to generate the signal
at an high IF well above the upper limit of the output band and
then downconverting. There is an advantage to the final mixer and
output filter to go above 3 times the high end of the output
band, e.g. 30.1 to 40 MHz for 0.1 to 10 MHz output.
--
-TV
Tauno Voipio
Apr 4, 2014, 2:23:59 AM4/4/14
to
On 3.4.14 21:21, Tim Wescott wrote:
.
>
> For "very linear" the diode-ring mixers (and, if you really want to work
> at it, FET-ring mixers) have been recognized to be superior to Gilbert
> cell mixers for years and years. I doubt that has changed, although
> knowing whether a diode ring is superior to a FET ring is beyond my pay
> scale at the moment.
>
> Gilbert cell mixers are better if you need to build an inexpensive small
> circuit that doesn't consume much power, though.
The last advance in very linear mixers seems to be using CMOS analog
.
>
> For "very linear" the diode-ring mixers (and, if you really want to work
> at it, FET-ring mixers) have been recognized to be superior to Gilbert
> cell mixers for years and years. I doubt that has changed, although
> knowing whether a diode ring is superior to a FET ring is beyond my pay
> scale at the moment.
>
> Gilbert cell mixers are better if you need to build an inexpensive small
> circuit that doesn't consume much power, though.
switches to do the switching instead of transistors or diodes.
Have a look at <http://www.norcalqrp.org/files/tayloe_mixer_x3a.pdf>.
--
-TV
Jan Panteltje
Apr 4, 2014, 6:23:31 AM4/4/14
to
On a sunny day (Thu, 03 Apr 2014 19:47:16 -0700) it happened RobertMacy
<robert...@gmail.com> wrote in <op.xdrt02pg2cx0wh@ajm>:
No reply received as of this morning.
try
double_balanced_mixer at p a n t e l t j e dot com
obvious space removed and at and dot replaced by .
<robert...@gmail.com> wrote in <op.xdrt02pg2cx0wh@ajm>:
try
double_balanced_mixer at p a n t e l t j e dot com
obvious space removed and at and dot replaced by .
George Herold
Apr 4, 2014, 9:04:18 AM4/4/14
to
On Thursday, April 3, 2014 2:21:45 PM UTC-4, Tim Wescott wrote:
> On Wed, 02 Apr 2014 17:26:21 -0700, RobertMacy wrote:
>
<snip>
> On Wed, 02 Apr 2014 17:26:21 -0700, RobertMacy wrote:
>
>
> For "very linear" the diode-ring mixers (and, if you really want to work
> at it, FET-ring mixers) have been recognized to be superior to Gilbert
> cell mixers for years and years. I doubt that has changed, although
> knowing whether a diode ring is superior to a FET ring is beyond my pay
> scale at the moment.
Tim, do you have any idea why that is. "In theory" I would think the Gilbert cell would be better, 'cause it doesn't have all the odd harmonics that the switchers have. In practice I can imagine that switching is much easier than multiplying.
> For "very linear" the diode-ring mixers (and, if you really want to work
> at it, FET-ring mixers) have been recognized to be superior to Gilbert
> cell mixers for years and years. I doubt that has changed, although
> knowing whether a diode ring is superior to a FET ring is beyond my pay
> scale at the moment.
George H.
amdx
Apr 4, 2014, 10:26:51 AM4/4/14
to
On 4/2/2014 2:18 PM, RobertMacy wrote:
> Anybody have experience/recommend an extremely linear double balanced
> mixer?
>
> DRIVE around 0+ dBm:
> prefer unbal single drive, but can live with bal drive.
> 0.1MHz < fc < 100MHz
>
> MODULATION 0+dBm
> prefer unbal modulation, but can live with bal drive.
> also can live with low end of 10kHz instead of DC
> DC < fm < 100kHz
>
> Need to create a lower SSB system WITH suppressed carrier, so linearity
> is important.
>
> Of course need the reverse, 'receiver' function, too, so component
> should not just be labeled "upconverter", or such.
>
>
> So far have identified Linear, Analog Devices, and TI as potential
> suppliers, I historically shy away from Maxim.
> But after sitting an hour at those websites have given up doing any
> effective searches there.
>
> Also, there might be some serious contenders who have real 'knock your
> socks off' niche components for this application.
>
>
Sounds like you want a single chip, But if not, here is the design I
> Anybody have experience/recommend an extremely linear double balanced
> mixer?
>
> DRIVE around 0+ dBm:
> prefer unbal single drive, but can live with bal drive.
> 0.1MHz < fc < 100MHz
>
> MODULATION 0+dBm
> prefer unbal modulation, but can live with bal drive.
> also can live with low end of 10kHz instead of DC
> DC < fm < 100kHz
>
> Need to create a lower SSB system WITH suppressed carrier, so linearity
> is important.
>
> Of course need the reverse, 'receiver' function, too, so component
> should not just be labeled "upconverter", or such.
>
>
> So far have identified Linear, Analog Devices, and TI as potential
> suppliers, I historically shy away from Maxim.
> But after sitting an hour at those websites have given up doing any
> effective searches there.
>
> Also, there might be some serious contenders who have real 'knock your
> socks off' niche components for this application.
>
>
always think about think about when I hear high performance , wide
dynamic range receiver.
http://www.robkalmeijer.nl/techniek/electronica/radiotechniek/hambladen/qst/1993/02/page23/
In following years new quad mixer chips have come out for further
improvements, that replace the SD8901.
I don't know if the later mixer chips make it to 100MHz. I suspect
they do.
I originally saw this in QST, I think this might be the author of that
article.
Mikek
Phil Hobbs
Apr 4, 2014, 10:27:48 AM4/4/14
to
work."
Doing that in any generality for strongly nonlinear circuits is an
intrinsically difficult problem, or so I would have thought.
Phil Hobbs
Apr 4, 2014, 10:34:56 AM4/4/14
to
amplifiers were using CMOS muxes as mixers in the 1970s, though discrete
MOSFET bridges were more common. For instance, here's the schematic of
a PARC 128A lock-in from 1983:
http://servv89pn0aj.sn.sourcedns.com/~gbpprorg/mil/cavity/lock/PAR_128_Lock-In_Amplifier.pdf
(See P. 54 of the service manual.)
Cheers
Phil Hobbs
Phil Hobbs
Apr 4, 2014, 10:45:21 AM4/4/14
to
On 04/04/2014 09:04 AM, George Herold wrote:
> On Thursday, April 3, 2014 2:21:45 PM UTC-4, Tim Wescott wrote:
>> On Wed, 02 Apr 2014 17:26:21 -0700, RobertMacy wrote:
>>
> <snip>
>>
>> For "very linear" the diode-ring mixers (and, if you really want to
>> work at it, FET-ring mixers) have been recognized to be superior to
>> Gilbert cell mixers for years and years. I doubt that has changed,
>> although knowing whether a diode ring is superior to a FET ring is
>> beyond my pay scale at the moment.
> Tim, do you have any idea why that is. "In theory" I would think the
> Gilbert cell would be better, 'cause it doesn't have all the odd
> harmonics that the switchers have. In practice I can imagine that
> switching is much easier than multiplying.
>
> George H.
I don't think it's the current steering (upper) stage of the Gilbert
> On Thursday, April 3, 2014 2:21:45 PM UTC-4, Tim Wescott wrote:
>> On Wed, 02 Apr 2014 17:26:21 -0700, RobertMacy wrote:
>>
> <snip>
>>
>> For "very linear" the diode-ring mixers (and, if you really want to
>> work at it, FET-ring mixers) have been recognized to be superior to
>> Gilbert cell mixers for years and years. I doubt that has changed,
>> although knowing whether a diode ring is superior to a FET ring is
>> beyond my pay scale at the moment.
> Tim, do you have any idea why that is. "In theory" I would think the
> Gilbert cell would be better, 'cause it doesn't have all the odd
> harmonics that the switchers have. In practice I can imagine that
> switching is much easier than multiplying.
>
> George H.
cell that's the problem--those devices are almost always ON or OFF.
The main issue is that the RF port drives a BJT diff pair that's in its
linear range, so that its output current goes as
tanh(e dV_BE/(2kT)), whereas in a diode mixer the RF port's load is the
input resistance of the first IF, plus the switch resistances. Diode
mixers work a lot better into a constant 50 ohms for this reason.
(Using a diplexer instead of a reflective filter after the mixer is a
big win for linearity, and this is why.)
A strong interfering signal will drive the RF diff pair of the Gilbert
cell nonlinear just the way it does an ordinary BJT amplifier. Running
the diff pair as a current mirror, as in the LM13700 OTA, nominally
turns the tanh characteristic back into a linear one, but that trick
isn't often used at RF, I don't think.
Phil Hobbs
Apr 4, 2014, 10:48:13 AM4/4/14
to
keywords as well as part numbers.
The SEO and redirect parasites have made Google dramatically less useful
in the last couple of years.
Tim Wescott
Apr 4, 2014, 11:30:36 AM4/4/14
to
So in a sense it's a FET-ring mixer that someone found inside of a little
bit of plastic.
--
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com
RobertMacy
Apr 4, 2014, 12:12:40 PM4/4/14
to
On Fri, 04 Apr 2014 03:23:31 -0700, Jan Panteltje <pant...@yahoo.com>
wrote:
> double_balanced_mixer at p a n t e l t j e dot com
embarrassed, I did NOT send the thank you, just did.
wrote:
> double_balanced_mixer at p a n t e l t j e dot com
RobertMacy
Apr 4, 2014, 12:47:52 PM4/4/14
to
On Thu, 03 Apr 2014 23:19:00 -0700, Tauno Voipio
<tauno....@notused.fi.invalid> wrote:
>> ...snip....
with accuracies better than tenths of a degree, over the band of 10Hz to
100kHz. Supplying the required modulation will NOT be the problem. My
problem is still with the carrier suppression and upper harmonic
generation that comes out of a mixer, instead of out of a multiplier
circuit. Perhaps what I really need is a 'perfect' analog multiplier that
can operate up to 100MHz, similar architecture to the MC1496, but better
specs??
'Interfering higher tones' is a major concern, need them down more than
90dBfs [would like more than 100dBfs]. It is surprising how difficult it
is to make physically realizable filters that can knock those 'specced'
tones down enough.
Two 2N3904's make a great 5 pole slightly ripply filter that operates up
to just beyond 10MHz [not 100MHz], but the limit of rejection in the
rolloff is ALWAYS the number of poles, no matter how you place them.
...Suppose could put two in series. It's too difficult to get all the
poles placed properly at around 7 poles and upwards. Need that broadband
capability, else would play games with notching out the tones. With active
circuitry, might be possible to make a 'tracking' filter that follows the
carrier. Not going to go there, yet.
<tauno....@notused.fi.invalid> wrote:
>> ...snip....
> We still have the problem of low frequencies. The OP wanted a band
> starting from DC, which is IMHO impossible. Even with a low limit
> of 10 Hz, the required bandwidth exceeds by a wide margin those
> which have been realized. I have the feeling that it is in practice
> impossible to get analog components that are matched well enough for
> the entire baseband width. Digitally, it may be possible if
> the band is limited at the low end, but the filters will be very
> long to provide the necessary delay difference at the low end
> of the range with the sampling rate needed for the high end.
>
> There are other methods with e.g. matched FIR filters, polyphase RC
> nets and mixing in the Weaver method.
>
> For the output band, the standard method is to generate the signal
> at an high IF well above the upper limit of the output band and
> then downconverting. There is an advantage to the final mixer and
> output filter to go above 3 times the high end of the output
> band, e.g. 30.1 to 40 MHz for 0.1 to 10 MHz output.
>
For modulation, I have the required 90 degree phase shift covered often
> starting from DC, which is IMHO impossible. Even with a low limit
> of 10 Hz, the required bandwidth exceeds by a wide margin those
> which have been realized. I have the feeling that it is in practice
> impossible to get analog components that are matched well enough for
> the entire baseband width. Digitally, it may be possible if
> the band is limited at the low end, but the filters will be very
> long to provide the necessary delay difference at the low end
> of the range with the sampling rate needed for the high end.
>
> There are other methods with e.g. matched FIR filters, polyphase RC
> nets and mixing in the Weaver method.
>
> For the output band, the standard method is to generate the signal
> at an high IF well above the upper limit of the output band and
> then downconverting. There is an advantage to the final mixer and
> output filter to go above 3 times the high end of the output
> band, e.g. 30.1 to 40 MHz for 0.1 to 10 MHz output.
>
with accuracies better than tenths of a degree, over the band of 10Hz to
100kHz. Supplying the required modulation will NOT be the problem. My
problem is still with the carrier suppression and upper harmonic
generation that comes out of a mixer, instead of out of a multiplier
circuit. Perhaps what I really need is a 'perfect' analog multiplier that
can operate up to 100MHz, similar architecture to the MC1496, but better
specs??
'Interfering higher tones' is a major concern, need them down more than
90dBfs [would like more than 100dBfs]. It is surprising how difficult it
is to make physically realizable filters that can knock those 'specced'
tones down enough.
Two 2N3904's make a great 5 pole slightly ripply filter that operates up
to just beyond 10MHz [not 100MHz], but the limit of rejection in the
rolloff is ALWAYS the number of poles, no matter how you place them.
...Suppose could put two in series. It's too difficult to get all the
poles placed properly at around 7 poles and upwards. Need that broadband
capability, else would play games with notching out the tones. With active
circuitry, might be possible to make a 'tracking' filter that follows the
carrier. Not going to go there, yet.
Phil Hobbs
Apr 4, 2014, 12:48:17 PM4/4/14
to
RobertMacy
Apr 4, 2014, 12:57:24 PM4/4/14
to
On Thu, 03 Apr 2014 23:40:51 -0700, josephkk
<joseph_...@sbcglobal.net> wrote:
>> ...snip...
have some advantages.
The required 90 degree phase shift in the modulation is covered. And, it's
pretty easy to supply LO with quadrature, too. It's the 'balance' in the
mixers I need to get that pesky carrier tone and 3rd harmonics down.
100MHz has 10ns period. 0.1nS is almost 4 degrees phase shift! Somehow
will have to make the system accomodate that kind of sloppy action, too
<joseph_...@sbcglobal.net> wrote:
>> ...snip...
> Well most double balanced mixers are very similar to Gilbert cell
> multipliers. The very wide bandwidth makes phasing type twin mixers
> impossible for conversion to SSB. The remaining choice it to mix to a
> much higher frequency (1 GHz) and use a SAW filter or similar to split
> the
> sidebands and mix down again with heavy filtering to do the image
> rejection.
>
> You need some more slack in the bandwidths to make this reasonable.
>
> ?-)
>
Forgot all about SAW filtering. I know it's fixed filtering, but might
> multipliers. The very wide bandwidth makes phasing type twin mixers
> impossible for conversion to SSB. The remaining choice it to mix to a
> much higher frequency (1 GHz) and use a SAW filter or similar to split
> the
> sidebands and mix down again with heavy filtering to do the image
> rejection.
>
> You need some more slack in the bandwidths to make this reasonable.
>
> ?-)
>
have some advantages.
The required 90 degree phase shift in the modulation is covered. And, it's
pretty easy to supply LO with quadrature, too. It's the 'balance' in the
mixers I need to get that pesky carrier tone and 3rd harmonics down.
100MHz has 10ns period. 0.1nS is almost 4 degrees phase shift! Somehow
will have to make the system accomodate that kind of sloppy action, too
RobertMacy
Apr 4, 2014, 1:00:59 PM4/4/14
to
On Fri, 04 Apr 2014 09:48:17 -0700, Phil Hobbs
<pcdhSpamM...@electrooptical.net> wrote:
> On 04/04/2014 12:47 PM, RobertMacy wrote:
>> ...snip....
<pcdhSpamM...@electrooptical.net> wrote:
> On 04/04/2014 12:47 PM, RobertMacy wrote:
>> Two 2N3904's make a great 5 pole slightly ripply filter that operates up
>> to just beyond 10MHz [not 100MHz], but the limit of rejection in the
>> rolloff is ALWAYS the number of poles, no matter how you place them.
>
> If only. The real limit is usually due to strays.
>
> Cheers
>
> Phil Hobbs
True. but I've designed systems that can 'see' signals below Tempest
>> to just beyond 10MHz [not 100MHz], but the limit of rejection in the
>> rolloff is ALWAYS the number of poles, no matter how you place them.
>
> If only. The real limit is usually due to strays.
>
> Cheers
>
> Phil Hobbs
levels. And they thought they were 'silencing' their electronics! NOT!
RobertMacy
Apr 4, 2014, 1:10:44 PM4/4/14
to
On Fri, 04 Apr 2014 07:26:51 -0700, amdx <noj...@knology.net> wrote:
>> ...snip...
I tried to search for an 'original' SD8901 data sheet and all I get are
wannabees.
Calogic is referenced as the originator, but couldn't find them, either.
Any EXACT URL to get me a 'clean' copy of that data sheet?
>> ...snip...
> Sounds like you want a single chip, But if not, here is the design I
> always think about think about when I hear high performance , wide
> dynamic range receiver.
> http://www.robkalmeijer.nl/techniek/electronica/radiotechniek/hambladen/qst/1993/02/page23/
> In following years new quad mixer chips have come out for further
> improvements, that replace the SD8901.
> I don't know if the later mixer chips make it to 100MHz. I suspect
> they do.
> I originally saw this in QST, I think this might be the author of that
> article.
> Mikek
>
great article, thank you.
> always think about think about when I hear high performance , wide
> dynamic range receiver.
> http://www.robkalmeijer.nl/techniek/electronica/radiotechniek/hambladen/qst/1993/02/page23/
> In following years new quad mixer chips have come out for further
> improvements, that replace the SD8901.
> I don't know if the later mixer chips make it to 100MHz. I suspect
> they do.
> I originally saw this in QST, I think this might be the author of that
> article.
> Mikek
>
I tried to search for an 'original' SD8901 data sheet and all I get are
wannabees.
Calogic is referenced as the originator, but couldn't find them, either.
Any EXACT URL to get me a 'clean' copy of that data sheet?
piglet
Apr 4, 2014, 1:16:28 PM4/4/14
to
On 04/04/2014 17:47, RobertMacy wrote:
> My problem is still with the carrier suppression and upper...
Would it be possible to wrap a feedback loop
around the mixer balance, something like a
carrier psd to servo carrier residual to minimum?
Just an idle thought.
piglet
> My problem is still with the carrier suppression and upper...
Would it be possible to wrap a feedback loop
around the mixer balance, something like a
carrier psd to servo carrier residual to minimum?
Just an idle thought.
piglet
upsid...@downunder.com
Apr 4, 2014, 1:29:51 PM4/4/14
to
On Wed, 02 Apr 2014 12:18:50 -0700, RobertMacy
<robert...@gmail.com> wrote:
>Anybody have experience/recommend an extremely linear double balanced
>mixer?
>
>DRIVE around 0+ dBm:
>prefer unbal single drive, but can live with bal drive.
>0.1MHz < fc < 100MHz
>
>MODULATION 0+dBm
>prefer unbal modulation, but can live with bal drive.
>also can live with low end of 10kHz instead of DC
>DC < fm < 100kHz
>
>Need to create a lower SSB system WITH suppressed carrier, so linearity is
>important.
I originally misread that you intended to use the carrier as some kind
>Anybody have experience/recommend an extremely linear double balanced
>mixer?
>
>DRIVE around 0+ dBm:
>prefer unbal single drive, but can live with bal drive.
>0.1MHz < fc < 100MHz
>
>MODULATION 0+dBm
>prefer unbal modulation, but can live with bal drive.
>also can live with low end of 10kHz instead of DC
>DC < fm < 100kHz
>
>Need to create a lower SSB system WITH suppressed carrier, so linearity is
>important.
of pilot tone (based on the WITH in upper case :-), but apparently you
want just ordinary SSB (J3E), upper or lower side band ?
There are a lot of "zero-IF" chips using I/Q (de)modulators, e.g. in
various WLAN, DVB-T and in some equipment in various license-free
bands. These might be usable.
Use an image canceling I/Q mixer on a fixed carrier frequency, say 300
MHz to generate a LSB SSB signal and put a simple band pass filter
after it. A simple LC filter with Q=100 would have a 3 MHz bandwidth
and quite a lot of attenuation at 600 MHz (wrap around carrier 3rd
overtone lower side band) and 900 MHz (upper side band of carrier
harmonics).
Using a VCO/NCO tunable between 300.1 and 400 MHz and using a simple
(non-I/Q) mixer followed by a simple LPF at 200 MHz, this will produce
the 0.1 to 100 MHz signal you wanted.
Please note that due to the high side LO, the sidebands are swapped,
so in order to generate the USB signal, the original I/Q mixer needs
to generate LSB.
RobertMacy
Apr 4, 2014, 1:33:36 PM4/4/14
to
On Fri, 04 Apr 2014 10:16:28 -0700, piglet <erichp...@hotmail.com>
wrote:
great thought! thanks. once thought of, becomes so obvious.
Might be easy to implement, too.
wrote:
Might be easy to implement, too.
RobertMacy
Apr 4, 2014, 1:35:20 PM4/4/14
to
Dropping another 20 dB lowers that noise floor just that much more!
piglet
Apr 4, 2014, 1:56:25 PM4/4/14
to
On 04/04/2014 18:35, RobertMacy wrote:
> great thought! thanks. might be easy to implement, too.
> Dropping another 20 dB lowers that noise floor just that much more!
Grin! Just slide your appreciation under the door.
> great thought! thanks. might be easy to implement, too.
> Dropping another 20 dB lowers that noise floor just that much more!
piglet
upsid...@downunder.com
Apr 4, 2014, 1:56:58 PM4/4/14
to
On Fri, 04 Apr 2014 09:47:52 -0700, RobertMacy
<robert...@gmail.com> wrote:
>> For the output band, the standard method is to generate the signal
>> at an high IF well above the upper limit of the output band and
>> then downconverting. There is an advantage to the final mixer and
>> output filter to go above 3 times the high end of the output
>> band, e.g. 30.1 to 40 MHz for 0.1 to 10 MHz output.
>>
>
>For modulation, I have the required 90 degree phase shift covered often
>with accuracies better than tenths of a degree, over the band of 10Hz to
>100kHz. Supplying the required modulation will NOT be the problem. My
>problem is still with the carrier suppression and upper harmonic
>generation that comes out of a mixer, instead of out of a multiplier
>circuit. Perhaps what I really need is a 'perfect' analog multiplier that
>can operate up to 100MHz, similar architecture to the MC1496, but better
>specs??
>
>'Interfering higher tones' is a major concern, need them down more than
>90dBfs [would like more than 100dBfs]. It is surprising how difficult it
>is to make physically realizable filters that can knock those 'specced'
>tones down enough.
>
What exactly is your problem, the suppressed carrier attenuation, the
<robert...@gmail.com> wrote:
>> For the output band, the standard method is to generate the signal
>> at an high IF well above the upper limit of the output band and
>> then downconverting. There is an advantage to the final mixer and
>> output filter to go above 3 times the high end of the output
>> band, e.g. 30.1 to 40 MHz for 0.1 to 10 MHz output.
>>
>
>For modulation, I have the required 90 degree phase shift covered often
>with accuracies better than tenths of a degree, over the band of 10Hz to
>100kHz. Supplying the required modulation will NOT be the problem. My
>problem is still with the carrier suppression and upper harmonic
>generation that comes out of a mixer, instead of out of a multiplier
>circuit. Perhaps what I really need is a 'perfect' analog multiplier that
>can operate up to 100MHz, similar architecture to the MC1496, but better
>specs??
>
>'Interfering higher tones' is a major concern, need them down more than
>90dBfs [would like more than 100dBfs]. It is surprising how difficult it
>is to make physically realizable filters that can knock those 'specced'
>tones down enough.
>
opposite (LSB) attenuation or some multiples of the carrier frequency
?
With analog components, the opposite side band attenuation in I/Q
modulators can be at least 40 dB for the audio band. Doing something
in digital domain will have better performance, e.g. by generating the
SSB signal at a lowish frequency (2 MHz) and then translate to the
final frequency.
Are there any gaps in the frequency band that could be used to handle
the intermediate frequency (e.g. between 2 and 3 MHz) ?
What is the modulating signal, is it analog or already digital ? This
might affect how to best optimize the whole _system_.
Phil Hobbs
Apr 4, 2014, 2:38:34 PM4/4/14
to
1988. I found the reference in this patent:
http://tinyurl.com/l9ddqwa .
The whole thing is in the 1989 databook, so I scanned it,
<http://electrooptical.net/www/mixers/EdOxnerHighDynamicRangeFET_Mixers1985.pdf>
along with the datasheet:
<http://electrooptical.net/www/mixers/Si8901MonolithicFET_RingDatasheet1989.pdf>
Enjoy!
Phil Hobbs
Apr 4, 2014, 2:42:05 PM4/4/14
to
On 04/04/2014 01:10 PM, RobertMacy wrote:
shelf. The similar-vintage SD5400 series quad SPST switches were also
commonly used as mixers.
After discontinuing it, the morons reused the part number for something
completely different.
amdx
Apr 4, 2014, 2:43:46 PM4/4/14
to
As I recall this is an old article and there are updates
with newer parts.
I'll look for the updated articles.
Here's a background article on mixers, modulators and Demodulators.
The example here uses the SD5000 DMOD FET array.
Oh, just found this as an update for the array, FST3125.
https://www.fairchildsemi.com/ds/FS/FST3125.pdf
I think there are other subs, unknown to me at this moment.
I'm just throwing what I've read over the years at you,
I don't know how useful it is.
Mikek
Glenn
Apr 4, 2014, 4:19:39 PM4/4/14
to
On 04/04/14 20.43, amdx wrote:
...
Found this some time ago (FST3125 test:
http://martein.home.xs4all.nl/pa3ake/hmode/fst3125.html):
H-Mode mixer (Conclusion):
http://www.xs4all.nl/~martein/pa3ake/hmode/conclusion.html
Quote: "...
I almost gave up with the SPDT type mixers as they could not match the
FST3125 or FSAV332 SPST mixers with regard to IMD. But when finally the
FSA3157 was tested that did even better than the FSAV332 and FST3125, I
was rather surprised.
Both the FSA3157 and the NC7SZ384 mixers did not suffer at all from the
fact that the mixer is constructed from two to four separate devices
rather then just one, with regard to IMD.
...
The 74AC04 fundamental squarer turned out to be a winning factor when it
comes to getting rid of the spurs of a rather "spurry" unfiltered AD9951
LO-DDS.
...
The FSA3157 in combination with the fundamental 74AC04 squarer and the
ADTT1-1 transformer is very good when it comes to the reduction of spurs.
...
There is no need anymore to try to make the RF-port "see" a 50 ohms
resistive termination at all times to reduce the spur level. No more
half or full diplexers anymore in front of the mixer. This greatly
simplifies things! And consequently no extra (although small) losses
introduced by those diplexers worsening the NF of the front-end.
...
The category 6 mini-circuits transformers (ADTT1-6, ADT1-6T, TT1-6,
T1-6T) turned out to be well worth investigating too. Although no
winners in spur reduction, these transformers are the best if spurs are
not the first concern. What these transformers have in common is that
for an unexplained reason they almost completely eliminate the mixers
bias point sensitivity.
...
It turns out to be quite possible to construct a frontend with a dynamic
range in excess of 120dB in SSB bandwidth on 40M.
..."
NC7SB3157, FSA3157
Low-Voltage SPDT Analog Switch or 2:1Multiplexer / De-multiplexer Bus
Switch:
https://www.fairchildsemi.com/ds/NC/NC7SB3157.pdf
Quote: "...
* Break-Before-Make Enable Circuitry
* 250MHz, 3dB Bandwidth
..."
( main page: http://martein.home.xs4all.nl/pa3ake/hmode/ )
-
SM 5 BSZ - I/Q mixer for direct conversion radio (Sept 28 2001):
http://www.sm5bsz.com/linuxdsp/iqmixer.htm
Quote: "...
The simple HCMOS mixer in fig 1. produces a second order spur at -50dB
when the level is adjusted for a standard low noise op-amp to degrade
the dynamic range by less than 3dB.
...
Fig 1. A quadrature nixer [mixer] using 74HC4052. The control inputs are
fed from a synchronous counter that divides a local oscillator by four.
As a result the switch loops through all four positions during a period
of 4 local oscillator cycles, see fig. 2. For each switch position only
one connection is made between the antenna and either output. This way
the make before break operation of the switch will not cause
shortcircuits across the coil as would be the case using 74HC4053 as in
an older design. The dynamic range of the 74HC4052 design is superior
to that of the 74HC4053 design.
..."
-
Optimization of Quadrature Modulator Performance:
https://web.archive.org/web/20050130041540/http://www.rfmd.com/DataBooks/db97/an0001.pdf
-
Down-converter to zero-IF:
LT5517 - 40MHz to 900MHz Quadrature Demodulator:
http://www.linear.com/product/LT5517
New 40MHz - 900MHz Quadrature Demodulator Offers High Spurious-Free
Dynamic Range:
http://www.linear.com/docs/10457
Up-converter from zero-IF:
LT5572 - 1.5GHz to 2.5GHz High Linearity Direct Quadrature Modulator:
http://www.linear.com/product/LT5572
Down-converter to IF:
LT5560 - 0.01MHz to 4GHz Low Power Active Mixer:
http://www.linear.com/product/LT5560
/Glenn
...
> Oh, just found this as an update for the array, FST3125.
> https://www.fairchildsemi.com/ds/FS/FST3125.pdf
> I think there are other subs, unknown to me at this moment.
> I'm just throwing what I've read over the years at you,
> I don't know how useful it is.
> Mikek
>
>
Supplementary:
> https://www.fairchildsemi.com/ds/FS/FST3125.pdf
> I think there are other subs, unknown to me at this moment.
> I'm just throwing what I've read over the years at you,
> I don't know how useful it is.
> Mikek
>
>
Found this some time ago (FST3125 test:
http://martein.home.xs4all.nl/pa3ake/hmode/fst3125.html):
H-Mode mixer (Conclusion):
http://www.xs4all.nl/~martein/pa3ake/hmode/conclusion.html
Quote: "...
I almost gave up with the SPDT type mixers as they could not match the
FST3125 or FSAV332 SPST mixers with regard to IMD. But when finally the
FSA3157 was tested that did even better than the FSAV332 and FST3125, I
was rather surprised.
Both the FSA3157 and the NC7SZ384 mixers did not suffer at all from the
fact that the mixer is constructed from two to four separate devices
rather then just one, with regard to IMD.
...
The 74AC04 fundamental squarer turned out to be a winning factor when it
comes to getting rid of the spurs of a rather "spurry" unfiltered AD9951
LO-DDS.
...
The FSA3157 in combination with the fundamental 74AC04 squarer and the
ADTT1-1 transformer is very good when it comes to the reduction of spurs.
...
There is no need anymore to try to make the RF-port "see" a 50 ohms
resistive termination at all times to reduce the spur level. No more
half or full diplexers anymore in front of the mixer. This greatly
simplifies things! And consequently no extra (although small) losses
introduced by those diplexers worsening the NF of the front-end.
...
The category 6 mini-circuits transformers (ADTT1-6, ADT1-6T, TT1-6,
T1-6T) turned out to be well worth investigating too. Although no
winners in spur reduction, these transformers are the best if spurs are
not the first concern. What these transformers have in common is that
for an unexplained reason they almost completely eliminate the mixers
bias point sensitivity.
...
It turns out to be quite possible to construct a frontend with a dynamic
range in excess of 120dB in SSB bandwidth on 40M.
..."
NC7SB3157, FSA3157
Low-Voltage SPDT Analog Switch or 2:1Multiplexer / De-multiplexer Bus
Switch:
https://www.fairchildsemi.com/ds/NC/NC7SB3157.pdf
Quote: "...
* Break-Before-Make Enable Circuitry
* 250MHz, 3dB Bandwidth
..."
( main page: http://martein.home.xs4all.nl/pa3ake/hmode/ )
-
SM 5 BSZ - I/Q mixer for direct conversion radio (Sept 28 2001):
http://www.sm5bsz.com/linuxdsp/iqmixer.htm
Quote: "...
The simple HCMOS mixer in fig 1. produces a second order spur at -50dB
when the level is adjusted for a standard low noise op-amp to degrade
the dynamic range by less than 3dB.
...
Fig 1. A quadrature nixer [mixer] using 74HC4052. The control inputs are
fed from a synchronous counter that divides a local oscillator by four.
As a result the switch loops through all four positions during a period
of 4 local oscillator cycles, see fig. 2. For each switch position only
one connection is made between the antenna and either output. This way
the make before break operation of the switch will not cause
shortcircuits across the coil as would be the case using 74HC4053 as in
an older design. The dynamic range of the 74HC4052 design is superior
to that of the 74HC4053 design.
..."
-
Optimization of Quadrature Modulator Performance:
https://web.archive.org/web/20050130041540/http://www.rfmd.com/DataBooks/db97/an0001.pdf
-
Down-converter to zero-IF:
LT5517 - 40MHz to 900MHz Quadrature Demodulator:
http://www.linear.com/product/LT5517
New 40MHz - 900MHz Quadrature Demodulator Offers High Spurious-Free
Dynamic Range:
http://www.linear.com/docs/10457
Up-converter from zero-IF:
LT5572 - 1.5GHz to 2.5GHz High Linearity Direct Quadrature Modulator:
http://www.linear.com/product/LT5572
Down-converter to IF:
LT5560 - 0.01MHz to 4GHz Low Power Active Mixer:
http://www.linear.com/product/LT5560
/Glenn
Glenn
Apr 4, 2014, 4:30:11 PM4/4/14
to
More:
H-Mode mixer - Introduction:
http://martein.home.xs4all.nl/pa3ake/hmode/hmode_intro.html
Quote: "...
Next came Colin's tip to use the 74AC74E (DIL14) or 74AC74M (SOIC14)
from Texas Instruments in the squarer as those parts have balanced
propagation delays in contrast to the ordinary 74AC74. This again
reduced the average spur levels.
..."
/Glenn
H-Mode mixer - Introduction:
http://martein.home.xs4all.nl/pa3ake/hmode/hmode_intro.html
Quote: "...
Next came Colin's tip to use the 74AC74E (DIL14) or 74AC74M (SOIC14)
from Texas Instruments in the squarer as those parts have balanced
propagation delays in contrast to the ordinary 74AC74. This again
reduced the average spur levels.
..."
/Glenn
Clifford Heath
Apr 4, 2014, 5:29:39 PM4/4/14
to
On 04/04/14 17:23, Tauno Voipio wrote:
> On 3.4.14 21:21, Tim Wescott wrote:
>> For "very linear" the diode-ring mixers (and, if you really want to work
>> at it, FET-ring mixers) have been recognized to be superior to Gilbert
>> cell mixers for years and years. I doubt that has changed, although
>> knowing whether a diode ring is superior to a FET ring is beyond my pay
>> scale at the moment.
>>
>> Gilbert cell mixers are better if you need to build an inexpensive small
>> circuit that doesn't consume much power, though.
>
> The last advance in very linear mixers seems to be using CMOS analog
> switches to do the switching instead of transistors or diodes.
>
> Have a look at <http://www.norcalqrp.org/files/tayloe_mixer_x3a.pdf>.
One of the Australian electronic magazines (EA, ETI, not sure) published
> On 3.4.14 21:21, Tim Wescott wrote:
>> For "very linear" the diode-ring mixers (and, if you really want to work
>> at it, FET-ring mixers) have been recognized to be superior to Gilbert
>> cell mixers for years and years. I doubt that has changed, although
>> knowing whether a diode ring is superior to a FET ring is beyond my pay
>> scale at the moment.
>>
>> Gilbert cell mixers are better if you need to build an inexpensive small
>> circuit that doesn't consume much power, though.
>
> The last advance in very linear mixers seems to be using CMOS analog
> switches to do the switching instead of transistors or diodes.
>
> Have a look at <http://www.norcalqrp.org/files/tayloe_mixer_x3a.pdf>.
a synchrodyne AM receiver in the 1970's using a quad CMOS switch, driven
by I/Q outputs from a counter with a 4x VCO. Not very new...
amdx
Apr 4, 2014, 6:36:51 PM4/4/14
to
Here's a link to rec.radio.amateur.homebrew is on this subject, a read
thru that will give you plenty to think about.
https://groups.google.com/forum/#!topic/rec.radio.amateur.homebrew/qWet-1h1dgg
Mikek
Clifford Heath
Apr 4, 2014, 6:52:55 PM4/4/14
to
On 05/04/14 01:45, Phil Hobbs wrote:
> The main issue is that the RF port drives a BJT diff pair that's in its
> linear range, so that its output current goes as
> tanh(e dV_BE/(2kT)),
[snip]
> The main issue is that the RF port drives a BJT diff pair that's in its
> linear range, so that its output current goes as
> tanh(e dV_BE/(2kT)),
> A strong interfering signal will drive the RF diff pair of the Gilbert
> cell nonlinear just the way it does an ordinary BJT amplifier. Running
> the diff pair as a current mirror, as in the LM13700 OTA, nominally
> turns the tanh characteristic back into a linear one, but that trick
> isn't often used at RF, I don't think.
I can't visualise how that would be done in the diff pair of a Gilbert
> cell nonlinear just the way it does an ordinary BJT amplifier. Running
> the diff pair as a current mirror, as in the LM13700 OTA, nominally
> turns the tanh characteristic back into a linear one, but that trick
> isn't often used at RF, I don't think.
cell. If these transistors are run at constant current, how do they
steer current into the upper stages?
Could you sketch it?
RobertMacy
Apr 4, 2014, 7:58:42 PM4/4/14
to
On Fri, 04 Apr 2014 07:27:48 -0700, Phil Hobbs
<pcdhSpamM...@electrooptical.net> wrote:
>> ...snip...
taken me over 2 months now [working part time on it is my only defense] As
far as I know, I'm the only person to do this, ...so far.
It does have a limitation that you have to stay in 'reasonable' operation.
I could zip up a sample of the MC1496 modeled fairly well for an
application I'm looking at. But the zipped file is somewhere between 30MB
and 65MB, so NO way I can post that sort of wad of bits.
I work a lot in 0.001 Hz up through 1MHz, right now. so it has really
helped to be able to 'add' in that pesky 1/f noise and watch it get
replicated elsewhere WITHOUT the need to calculate what's going on [well,
until later for verification].
<pcdhSpamM...@electrooptical.net> wrote:
>> ...snip...
> Boom-kssssshhhh.... I meant, "how does it work", not "how well does it
> work."
>
> Doing that in any generality for strongly nonlinear circuits is an
> intrinsically difficult problem, or so I would have thought.
>
> Cheers
>
> Phil Hobbs
>
couldn't resist that one! Yes, it is difficult. I'm embarrassed to say has
> work."
>
> Doing that in any generality for strongly nonlinear circuits is an
> intrinsically difficult problem, or so I would have thought.
>
> Cheers
>
> Phil Hobbs
>
taken me over 2 months now [working part time on it is my only defense] As
far as I know, I'm the only person to do this, ...so far.
It does have a limitation that you have to stay in 'reasonable' operation.
I could zip up a sample of the MC1496 modeled fairly well for an
application I'm looking at. But the zipped file is somewhere between 30MB
and 65MB, so NO way I can post that sort of wad of bits.
I work a lot in 0.001 Hz up through 1MHz, right now. so it has really
helped to be able to 'add' in that pesky 1/f noise and watch it get
replicated elsewhere WITHOUT the need to calculate what's going on [well,
until later for verification].
RobertMacy
Apr 4, 2014, 8:18:25 PM4/4/14
to
On Fri, 04 Apr 2014 15:52:55 -0700, Clifford Heath <no....@please.net>
wrote:
>> ...snip...
around 485u each so you can only 'tilt' approx 1/2 that before really
start to see the distortion.
wrote:
>> ...snip...
> I can't visualise how that would be done in the diff pair of a Gilbert
> cell. If these transistors are run at constant current, how do they
> steer current into the upper stages?
>
> Could you sketch it?
>
Data sheet of the MC1496 shows it very well. The sample runs at bias of
> cell. If these transistors are run at constant current, how do they
> steer current into the upper stages?
>
> Could you sketch it?
>
around 485u each so you can only 'tilt' approx 1/2 that before really
start to see the distortion.
Phil Hobbs
Apr 4, 2014, 8:48:44 PM4/4/14
to
Phil Hobbs
Apr 4, 2014, 9:26:16 PM4/4/14
to
or the following listing.
Cheers
Phil Hobbs
Version 4
SHEET 1 1044 680
WIRE 112 -384 -96 -384
WIRE 368 -384 112 -384
WIRE 448 -384 368 -384
WIRE 112 -368 112 -384
WIRE -96 -352 -96 -384
WIRE 368 -336 368 -384
WIRE -96 -240 -96 -272
WIRE 112 -240 112 -288
WIRE 224 -240 112 -240
WIRE 368 -240 368 -256
WIRE 368 -240 288 -240
WIRE 112 -192 112 -240
WIRE 448 -192 112 -192
WIRE 368 -160 368 -240
WIRE 368 -160 240 -160
WIRE 112 -96 112 -192
WIRE 240 -96 240 -160
WIRE 368 -96 368 -160
WIRE 448 -96 448 -192
WIRE 48 -48 -160 -48
WIRE 176 -48 48 -48
WIRE 512 -48 432 -48
WIRE 640 -48 512 -48
WIRE 240 32 240 0
WIRE 448 32 448 0
WIRE 448 32 240 32
WIRE -160 48 -160 32
WIRE 640 48 640 32
WIRE 112 64 112 0
WIRE 368 64 368 0
WIRE 368 64 112 64
WIRE 576 96 -144 96
WIRE 720 96 576 96
WIRE -144 128 -144 96
WIRE 64 144 -32 144
WIRE 144 144 64 144
WIRE 576 144 576 96
WIRE 576 144 496 144
WIRE 720 176 720 96
WIRE 64 192 64 144
WIRE 240 192 240 32
WIRE 368 192 368 64
WIRE 576 192 576 144
WIRE 144 240 144 144
WIRE 144 240 128 240
WIRE 176 240 144 240
WIRE 496 240 496 144
WIRE 496 240 432 240
WIRE 512 240 496 240
WIRE -144 272 -144 208
WIRE -32 272 -32 144
WIRE -32 272 -144 272
WIRE -32 288 -32 272
WIRE 576 320 576 288
WIRE 720 320 720 256
WIRE 720 320 576 320
WIRE 64 336 64 288
WIRE 576 336 576 320
WIRE 240 368 240 288
WIRE 304 368 240 368
WIRE 368 368 368 288
WIRE 368 368 304 368
WIRE -32 432 -32 368
WIRE 304 432 304 368
WIRE 304 560 304 512
FLAG 64 336 0
FLAG 576 336 0
FLAG 304 560 0
FLAG -32 432 0
FLAG 640 48 0
FLAG -96 -240 0
FLAG -160 48 0
SYMBOL npn 176 192 R0
SYMATTR InstName Q1
SYMATTR Value 2N2369
SYMBOL npn 432 192 M0
WINDOW 3 -49 126 Left 2
SYMATTR InstName Q2
SYMATTR Value 2N2369
SYMBOL npn 128 192 M0
SYMATTR InstName Q3
SYMATTR Value 2N2369
SYMBOL npn 512 192 R0
WINDOW 3 78 98 Left 2
SYMATTR InstName Q4
SYMATTR Value 2N2369
SYMBOL current 304 432 R0
SYMATTR InstName I1
SYMATTR Value 3m
SYMBOL npn 176 -96 R0
WINDOW 3 72 109 Left 2
SYMATTR InstName Q5
SYMATTR Value 2N2369
SYMBOL npn 432 -96 M0
SYMATTR InstName Q6
SYMATTR Value 2N2369
SYMBOL npn 48 -96 R0
SYMATTR InstName Q7
SYMATTR Value 2N2369
SYMBOL npn 512 -96 M0
WINDOW 3 -54 125 Left 2
SYMATTR InstName Q8
SYMATTR Value 2N2369
SYMBOL res 96 -384 R0
SYMATTR InstName R1
SYMATTR Value 400
SYMBOL res 352 -352 R0
SYMATTR InstName R2
SYMATTR Value 400
SYMBOL current -144 128 R0
WINDOW 3 55 -7 Left 2
WINDOW 123 24 108 Left 2
WINDOW 39 0 0 Left 2
SYMATTR InstName I2
SYMATTR Value SINE(0 5u 1meg)
SYMATTR Value2 AC 1
SYMBOL current 720 256 R180
WINDOW 0 24 80 Left 2
WINDOW 3 24 0 Left 2
WINDOW 123 0 0 Left 2
WINDOW 39 0 0 Left 2
SYMATTR InstName I3
SYMATTR Value 100µ
SYMBOL current -32 368 R180
WINDOW 0 24 80 Left 2
WINDOW 3 24 0 Left 2
WINDOW 123 0 0 Left 2
WINDOW 39 0 0 Left 2
SYMATTR InstName I4
SYMATTR Value 100µ
SYMBOL voltage 640 -64 R0
WINDOW 123 0 0 Left 2
WINDOW 39 0 0 Left 2
SYMATTR InstName V1
SYMATTR Value SINE(2.5 500m 1.05meg 0 0 0)
SYMBOL voltage -96 -368 R0
SYMATTR InstName V3
SYMATTR Value 5
SYMBOL voltage -160 -64 R0
WINDOW 123 0 0 Left 2
WINDOW 39 0 0 Left 2
SYMATTR InstName V4
SYMATTR Value SINE(2.5 500m 1.05meg 0 0 180)
SYMBOL cap 288 -256 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 32 32 VTop 2
SYMATTR InstName C1
SYMATTR Value 10n
TEXT 560 568 Left 2 !.tran 100u
TEXT -496 160 Left 2 ;Signal input (differential)
TEXT 144 -296 Left 2 ;Output(differential)
josephkk
Apr 4, 2014, 11:52:21 PM4/4/14
to
On Fri, 04 Apr 2014 09:19:00 +0300, Tauno Voipio
<tauno....@notused.fi.invalid> wrote:
>On 4.4.14 03:17, Clifford Heath wrote:
>> On 03/04/14 17:16, Tauno Voipio wrote:
>>> On 3.4.14 03:00, RobertMacy wrote:
>>>> Is the bandwidth too much for the baseband, or for the carrier?
>>>
>>> Both, but for different reasons:
>>>
>>> - Creating the analytic signal (two copies at 90 degrees difference)
>>> makes the requirement of limited baseband. It must be limited at
>>> the low end to make the phase shifting possible, 0 Hz does not go.
>>> The minimum delay for the filters is a quarter wave at he lowest
>>> frequency.
>>
>> You can get pretty accurate 90 degree phase shifts across a decade of
>> frequency using a fourth-order quadrature hybrid, but I really wonder if
>> it wouldn't be easier to generate 5.1-15MHz SSB and downconvert it to
>> 0.1-10MHz.
analysis purposes substitute as follows:
For Cos(nearly to real DC) the DC voltage;
For Sin(nearly to real DC) 0 volts.
I got this from theory of limits. I presume that you have the ability to
fill it in on your own from this.
>Even with a low limit
>of 10 Hz, the required bandwidth exceeds by a wide margin those
>which have been realized. I have the feeling that it is in practice
>impossible to get analog components that are matched well enough for
>the entire baseband width. Digitally, it may be possible if
>the band is limited at the low end, but the filters will be very
>long to provide the necessary delay difference at the low end
>of the range with the sampling rate needed for the high end.
>
>There are other methods with e.g. matched FIR filters, polyphase RC
>nets and mixing in the Weaver method.
>
<tauno....@notused.fi.invalid> wrote:
>On 4.4.14 03:17, Clifford Heath wrote:
>> On 03/04/14 17:16, Tauno Voipio wrote:
>>> On 3.4.14 03:00, RobertMacy wrote:
>>>> Is the bandwidth too much for the baseband, or for the carrier?
>>>
>>> Both, but for different reasons:
>>>
>>> - Creating the analytic signal (two copies at 90 degrees difference)
>>> makes the requirement of limited baseband. It must be limited at
>>> the low end to make the phase shifting possible, 0 Hz does not go.
>>> The minimum delay for the filters is a quarter wave at he lowest
>>> frequency.
>>
>> You can get pretty accurate 90 degree phase shifts across a decade of
>> frequency using a fourth-order quadrature hybrid, but I really wonder if
>> it wouldn't be easier to generate 5.1-15MHz SSB and downconvert it to
>> 0.1-10MHz.
>
>
>We still have the problem of low frequencies. The OP wanted a band
>starting from DC, which is IMHO impossible.
I will dispute that with you analytically, perhaps even practically. For
>
>We still have the problem of low frequencies. The OP wanted a band
>starting from DC, which is IMHO impossible.
analysis purposes substitute as follows:
For Cos(nearly to real DC) the DC voltage;
For Sin(nearly to real DC) 0 volts.
I got this from theory of limits. I presume that you have the ability to
fill it in on your own from this.
>Even with a low limit
>of 10 Hz, the required bandwidth exceeds by a wide margin those
>which have been realized. I have the feeling that it is in practice
>impossible to get analog components that are matched well enough for
>the entire baseband width. Digitally, it may be possible if
>the band is limited at the low end, but the filters will be very
>long to provide the necessary delay difference at the low end
>of the range with the sampling rate needed for the high end.
>
>There are other methods with e.g. matched FIR filters, polyphase RC
>nets and mixing in the Weaver method.
>
Clifford Heath
Apr 5, 2014, 2:43:13 AM4/5/14
to
On 05/04/14 12:26, Phil Hobbs wrote:
> On 4/4/2014 8:18 PM, RobertMacy wrote:
>> On Fri, 04 Apr 2014 15:52:55 -0700, Clifford Heath <no....@please.net>
>> wrote:
>>
>>>> ...snip...
>>> I can't visualise how that would be done in the diff pair of a Gilbert
>>> cell. If these transistors are run at constant current, how do they
>>> steer current into the upper stages?
>>>
>>> Could you sketch it?
>>>
>>
>> Data sheet of the MC1496 shows it very well. The sample runs at bias of
>> around 485u each so you can only 'tilt' approx 1/2 that before really
>> start to see the distortion.
>
> The MC1496 doesn't use the diode trick. Check out the LM13700 datasheet
> or the following listing.
Impressive, thanks!
> On 4/4/2014 8:18 PM, RobertMacy wrote:
>> On Fri, 04 Apr 2014 15:52:55 -0700, Clifford Heath <no....@please.net>
>> wrote:
>>
>>>> ...snip...
>>> I can't visualise how that would be done in the diff pair of a Gilbert
>>> cell. If these transistors are run at constant current, how do they
>>> steer current into the upper stages?
>>>
>>> Could you sketch it?
>>>
>>
>> Data sheet of the MC1496 shows it very well. The sample runs at bias of
>> around 485u each so you can only 'tilt' approx 1/2 that before really
>> start to see the distortion.
>
> The MC1496 doesn't use the diode trick. Check out the LM13700 datasheet
> or the following listing.
--
Clifford Heath.
upsid...@downunder.com
Apr 5, 2014, 3:48:03 AM4/5/14
to
On Fri, 04 Apr 2014 16:58:42 -0700, RobertMacy
<robert...@gmail.com> wrote:
>I work a lot in 0.001 Hz up through 1MHz, right now. so it has really
>helped to be able to 'add' in that pesky 1/f noise and watch it get
>replicated elsewhere WITHOUT the need to calculate what's going on [well,
>until later for verification].
If the base band requirement is 0.001 Hz (and 1/f issues) I do not
<robert...@gmail.com> wrote:
>I work a lot in 0.001 Hz up through 1MHz, right now. so it has really
>helped to be able to 'add' in that pesky 1/f noise and watch it get
>replicated elsewhere WITHOUT the need to calculate what's going on [well,
>until later for verification].
understand how you are going to transfer this on radio waves on SSB.
At least a pilot tone (carrier) is absolutely required. Any Tx/Rx
frequency error will destroy the low frequency accuracy.
I would suggest going digital with the DC to 1 MHz signal and then
think about how to transfer it over radio waves (there are a huge
number of alternatives).
josephkk
Apr 5, 2014, 6:26:35 AM4/5/14
to
On Fri, 04 Apr 2014 10:45:21 -0400, Phil Hobbs
<pcdhSpamM...@electrooptical.net> wrote:
>On 04/04/2014 09:04 AM, George Herold wrote:
>> On Thursday, April 3, 2014 2:21:45 PM UTC-4, Tim Wescott wrote:
>>> On Wed, 02 Apr 2014 17:26:21 -0700, RobertMacy wrote:
>>>
>> <snip>
>> Gilbert cell would be better, 'cause it doesn't have all the odd
>> harmonics that the switchers have. In practice I can imagine that
>> switching is much easier than multiplying.
>>
>> George H.
>
>I don't think it's the current steering (upper) stage of the Gilbert
>cell that's the problem--those devices are almost always ON or OFF.
Not at all. They are operating in the linear portion of the tanh(x)
curve. They could not do the job at all if they were bouncing in and out
of cutoff and saturation. They have to stay in the active region or
create all kind of nasty non-linearities and harmonics.
BTW for all the people recommending saturated switching mixing, this is
always followed be methods to suppress the harmonics generated. Which
would likely reduce the accuracy of the measurements compared to not
generating them in the first place.
>input resistance of the first IF, plus the switch resistances. Diode
>mixers work a lot better into a constant 50 ohms for this reason.
>
>(Using a diplexer instead of a reflective filter after the mixer is a
>big win for linearity, and this is why.)
balanced from DC to past 100 MHz. The Gilbert cell is all about the
tanh(x) properties of differential pairs. That is the basis of how it
works.
>
>Cheers
>
>Phil Hobbs
<pcdhSpamM...@electrooptical.net> wrote:
>On 04/04/2014 09:04 AM, George Herold wrote:
>> On Thursday, April 3, 2014 2:21:45 PM UTC-4, Tim Wescott wrote:
>>> On Wed, 02 Apr 2014 17:26:21 -0700, RobertMacy wrote:
>>>
>> <snip>
>>>
>>> For "very linear" the diode-ring mixers (and, if you really want to
>>> work at it, FET-ring mixers) have been recognized to be superior to
>>> Gilbert cell mixers for years and years. I doubt that has changed,
>>> although knowing whether a diode ring is superior to a FET ring is
>>> beyond my pay scale at the moment.
>
>> Tim, do you have any idea why that is. "In theory" I would think the
>>> For "very linear" the diode-ring mixers (and, if you really want to
>>> work at it, FET-ring mixers) have been recognized to be superior to
>>> Gilbert cell mixers for years and years. I doubt that has changed,
>>> although knowing whether a diode ring is superior to a FET ring is
>>> beyond my pay scale at the moment.
>
>> Gilbert cell would be better, 'cause it doesn't have all the odd
>> harmonics that the switchers have. In practice I can imagine that
>> switching is much easier than multiplying.
>>
>> George H.
>
>I don't think it's the current steering (upper) stage of the Gilbert
>cell that's the problem--those devices are almost always ON or OFF.
Not at all. They are operating in the linear portion of the tanh(x)
curve. They could not do the job at all if they were bouncing in and out
of cutoff and saturation. They have to stay in the active region or
create all kind of nasty non-linearities and harmonics.
BTW for all the people recommending saturated switching mixing, this is
always followed be methods to suppress the harmonics generated. Which
would likely reduce the accuracy of the measurements compared to not
generating them in the first place.
>
>The main issue is that the RF port drives a BJT diff pair that's in its
>linear range, so that its output current goes as
>tanh(e dV_BE/(2kT)), whereas in a diode mixer the RF port's load is the
>The main issue is that the RF port drives a BJT diff pair that's in its
>linear range, so that its output current goes as
>input resistance of the first IF, plus the switch resistances. Diode
>mixers work a lot better into a constant 50 ohms for this reason.
>
>(Using a diplexer instead of a reflective filter after the mixer is a
>big win for linearity, and this is why.)
>
>A strong interfering signal will drive the RF diff pair of the Gilbert
>cell nonlinear just the way it does an ordinary BJT amplifier. Running
>the diff pair as a current mirror, as in the LM13700 OTA, nominally
>turns the tanh characteristic back into a linear one, but that trick
>isn't often used at RF, I don't think.
Really? There are plenty of Gilbert cell multipliers that stay well
>A strong interfering signal will drive the RF diff pair of the Gilbert
>cell nonlinear just the way it does an ordinary BJT amplifier. Running
>the diff pair as a current mirror, as in the LM13700 OTA, nominally
>turns the tanh characteristic back into a linear one, but that trick
>isn't often used at RF, I don't think.
balanced from DC to past 100 MHz. The Gilbert cell is all about the
tanh(x) properties of differential pairs. That is the basis of how it
works.
>
>Cheers
>
>Phil Hobbs
Phil Hobbs
Apr 5, 2014, 7:10:34 AM4/5/14
to
On 4/5/2014 6:26 AM, josephkk wrote:
> On Fri, 04 Apr 2014 10:45:21 -0400, Phil Hobbs
> <pcdhSpamM...@electrooptical.net> wrote:
>
>> On 04/04/2014 09:04 AM, George Herold wrote:
>>> On Thursday, April 3, 2014 2:21:45 PM UTC-4, Tim Wescott wrote:
>>>> On Wed, 02 Apr 2014 17:26:21 -0700, RobertMacy wrote:
>>>>
>>> <snip>
>>>>
>>>> For "very linear" the diode-ring mixers (and, if you really want to
>>>> work at it, FET-ring mixers) have been recognized to be superior to
>>>> Gilbert cell mixers for years and years. I doubt that has changed,
>>>> although knowing whether a diode ring is superior to a FET ring is
>>>> beyond my pay scale at the moment.
>>
>>> Tim, do you have any idea why that is. "In theory" I would think the
>>> Gilbert cell would be better, 'cause it doesn't have all the odd
>>> harmonics that the switchers have. In practice I can imagine that
>>> switching is much easier than multiplying.
>>>
>>> George H.
>>
>> I don't think it's the current steering (upper) stage of the Gilbert
>> cell that's the problem--those devices are almost always ON or OFF.
>
> Not at all. They are operating in the linear portion of the tanh(x)
> curve. They could not do the job at all if they were bouncing in and out
> of cutoff and saturation. They have to stay in the active region or
> create all kind of nasty non-linearities and harmonics.
A barefoot Gilbert cell is a sorta-kinda multiplier. To make a good
> On Fri, 04 Apr 2014 10:45:21 -0400, Phil Hobbs
> <pcdhSpamM...@electrooptical.net> wrote:
>
>> On 04/04/2014 09:04 AM, George Herold wrote:
>>> On Thursday, April 3, 2014 2:21:45 PM UTC-4, Tim Wescott wrote:
>>>> On Wed, 02 Apr 2014 17:26:21 -0700, RobertMacy wrote:
>>>>
>>> <snip>
>>>>
>>>> For "very linear" the diode-ring mixers (and, if you really want to
>>>> work at it, FET-ring mixers) have been recognized to be superior to
>>>> Gilbert cell mixers for years and years. I doubt that has changed,
>>>> although knowing whether a diode ring is superior to a FET ring is
>>>> beyond my pay scale at the moment.
>>
>>> Tim, do you have any idea why that is. "In theory" I would think the
>>> Gilbert cell would be better, 'cause it doesn't have all the odd
>>> harmonics that the switchers have. In practice I can imagine that
>>> switching is much easier than multiplying.
>>>
>>> George H.
>>
>> I don't think it's the current steering (upper) stage of the Gilbert
>> cell that's the problem--those devices are almost always ON or OFF.
>
> Not at all. They are operating in the linear portion of the tanh(x)
> curve. They could not do the job at all if they were bouncing in and out
> of cutoff and saturation. They have to stay in the active region or
> create all kind of nasty non-linearities and harmonics.
linear multiplier, you have to work quite a bit harder than that.
Gilbert cell frequency mixers are generally run with the lower stage
linear and the upper stage switching. The upper stage works very much
like a MOSFET- or diode-ring mixer. Switching mixers do give you
intermod products of the form M*f_RF + N*f_LO, but you have to deal with
that anyway, because even a 1%-accuracy multiplier only gives you ~40 dB
spurious suppression, which generally is far too little. You deal with
it by picking a good frequency plan and building good filters.
The crucial advantage of switching mixers is that they don't give you a
lot of intermodulation between different signals on the RF port, i.e.
they suppress spurs of the form M*f_RF1 + N*f_RF2. Lots of those wind
up in-band, so filtering doesn't help nearly as much as a nice strong mixer.
And nobody in his right mind lets any of the transistors saturate.
Saturation is slow and noisy, whereas cutoff is well-behaved. If you're
running the Gilbert cell at, say, 3 mA total, then when the lower stage
is in balance, each of the upper stage transistors switch back and forth
from 1.5 mA to cutoff.
>
> BTW for all the people recommending saturated switching mixing, this is
> always followed be methods to suppress the harmonics generated. Which
> would likely reduce the accuracy of the measurements compared to not
> generating them in the first place.
mixer in a receiver has a difficult life. Once you're past the first IF
filter, life gets a lot calmer. In an instrument application, where you
control what's in the input, you can sometimes design around the
problem. But linear multipliers aren't that great, especially not quick
ones. It's a lot easier to get 0.1% accuracy using a nice fast
switching mixer than a Gilbert cell.
>>
>> The main issue is that the RF port drives a BJT diff pair that's in its
>> linear range, so that its output current goes as
>> tanh(e dV_BE/(2kT)), whereas in a diode mixer the RF port's load is the
>> input resistance of the first IF, plus the switch resistances. Diode
>> mixers work a lot better into a constant 50 ohms for this reason.
>>
>> (Using a diplexer instead of a reflective filter after the mixer is a
>> big win for linearity, and this is why.)
>>
>> A strong interfering signal will drive the RF diff pair of the Gilbert
>> cell nonlinear just the way it does an ordinary BJT amplifier. Running
>> the diff pair as a current mirror, as in the LM13700 OTA, nominally
>> turns the tanh characteristic back into a linear one, but that trick
>> isn't often used at RF, I don't think.
>
>
> Really? There are plenty of Gilbert cell multipliers that stay well
> balanced from DC to past 100 MHz. The Gilbert cell is all about the
> tanh(x) properties of differential pairs. That is the basis of how it
> works.
interfering signals present. The Maclaurin series for tanh is
tanh(x) = x - (1/3) x**3 + (2/15) x**5 ....
For BJT pairs, delta I_C = I_tail * tanh(delta V_BE*e/(2kT)), so the
scale factor is about 50 mV.
The third order term reaches 1% of the desired signal when the input
amplitude gets up to
50 mV * sqrt(3/100) ~ 9 mV.
A properly designed linear multiplier does quite a bit better than this.
The diode trick that I posted is one of the simplest ways of patching
up the tanh characteristic, but there are lots of improvements you could
add, such as cascoding the lower stage. In a vanilla Gilbert cell with
sinusoidal LO, the lower stage collectors bounce up and down at twice
the LO frequency, which is another source of nonlinearity.
For receivers, there are other things that might be useful:
It would be sort of interesting to try replacing the lower stage of the
Gilbert cell with a center-tapped transformer secondary.
In a MOSFET bridge, it might be better to give each of the FETs its own
LO winding, connected G-S as in HV SMPSes. Driving the gates in pairs
gets rid of the even order distortion terms but not the odd order ones,
whereas separate gate drive should pretty well get rid of both.
Cheers
Phil Hobbs
Kevin Aylward
Apr 5, 2014, 9:40:05 AM4/5/14
to
"Phil Hobbs" wrote in message news:533FE4AA...@electrooptical.net...
On 4/5/2014 6:26 AM, josephkk wrote:
> On Fri, 04 Apr 2014 10:45:21 -0400, Phil Hobbs
> <pcdhSpamM...@electrooptical.net> wrote:
>
>> The Gilbert cell is all about the
>> tanh(x) properties of differential pairs. That is the basis of how it
> >works.
Ahmm... lets be accurate shall we... :-)
The basic multiplier is the "Howard Jones cell" invented by said person in
1963. It consists of the x-coupled stacked differential pairs erroneously
named the Gilbert cell.
The true Gilbert Cell is the the Howard Jones cell with logging diode
(transistors) on the input. It is the logging bit (aforementioned trick)
that Barrie introduced to get, in the ideal limit, perfect multiplication.
So, in reality, the Gilbert Cell is actually very linear, assuming a decent
V-I converter feeding the diodes. It is the Howard cell that is nonlinear.
Kevin Aylward B.Sc.
www.kevinaylward.co.uk
On 4/5/2014 6:26 AM, josephkk wrote:
> On Fri, 04 Apr 2014 10:45:21 -0400, Phil Hobbs
> <pcdhSpamM...@electrooptical.net> wrote:
>
>> The Gilbert cell is all about the
>> tanh(x) properties of differential pairs. That is the basis of how it
> >works.
> The diode trick that I posted is one of the simplest ways of patching up
> the tanh characteristic, In a vanilla Gilbert cell with
Ahmm... lets be accurate shall we... :-)
The basic multiplier is the "Howard Jones cell" invented by said person in
1963. It consists of the x-coupled stacked differential pairs erroneously
named the Gilbert cell.
The true Gilbert Cell is the the Howard Jones cell with logging diode
(transistors) on the input. It is the logging bit (aforementioned trick)
that Barrie introduced to get, in the ideal limit, perfect multiplication.
So, in reality, the Gilbert Cell is actually very linear, assuming a decent
V-I converter feeding the diodes. It is the Howard cell that is nonlinear.
Kevin Aylward B.Sc.
www.kevinaylward.co.uk
josephkk
Apr 5, 2014, 2:11:45 PM4/5/14
to
>On Thu, 03 Apr 2014 23:40:51 -0700, josephkk
><joseph_...@sbcglobal.net> wrote:
>
>>> ...snip...
>> Well most double balanced mixers are very similar to Gilbert cell
>> multipliers. The very wide bandwidth makes phasing type twin mixers
>> impossible for conversion to SSB. The remaining choice it to mix to a
>> much higher frequency (1 GHz) and use a SAW filter or similar to split
>> the
>> sidebands and mix down again with heavy filtering to do the image
>> rejection.
>>
>> You need some more slack in the bandwidths to make this reasonable.
>>
>> ?-)
>>
>
>
>Forgot all about SAW filtering. I know it's fixed filtering, but might
>have some advantages.
>
>
>The required 90 degree phase shift in the modulation is covered. And, it's
>pretty easy to supply LO with quadrature, too. It's the 'balance' in the
>mixers I need to get that pesky carrier tone and 3rd harmonics down.
>
>100MHz has 10ns period. 0.1nS is almost 4 degrees phase shift! Somehow
>will have to make the system accomodate that kind of sloppy action, too
LO never was a worry, the 9 decades band on the input was the bother.
Gettting 90 degrees over more than three decades (from a single source) is
all but impossible.
That said there are multipliers that are good from DC to 250 MHz or more.
I do not always like square wave drive on diode rings because of the
harmonics which are usually just filtered out but that doesn't seem like
it would work in your application as stated so far.
If you can keep f(m) << f(c) the designs get a lot easier. If you can do
band switching for <10 Hz and > 100 kHz it could really help.
?-)
><joseph_...@sbcglobal.net> wrote:
>
>>> ...snip...
>> Well most double balanced mixers are very similar to Gilbert cell
>> multipliers. The very wide bandwidth makes phasing type twin mixers
>> impossible for conversion to SSB. The remaining choice it to mix to a
>> much higher frequency (1 GHz) and use a SAW filter or similar to split
>> the
>> sidebands and mix down again with heavy filtering to do the image
>> rejection.
>>
>> You need some more slack in the bandwidths to make this reasonable.
>>
>> ?-)
>>
>
>
>Forgot all about SAW filtering. I know it's fixed filtering, but might
>have some advantages.
>
>
>The required 90 degree phase shift in the modulation is covered. And, it's
>pretty easy to supply LO with quadrature, too. It's the 'balance' in the
>mixers I need to get that pesky carrier tone and 3rd harmonics down.
>
>100MHz has 10ns period. 0.1nS is almost 4 degrees phase shift! Somehow
>will have to make the system accomodate that kind of sloppy action, too
LO never was a worry, the 9 decades band on the input was the bother.
Gettting 90 degrees over more than three decades (from a single source) is
all but impossible.
That said there are multipliers that are good from DC to 250 MHz or more.
I do not always like square wave drive on diode rings because of the
harmonics which are usually just filtered out but that doesn't seem like
it would work in your application as stated so far.
If you can keep f(m) << f(c) the designs get a lot easier. If you can do
band switching for <10 Hz and > 100 kHz it could really help.
?-)
RobertMacy
Apr 5, 2014, 11:04:51 PM4/5/14
to
On Fri, 04 Apr 2014 18:26:16 -0700, Phil Hobbs <ho...@electrooptical.net>
wrote:
>> ...snip....
That appears quite a bit more 'linear' than the MC1496. Can you confirm
that it is?
wrote:
>> ...snip....
> The MC1496 doesn't use the diode trick. Check out the LM13700 datasheet
> or the following listing.
>
Wow!
> or the following listing.
>
That appears quite a bit more 'linear' than the MC1496. Can you confirm
that it is?
Phil Hobbs
Apr 6, 2014, 10:00:28 AM4/6/14
to
ohms, so replace them with 250 ohm resistors and run an apples-to-apples
comparison.
Cheers
Phil Hobbs
Clifford Heath
Apr 6, 2014, 8:30:31 PM4/6/14
to
On 07/04/14 00:00, Phil Hobbs wrote:
> On 4/5/2014 11:04 PM, RobertMacy wrote:
>> On Fri, 04 Apr 2014 18:26:16 -0700, Phil Hobbs
>> <ho...@electrooptical.net> wrote:
>>
>>>> ...snip....
>>> The MC1496 doesn't use the diode trick. Check out the LM13700
>>> datasheet or the following listing.
>>>
>> Wow!
>>
>> That appears quite a bit more 'linear' than the MC1496. Can you confirm
>> that it is?
>
> Sure. The 'diodes' are running at a dynamic impedance of around 250
> ohms, so replace them with 250 ohm resistors and run an apples-to-apples
> comparison.
I don't think I understand. I removed Q3 and Q4 and replaced them by 250
> On 4/5/2014 11:04 PM, RobertMacy wrote:
>> On Fri, 04 Apr 2014 18:26:16 -0700, Phil Hobbs
>> <ho...@electrooptical.net> wrote:
>>
>>>> ...snip....
>>> The MC1496 doesn't use the diode trick. Check out the LM13700
>>> datasheet or the following listing.
>>>
>> Wow!
>>
>> That appears quite a bit more 'linear' than the MC1496. Can you confirm
>> that it is?
>
> Sure. The 'diodes' are running at a dynamic impedance of around 250
> ohms, so replace them with 250 ohm resistors and run an apples-to-apples
> comparison.
ohms to ground, and I can see no significant change in either the
waveform or the FFT.
Clifford Heath.
Phil Hobbs
Apr 7, 2014, 8:39:36 AM4/7/14
to
stepping that and see what happens in each case.
RobertMacy
Apr 7, 2014, 11:10:06 AM4/7/14
to
On Fri, 04 Apr 2014 17:48:44 -0700, Phil Hobbs <ho...@electrooptical.net>
wrote:
>> ...snip...
appropriate noise source(s) to every 'important' component. But you saw
that already, right?
wrote:
>> ...snip...
> So how does it work?
>
> Cheers
>
> Phil Hobbs
>
Phil, there is NO magic in what I did. Simply brute force...'add' the
>
> Cheers
>
> Phil Hobbs
>
appropriate noise source(s) to every 'important' component. But you saw
that already, right?
It is loading more messages.
0 new messages