Notches in ground planes for multi-power multi-channel board
namde...@gmail.com
I am designing a multi-channel board and I have individual linear
regulators for each channel running off a common switcher.
Now, my concern is that due to the split power planes, which are like
thin fingers running through the board, I will cause potential
differences based on the current in different channels, especially,
because I have a common ground plane. Now, I was thinking my choices
are:
1) Split the ground plane as well, so I would have thin finger like
power and ground plane separated by narrow notch but connected at the
linear regulator end and at the input end. (All the channels are w.r.t
common ground plane.)
The trouble is, that I have common (to all channels) control lines
runnning across my planes (though separated by 2 layers) and even if I
isolate the grounds, some noise is bound to take the signal lines.
2) Keep split power planes, but add a narrow slice ground plane
between the power plane 'fingers' in between them on the same layer as
the power planes.
3) Keep 1, but put connections between ground planes at regular
intervals.
Anyone have thoughts on this. The voltages are +5 or less and it is
all analog..there are digital lines and a digital section, but they
come into the analog area through an isolator.
Joerg
> All,
>
> I am designing a multi-channel board and I have individual linear
> regulators for each channel running off a common switcher.
>
> Now, my concern is that due to the split power planes, which are like
> thin fingers running through the board, I will cause potential
> differences based on the current in different channels, especially,
> because I have a common ground plane. Now, I was thinking my choices
> are:
>
> 1) Split the ground plane as well, so I would have thin finger like
> power and ground plane separated by narrow notch but connected at the
> linear regulator end and at the input end. (All the channels are w.r.t
> common ground plane.)
>
Splitting a ground plane is usually a recipe for disaster. I have yet to
see a case where that really worked and it's been decades now. In audio
designs it can work but only until a strong RF field shows up, upon
which all hell breaks loose.
OTOH if nobody split planes anymore I'd have less work so it does have
some upsides ;-)
> The trouble is, that I have common (to all channels) control lines
> runnning across my planes (though separated by 2 layers) and even if I
> isolate the grounds, some noise is bound to take the signal lines.
>
That's one of the reasons why it's a recipe for disaster ;-)
> 2) Keep split power planes, but add a narrow slice ground plane
> between the power plane 'fingers' in between them on the same layer as
> the power planes.
>
Copper pour on the same plane as power? You can do that but via it
through here and there. And no thermal reliefs for those vias unless
something needs to be soldered in there.
> 3) Keep 1, but put connections between ground planes at regular
> intervals.
>
That's like saying "Let's split it, oh, wait, maybe split them only a
little". I would not split the ground at all.
> Anyone have thoughts on this. The voltages are +5 or less and it is
> all analog..there are digital lines and a digital section, but they
> come into the analog area through an isolator.
Make sure your power is clean and the switcher doesn't chatter through.
Use shielded versions for the magnetics if it's close by. Design your
stuff with reasonable PSRR or bypass extra good where you can't, like
for transistor stages.
--
Regards, Joerg
legg
Best just to reference and decouple the output regulators to the load
at the ground plane of the load. input decoupling will only work with
an intervening impedance at the unregulated input.
If the switcher outputs had individual returns (ie were isolated) you
might do some funny stuff with the grounds before the regulators, but
not with a common return. You might consider runnung the switcher
return to an intelligent board location, before it hits the ground
plane.
RL
John Larkin
Keep a single, solid ground plane. Break the power plane up into
individual pours, each fed from its own regulator.
That's the best general advice I can give without getting involved in
the detailed design of your product (and you probably can't afford
me.)
John
namde...@gmail.com
wrote:
> namdegue...@gmail.com wrote:
> > All,
>
> > I am designing amulti-channelboard and I have individual linear
>
> - Show quoted text -
Well, actually, I am not splitting the ground planes, in the sense
that they are connected at the same potential at all the linear
regulators and at the input. But with my control lines, crossing
across my parts, my split is not going to have much effect. I could
put de-coupling caps across the notch to allow high frequencies to
pass, but again they might go through my control lines and not bother
with the caps...basically it is big ?..so yeah, I think I am not going
to split the ground...best to do it when I have free time as a test
project.
For my thin ground slice in between the power planes (on the same
layer0, I put vias every 50 mils or so and no themal reliefs and it is
a relatively unpopulated location, so that will help. My switcher and
linear regulator are LT parts 3430/1763, but I have been told by just
about every amplifier part vendor (and LT themselves!) that linear
regulators don't attenuate frequencies above 100 kHz very well, so I
might keep a 100 kHz switching frequency and maybe put ferrites at the
linear regulator input.
I am also considering using ceramic 10 uF caps instead of the tantalum
datasheets reference. Also some 0508/0402 caps.
namde...@gmail.com
>
> - Show quoted text -
Unfortunately, my switcher is teh standard run of the mill, I could
use a DC-DC converter with isolation probably. My switcher return is
almost immediately goes to the ground plane as per the manufacturer
(LT) recommendation, however I can connect the grounds of the linear
regulator and the switcher to a common location and then send it to
the ground plane.
Well
namde...@gmail.com
<jjlar...@highNOTlandTHIStechnologyPART.com> wrote:> >I am designing amulti-channelboard and I have individual linear
>
> - Show quoted text -
Thanks, that is very helpful, I will stick with one plane and
hopefully the ground 'slice' between the power planes will help as
well. Also, I am wondering if putting decoupling caps between two
power plane pours will help further
John Larkin
Do NOT slice the ground plane. Bypass each power pour to ground.
Over and out.
John
Guy Macon
namde...@gmail.com wrote:
>Well, actually, I am not splitting the ground planes, in the sense
>that they are connected at the same potential at all the linear
>regulators and at the input.
Only one of the following can be true:
[A] Your ground plane has the bare minimum of interruptions needed
to accommodate signal vias, etc., in which case you have not split
the ground plane.
[B] you have purposely added some sort of notch, gap, hole, or
other shape of non-copper that you could have filled with copper,
in which case you have split the ground plane.
Some splits are worse than others, but case [A] is almost
always the best choice unless you are doing something really
specialized, know what you are doing, and are willing to
build and test the proposed topology to prove that case [A]
doesn't beat it. And even then case [A] tends to be better
when conditions you didn't envision arise.
>I could put de-coupling caps across the notch to allow high
>frequencies to pass
Replace the caps with copper (no notches). Use those caps to
bypass high frequencies your power plane(s) to your ground plane.
Unlike splitting the ground plane, having a split power plane
with separate onboard regulators for different sections is
usually a good idea, depending on your cost constraints.
>My switcher and linear regulator are LT parts 3430/1763,
>but I have been told by just about every amplifier part
>vendor (and LT themselves!) that linear regulators don't
>attenuate frequencies above 100 kHz very well,
It is always better to kill as much switching noise as
possible before feeding it into an analog regulator.
>so I might keep a 100 kHz switching frequency
That's a trade-off too complex to discuss here. Higher
switcher frequencies are better at blowing past linear
regulators, jumping through the air, etc. but are easier
to kill at susceptible IC or transistor. Lower frequencies
are just the opposite.
>and maybe put ferrites at the linear regulator input.
If you are building a switcher on the board, use the layout
from the app notes and data sheets for your switcher. If
it is the usual separate box, putting ferrites on the wires
is a cheap way of stopping some of the noise. Please note
that choosing ferrites and deciding where to put them is
an engineering task in itself. How big? What frequency?
Differential mode (bead per wire) or common mode (power and
return through one bigger bead)? Saturation problems?
Look for app notes that will help with those decisions.
>I am also considering using ceramic 10 uF caps instead of
>the tantalum datasheets reference.
Why? Do you really think that you know more than the
engineer who wrote the datasheet?
>Also some 0508/0402 caps.
It is almost always a good idea to put a 0.1 ceramic
bypass cap at every power pin of every IC, depending
on your cost constraints.
>For my thin ground slice in between the power planes (on the
>same layer0, I put vias every 50 mils or so and no themal
>reliefs and it is a relatively unpopulated location, so
>that will help.
[from a later post]
>Thanks, that is very helpful, I will stick with one plane and
>hopefully the ground 'slice' between the power planes will help as
>well.
In general, that space is better used to make the power planes
bigger.
>Also, I am wondering if putting decoupling caps between two
>power plane pours will help further
Not a good plan. Put those decoupling caps between each
power plane and the ground plane instead.
--
Guy Macon <http://www.guymacon.com/> Guy Macon <http://www.guymacon.com/>
Guy Macon <http://www.guymacon.com/> Guy Macon <http://www.guymacon.com/>
Guy Macon <http://www.guymacon.com/> Guy Macon <http://www.guymacon.com/>
Guy Macon <http://www.guymacon.com/> Guy Macon <http://www.guymacon.com/>
MooseFET
> All,
>
> I am designing a multi-channel board and I have individual linear
> regulators for each channel running off a common switcher.
>
> Now, my concern is that due to the split power planes, which are like
> thin fingers running through the board, I will cause potential
> differences based on the current in different channels, especially,
> because I have a common ground plane. Now, I was thinking my choices
> are:
>
> 1) Split the ground plane as well, so I would have thin finger like
> power and ground plane separated by narrow notch but connected at the
> linear regulator end and at the input end. (All the channels are w.r.t
> common ground plane.)
>
> The trouble is, that I have common (to all channels) control lines
> runnning across my planes (though separated by 2 layers) and even if I
> isolate the grounds, some noise is bound to take the signal lines.
>
> 2) Keep split power planes, but add a narrow slice ground plane
> between the power plane 'fingers' in between them on the same layer as
> the power planes.
>
> 3) Keep 1, but put connections between ground planes at regular
> intervals.
4) Add yet another layer to the design. This extra layer would be the
"signal return" layer and won't carry the power currents.
Joerg
Easy fix if you need good rejection: Use a npn-Darlington with enough
heat dissipation capability to drop around 1.5V at your maximum current.
Collector to switcher output, emitter to linear regulator input, 1K
resistor from base to collector, 47uF cap from base to ground. Plus a
couple discharge diodes. That'll behave like having a shoe size 15
electrolytic in there.
Also, consider using the LM317. With a cap at the adjust pin it'll
maintain a low output impedance all the way up to a MHz:
http://www.national.com/ds/LM/LM117.pdf
Personally I'd resist the temptation to use an LDO. I've seen them bring
a lot of grief.
> I am also considering using ceramic 10 uF caps instead of the tantalum
> datasheets reference. Also some 0508/0402 caps.
Tantalums in bypass applications are quite evil IMHO. Seen too many
orange-greenish clouds wafting up after the explosion and so on.
Joerg
Careful, that can develop it's own nasty form of life. Just like a
subway station in a seedy part of town.
John Larkin
>I am also considering using ceramic 10 uF caps instead of the tantalum
>datasheets reference. Also some 0508/0402 caps.
Many linear regs are unstable with ceramics on the output. The
suggestion to use tantalums is their Zen way of saying so without
saying so.
John
qrk
Keep one solid ground plane.
Using a linear regulator on each channel is a good way to isolate
power supply cross talk, so, keep that in.
You will have more problems with that switcher getting into things.
The switcher needs to be properly laid out on the pcb or you will get
switching noise into places you don't want it. Linear Tech. has the
layout of their reference designs for most or all of their switchers.
If you can't find it on LT's web site, call your LT rep. LT has
excellent customer support for the small guy. It is a good idea to use
LT's layout as a reference. If you can, use one of the MHz plus
switchers, then, apply ferrite beads to input and output of the
switcher.
---
Mark
Joel Koltner
news:r791r3p5jvbprji4m...@4ax.com...
> Many linear regs are unstable with ceramics on the output. The
> suggestion to use tantalums is their Zen way of saying so without
> saying so.
It'd be handy if someone started collecting a "decoder ring" between what data
sheets say and what they really mean.
"It is suggested to use tantalum caps on the output of this regulator" -->
"This regulator is unstable with too low of ESR output caps."
Ever use chips by FTDI? They take the approach that if you want to do
something "easy" based on the register descriptions... but due to chip bugs it
doesn't actually WORK that way... they never mention there IS a bug, but
instead provide circuitous example code that works around it!
Joerg
Thou shalt not use LDOs ;-)
Joel Koltner
news:K33sj.4087$xq2....@newssvr21.news.prodigy.net...
> Thou shalt not use LDOs ;-)
Every time someone says they don't believe in LDOs a designer at Analog
Devices, Linear, or Maxim dies. :-)
I suspect you've rolled a few discrete transistor-based LDO designs in your
time, Joerg!
John Larkin
<zapwireD...@yahoo.com> wrote:
>"John Larkin" <jjla...@highNOTlandTHIStechnologyPART.com> wrote in message
>news:r791r3p5jvbprji4m...@4ax.com...
>> Many linear regs are unstable with ceramics on the output. The
>> suggestion to use tantalums is their Zen way of saying so without
>> saying so.
>
>It'd be handy if someone started collecting a "decoder ring" between what data
>sheets say and what they really mean.
>
>"It is suggested to use tantalum caps on the output of this regulator" -->
>"This regulator is unstable with too low of ESR output caps."
Eventually one becomes sensitized to this sort of new-speak. Even more
fun is when you have to research the part numbers that they call out
on the eval board, because *their* esr's are the only hint about what
might work. Micrel is especially good at this trick.
>
>Ever use chips by FTDI? They take the approach that if you want to do
>something "easy" based on the register descriptions... but due to chip bugs it
>doesn't actually WORK that way... they never mention there IS a bug, but
>instead provide circuitous example code that works around it!
>
This sort of obscurity means that many, many users spend many, many
man-weeks and board spins discovering these facts over and over.
John
Joerg
> "Joerg" <notthis...@removethispacbell.net> wrote in message
> news:K33sj.4087$xq2....@newssvr21.news.prodigy.net...
>> Thou shalt not use LDOs ;-)
>
> Every time someone says they don't believe in LDOs a designer at Analog
> Devices, Linear, or Maxim dies. :-)
>
And if I'd sing my song about Maxim their CEO might clutch his chest ;-)
> I suspect you've rolled a few discrete transistor-based LDO designs in your
> time, Joerg!
>
Yes, I did. But then you know how it'll behave. Problem with modern
chips is that they don't give out much information. For example, TI
refused to release the SPICE model for one of their TPS chips after some
went unexplicably "exotherm" on our boards. They also refused to run our
simple circuitry around it. So I gave it the ax, for good. I sure won't
use one anytime soon.
Joerg
> On Mon, 11 Feb 2008 12:12:15 -0800, "Joel Koltner"
> <zapwireD...@yahoo.com> wrote:
>
>> "John Larkin" <jjla...@highNOTlandTHIStechnologyPART.com> wrote in message
>> news:r791r3p5jvbprji4m...@4ax.com...
>>> Many linear regs are unstable with ceramics on the output. The
>>> suggestion to use tantalums is their Zen way of saying so without
>>> saying so.
>> It'd be handy if someone started collecting a "decoder ring" between what data
>> sheets say and what they really mean.
>>
>> "It is suggested to use tantalum caps on the output of this regulator" -->
>> "This regulator is unstable with too low of ESR output caps."
>
> Eventually one becomes sensitized to this sort of new-speak. Even more
> fun is when you have to research the part numbers that they call out
> on the eval board, because *their* esr's are the only hint about what
> might work. Micrel is especially good at this trick.
>
Even if they give you an ESR range it'll be in the figure section and
thus typical, not guaranteed. Personally I do not use LDOs with
non-guaranteed ESR ranges. Which pretty much excludes most of them. The
ones that aren't excluded (stable down to zero and stuff) are often out
of budget range.
>> Ever use chips by FTDI? They take the approach that if you want to do
>> something "easy" based on the register descriptions... but due to chip bugs it
>> doesn't actually WORK that way... they never mention there IS a bug, but
>> instead provide circuitous example code that works around it!
>>
>
> This sort of obscurity means that many, many users spend many, many
> man-weeks and board spins discovering these facts over and over.
>
If there is a better device from the competition just move on.
Joerg
And by golly, use shielded inductors on those. Else you might sing the
blues at the EMC lab.
John Larkin
I recently did a 3.3 to 2.5 volt "regulator" to power VccAux of a
Xilinx FPGA. I used a zener diode. In series. Backwards.
John
Joel Koltner
> I recently did a 3.3 to 2.5 volt "regulator" to power VccAux of a
> Xilinx FPGA. I used a zener diode. In series. Backwards.
Nice... but why not just a regular silicon diode? Better regulation or
something?
I do like the trick of using LEDs as a combination means of setting up biases
as well as indcating power/activity/whatever.
Joerg
> "John Larkin" <jjla...@highNOTlandTHIStechnologyPART.com> wrote in message
> news:dbu1r3tiuekmbilnp...@4ax.com...
>> I recently did a 3.3 to 2.5 volt "regulator" to power VccAux of a
>> Xilinx FPGA. I used a zener diode. In series. Backwards.
>
> Nice... but why not just a regular silicon diode? Better regulation or
> something?
>
Or two Schottkys. But I guess the zener costs less.
> I do like the trick of using LEDs as a combination means of setting up biases
> as well as indcating power/activity/whatever.
>
Be careful. Suddenly someone releases another kind of red (Oh, it looks
much nicer ...) and your biases are all off the rocker.
John Larkin
The voltage drop came out right, 0.8 exactly. Plus, the melf package
we have in stock exactly bridged the pads of the LDO that didn't work.
Yeah, an led can make a useful bias voltage, and free light.
John
Joerg
What did the LDO do? The usual, such as oscillating?
> Yeah, an led can make a useful bias voltage, and free light.
>
--
Regards, Joerg
John Larkin
<notthis...@removethispacbell.net> wrote:
>John Larkin wrote:
>> On Mon, 11 Feb 2008 17:54:48 -0800, "Joel Koltner"
>> <zapwireD...@yahoo.com> wrote:
>>
>>> "John Larkin" <jjla...@highNOTlandTHIStechnologyPART.com> wrote in message
>>> news:dbu1r3tiuekmbilnp...@4ax.com...
>>>> I recently did a 3.3 to 2.5 volt "regulator" to power VccAux of a
>>>> Xilinx FPGA. I used a zener diode. In series. Backwards.
>>> Nice... but why not just a regular silicon diode? Better regulation or
>>> something?
>>>
>>> I do like the trick of using LEDs as a combination means of setting up biases
>>> as well as indcating power/activity/whatever.
>>>
>>
>> The voltage drop came out right, 0.8 exactly. Plus, the melf package
>> we have in stock exactly bridged the pads of the LDO that didn't work.
>>
>
>What did the LDO do? The usual, such as oscillating?
>
>
It dropped out! Actually, I used an LM1117, which has a dropout of
around 1 volt. My bad.
The diode seems remarkably stable over a wide range of cap esr's.
John
John Larkin
>John Larkin a écrit :
>Then try a TD or a GUNN :-)
I used to love TD's. Too bad nobody makes them any more; GPD used to,
but I don't think they still do. The fabrication process was insane,
but it's something that an amateur could potentially do.
John
Joerg
Happened to all the boutique diodes. AFAIR you had found a source for
lower cost SRDs a while ago. But the thing is, if Digikey doesn't carry
something chances are it'll go lalaland soon.
Jim Thompson
There's been a trend in I/C design going on, actually, for quite some
time... driven by you parts consumers...
You want to buy a circuit function that requires NO ADDITIONAL PARTS
or PINS.
Therefore your LDO can't have the proper feed-forward compensation to
adequately retain stability with capacitive loads... the manufacturer
fudges by specifying ESR.
So it's your own fault ;-)
...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona Voice:(480)460-2350 | |
| E-mail Address at Website Fax:(480)460-2142 | Brass Rat |
| http://www.analog-innovations.com | 1962 |
America: Land of the Free, Because of the Brave
Phil Hobbs
Nice and quiet, too, if you can spend some current on it, especially if
you use a series string to make higher voltages instead of multiplying
it up.
Cheers,
Phil Hobbs
Phil Hobbs
Capacitance multipliers work dramatically better than VRs for cleaning
up SMPS crap.
Cheers,
Phil Hobbs
JosephKK
Hey gotta keep the economy going.
JosephKK
> Joel Koltner wrote:
>> "Joerg" <notthis...@removethispacbell.net> wrote in message
>> news:K33sj.4087$xq2....@newssvr21.news.prodigy.net...
>>> Thou shalt not use LDOs ;-)
>>
>> Every time someone says they don't believe in LDOs a designer at
>> Analog Devices, Linear, or Maxim dies. :-)
>>
>
> And if I'd sing my song about Maxim their CEO might clutch his chest ;-)
>
You simply must give us the libretto now.
Joerg
> On Mon, 11 Feb 2008 14:12:54 -0800, Joerg
> <notthis...@removethispacbell.net> wrote:
>
>> Joel Koltner wrote:
>>> "Joerg" <notthis...@removethispacbell.net> wrote in message
>>> news:K33sj.4087$xq2....@newssvr21.news.prodigy.net...
>>>> Thou shalt not use LDOs ;-)
>>> Every time someone says they don't believe in LDOs a designer at Analog
>>> Devices, Linear, or Maxim dies. :-)
>>>
>> And if I'd sing my song about Maxim their CEO might clutch his chest ;-)
>>
>>
>>> I suspect you've rolled a few discrete transistor-based LDO designs in your
>>> time, Joerg!
>>>
>> Yes, I did. But then you know how it'll behave. Problem with modern
>> chips is that they don't give out much information. For example, TI
>> refused to release the SPICE model for one of their TPS chips after some
>> went unexplicably "exotherm" on our boards. They also refused to run our
>> simple circuitry around it. So I gave it the ax, for good. I sure won't
>> use one anytime soon.
>
> There's been a trend in I/C design going on, actually, for quite some
> time... driven by you parts consumers...
>
> You want to buy a circuit function that requires NO ADDITIONAL PARTS
> or PINS.
>
Yeah but you's guys, the chip designers, typically don't ask us guys
before releasing a chip to the public. I would not mind at all if an RC
or whatever has to be connected if that makes an LDO unconditionally
stable. After all we are all used to that from PWM chips for switchers.
Before doing the architecture have your people call my people ;-)
> Therefore your LDO can't have the proper feed-forward compensation to
> adequately retain stability with capacitive loads... the manufacturer
> fudges by specifying ESR.
>
> So it's your own fault ;-)
>
<harumph ... grumble>
Joerg
> Joerg wrote:
>> Joel Koltner wrote:
>>> "Joerg" <notthis...@removethispacbell.net> wrote in message
>>> news:K33sj.4087$xq2....@newssvr21.news.prodigy.net...
>>>> Thou shalt not use LDOs ;-)
>>>
>>> Every time someone says they don't believe in LDOs a designer at
>>> Analog Devices, Linear, or Maxim dies. :-)
>>>
>>
>> And if I'd sing my song about Maxim their CEO might clutch his chest ;-)
>>
>
> You simply must give us the libretto now.
>
In the army our drill sergeant told two guys to stop singing. One of
them is yours truly.
>>
>>> I suspect you've rolled a few discrete transistor-based LDO designs
>>> in your time, Joerg!
>>>
>>
>> Yes, I did. But then you know how it'll behave. Problem with modern
>> chips is that they don't give out much information. For example, TI
>> refused to release the SPICE model for one of their TPS chips after
>> some went unexplicably "exotherm" on our boards. They also refused to
>> run our simple circuitry around it. So I gave it the ax, for good. I
>> sure won't use one anytime soon.
>>
Jim Thompson
Takes an extra pin.
>
>Before doing the architecture have your people call my people ;-)
My people don't decide the architecture. You should know that well.
MARKETING decides ;-)
>
>
>> Therefore your LDO can't have the proper feed-forward compensation to
>> adequately retain stability with capacitive loads... the manufacturer
>> fudges by specifying ESR.
>>
>> So it's your own fault ;-)
>>
>
><harumph ... grumble>
<sno-o-o-o-ort>
JosephKK
So when do we get the libretto?
Joerg
Like in singing a requiem? No way my vocal cords would even get up there.
Frank Raffaeli
> On Feb 10, 8:49 pm,John Larkin
>
>
>
> <jjlar...@highNOTlandTHIStechnologyPART.com> wrote:> > >I am designing amulti-channelboard and I have individual linear
> > >regulators for each channel running off a common switcher.
>
> > >Now, my concern is that due to the split power planes, which are like
> > >thin fingers running through the board, I will cause potential
> > >differences based on the current in different channels, especially,
> > >because I have a common ground plane. Now, I was thinking my choices
> > >are:
>
> > >1) Split the ground plane as well, so I would have thin finger like
> > >power and ground plane separated by narrow notch but connected at the
> > >linear regulator end and at the input end. (All the channels are w.r.t
> > >common ground plane.)
>
> > >The trouble is, that I have common (to all channels) control lines
> > >runnning across my planes (though separated by 2 layers) and even if I
> > >isolate the grounds, some noise is bound to take the signal lines.
>
> > >2) Keep split power planes, but add a narrow slice ground plane
> > >between the power plane 'fingers' in between them on the same layer as
> > >the power planes.
>
> > >3) Keep 1, but put connections between ground planes at regular
> > >intervals.
>
> > >Anyone have thoughts on this. The voltages are +5 or less and it is
> > >all analog..there are digital lines and a digital section, but they
> > >come into the analog area through an isolator.
>
> > individual pours, each fed from its own regulator.
>
> > That's the best general advice I can give without getting involved in
> > the detailed design of your product (and you probably can't afford
> > me.)
>
> > John- Hide quoted text -
>
> > - Show quoted text -
>
> Thanks, that is very helpful, I will stick with one plane and
> hopefully the ground 'slice' between the power planes will help as
> well. Also, I am wondering if putting decoupling caps between two
> power plane pours will help further
It seems intuitive that a gap across copper will reduce current flow
across the gap. This is true only at low frequencies. If you have any
digital logic on this board, and there are any signal wires across the
"gap", you've made an efficient transmitting / receive antenna. I
think Joerg said if people didn't put gaps in their ground planes, he
would have less work ;-) Sometimes the max current flows across the
center of the gap. It's frequency selective.
Frank
John Larkin
Slot antennas and slotline transmission lines both suggest modes
better left alone.
John
Didi
> see a case where that really worked and it's been decades now.
I have pointed you to that before so it is not that you have yet to
see one, it is that you have yet to look at one :-).
http://tgi-sci.com/tgi/hstb.htm
I gree in the context of the OP splitting the GND plane is a bad idea,
probably very bad. My guess is the split power plane can still be
made having wide enough sections so this will not be an issue, but
I did not get into all the details.
Spliiting the GND plane gets useful way beyond commonplace resolution
and sensitivity, this is why most people have no problem without
splitting it, and being happy with their results think this is a
usiversal rule.
It is not.
Dimiter
------------------------------------------------------
Dimiter Popoff Transgalactic Instruments
http://www.tgi-sci.com http://www.tgi-sci.com/dsv/ <-- new
section....
------------------------------------------------------
http://www.flickr.com/photos/didi_tgi/sets/72157600228621276/
Joerg wrote:
> namde...@gmail.com wrote:
> > All,
> >
> > I am designing a multi-channel board and I have individual linear
> > regulators for each channel running off a common switcher.
> >
> > Now, my concern is that due to the split power planes, which are like
> > thin fingers running through the board, I will cause potential
> > differences based on the current in different channels, especially,
> > because I have a common ground plane. Now, I was thinking my choices
> > are:
> >
> > 1) Split the ground plane as well, so I would have thin finger like
> > power and ground plane separated by narrow notch but connected at the
> > linear regulator end and at the input end. (All the channels are w.r.t
> > common ground plane.)
> >
>
> Splitting a ground plane is usually a recipe for disaster. I have yet to
> see a case where that really worked and it's been decades now. In audio
> designs it can work but only until a strong RF field shows up, upon
> which all hell breaks loose.
>
> OTOH if nobody split planes anymore I'd have less work so it does have
> some upsides ;-)
>
>
> > The trouble is, that I have common (to all channels) control lines
> > runnning across my planes (though separated by 2 layers) and even if I
> > isolate the grounds, some noise is bound to take the signal lines.
> >
>
> That's one of the reasons why it's a recipe for disaster ;-)
>
>
> > 2) Keep split power planes, but add a narrow slice ground plane
> > between the power plane 'fingers' in between them on the same layer as
> > the power planes.
> >
>
> Copper pour on the same plane as power? You can do that but via it
> through here and there. And no thermal reliefs for those vias unless
> something needs to be soldered in there.
>
>
> > 3) Keep 1, but put connections between ground planes at regular
> > intervals.
> >
>
> That's like saying "Let's split it, oh, wait, maybe split them only a
> little". I would not split the ground at all.
>
>
> > Anyone have thoughts on this. The voltages are +5 or less and it is
> > all analog..there are digital lines and a digital section, but they
> > come into the analog area through an isolator.
>
>
> Make sure your power is clean and the switcher doesn't chatter through.
> Use shielded versions for the magnetics if it's close by. Design your
> stuff with reasonable PSRR or bypass extra good where you can't, like
> for transistor stages.
Joerg
Mostly people end up with a huge loop antenna that picks up just about
anything. That is because stuff will have to be connected to either side
of a split, usually. Now throw something like Sutro tower into the mix
and the circuit noise can reach the signal levels of a Led Zeppelin concert.
John Larkin
wrote:
>> Splitting a ground plane is usually a recipe for disaster. I have yet to
>> see a case where that really worked and it's been decades now.
>
>I have pointed you to that before so it is not that you have yet to
>see one, it is that you have yet to look at one :-).
>http://tgi-sci.com/tgi/hstb.htm
Interesting spec... Zero Offset: None
John
Joel Koltner
message news:8774r31ajh7eaii8v...@4ax.com...
> My people don't decide the architecture. You should know that well.
> MARKETING decides ;-)
I'm sure it has something to do with the desire from many companies'
management to be able to hire relatively unskilled engineers (...they're
cheap!) and still have a product.
Arguably that's not a particularly awful incentive, but few companies seem to
realize that the whole approach fails as soon as you want to do something just
a little bit different than what the whiz-bang "all-in-one" IC does.
That's when people give Joerg a call.
Or the company I work at, for that matter. We spend plenty of time these days
building radios from all the discrete mixers/amps/etc. that are now readily
available, although unfortuntaely it's often difficult to compete on power
relative to what some of those old Philips (NXP) parts provided. (Remeber the
UAA2080? Beautiful part... very low power, good quality filters using
gyrators...)
At least battery technology has progressed a bit.
---Joel
Richard Henry
wrote:
>
> - Show quoted text -
On the other hand, if your circuitry is completely isolated into
physical/electrical/logical domains (such as in red/black segregation)
then splitting or slotting the ground plane makes sense and may be
required by the physical and/or regulatory realities.
Didi
Pretty unique, yes. And guaranteed...
This is the only MCA on the market which allows you to calibrate
for energy using a single point.
Dimiter
------------------------------------------------------
Dimiter Popoff Transgalactic Instruments
http://www.tgi-sci.com
------------------------------------------------------
http://www.flickr.com/photos/didi_tgi/sets/72157600228621276/
Joerg
> "Jim Thompson" <To-Email-Use-Th...@My-Web-Site.com> wrote in
> message news:8774r31ajh7eaii8v...@4ax.com...
>> My people don't decide the architecture. You should know that well.
>> MARKETING decides ;-)
>
> I'm sure it has something to do with the desire from many companies'
> management to be able to hire relatively unskilled engineers (...they're
> cheap!) and still have a product.
>
> Arguably that's not a particularly awful incentive, but few companies seem to
> realize that the whole approach fails as soon as you want to do something just
> a little bit different than what the whiz-bang "all-in-one" IC does.
>
> That's when people give Joerg a call.
>
Sometimes they give me a call a bit late, after the purchasing guy
became unable to source some super-duper panacea chip. Then it's often
"tabula rasa", scrap the whole design and start from scratch.
> Or the company I work at, for that matter. We spend plenty of time these days
> building radios from all the discrete mixers/amps/etc. that are now readily
> available, although unfortuntaely it's often difficult to compete on power
> relative to what some of those old Philips (NXP) parts provided. (Remeber the
> UAA2080? Beautiful part... very low power, good quality filters using
> gyrators...)
>
Personally I am not a fan of such "panacea chips". The millisecond they
lose a key account the manufacturers might stop production and soon
after purge the datasheet.
> At least battery technology has progressed a bit.
>
However, electronics and especially the SW guys keep bloating stuff
faster than battery technology can keep up with. Case in point: My early
90's laptop ran up to six hours on one NiCd charge and so I opted not to
buy the extra capacity battery. Nowadays it's down to two hours with an
expensive LiIon battery that has many times the capacity. And I do the
same type of work on them :-)
The concept of a mobile device needing a cooling fan is IMHO just plain
sick.
Or take our Dimango remote doorbell. 20 bucks, 10 (!) years on the first
set of batteries and no slowdown in sight. Of course, the receiver is
completely discrete. This year I am going to change the batteries anyhow
because they might leak just from old age.
Joerg
C'mon, Didi lives and works in Bulgaria and probably speaks a language
of Macedonian origin all day long. Quite different from our Anglo-Saxon
language roots. It's much tougher for those guys to master English than
it is for Western Europeans. Guess he meant offset from the zero-line.
Didi
> of Macedonian origin all day long. Quite different from our Anglo-Saxon
> language roots. It's much tougher for those guys to master English than
> it is for Western Europeans. Guess he meant offset from the zero-line.
Thanks Joerg,
it is not down to my English, I suppose it is good enough.
I have just switched words, the spec should have been called
offset-0, this is how it is known. I suppose people who know about
MCAs would not pay much attention to my mistake, this is a common
spec on that stuff (exactly what you guessed). I'll fix it, though.
Dimiter
------------------------------------------------------
Dimiter Popoff Transgalactic Instruments
http://www.tgi-sci.com
------------------------------------------------------
http://www.flickr.com/photos/didi_tgi/sets/72157600228621276/
John Larkin
wrote:
>> Interesting spec... Zero Offset: None
>
>Pretty unique, yes. And guaranteed...
>This is the only MCA on the market which allows you to calibrate
>for energy using a single point.
>
>Dimiter
What's the percentage tolerance spec on your zero?
John
Didi
It can be seen (and specified) as 0%. The INL and DNL are non-0, of
course, so it
is safe to specify that (which, in essence, is what counts - you can
calibrate by 1 point without giving it much thought).
The 0 is mathematically correct, and while the input data will always
carry some error, it is randomized and spread over many samples
(hundreds or thousands), so that specifying a non-0 number (say, 0.1%
LSB...)
would make no sense (and would convey perhaps the wrong message
to users, they would have to look for error where there is none).
The INL is within 1 LSB or so, it safely covers for the offset-0 being
specified as
100% 0. So does even the DNL, BTW, I have specified a conservative
number,
yet the fact is it is so low I could not measure it when I did that
(and I could
measure 0.5% LSB DNL on my older, Wilkinson based devices).
Dimiter
------------------------------------------------------
Dimiter Popoff Transgalactic Instruments
http://www.tgi-sci.com
------------------------------------------------------
http://www.flickr.com/photos/didi_tgi/sets/72157600228621276/
Michael A. Terrell
>
> In the army our drill sergeant told two guys to stop singing. One of
> them is yours truly.
That's because he as afraid everyone else would go AWOL, or give you
two a GI shower and blanket party. ;-)
--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.
Michael A. Terrell
Central Florida
JosephKK
I want the words, the prosody, not the music nor the melody for whatever
voice.
Joerg
Well, since I can't sing I don't have any music texts. Just like I don't
have sheet music.
Joerg
> Say Joerg,
>
> I just thought I'd mention...
>
> Some purchasing guy was talking with me about various parts for some project
> yesterday and I noticed that we're using Linear Tech's LTC1844 in numerous
> different products and at least so far no one's reporting seeing one
> misbehave. This is a 150mA LDO (90mV headroom required at 150mA, 30mV at
> 50mA). The data sheet has a chart that claims it's guaranteed stable so long
> as the output capacitance is >=1uF and the cap's ESR is <=300milliohms.
>
Thanks for the hint. However, quote from datasheet: "If the input supply
voltage drops too low for the LTC1844 to maintain regulation, the
internal feedback loop goes into dropout and the internal pass
transistor turns fully on. If the input supply then suddenly rises, the
output may briefly overshoot the intended output voltage ..."
Can be muffled with a big cap on the output, probably driving the
p-channel into its "ouch" region. I don't really like such behavior.
Nah, not my cup of tea.
I've also had my share of experiences with fast input voltage rise
times. A TPS regulator reacted to that with phssst ... BANG. TI couldn't
explain it, refused to give me the innards for SPICE, refused to run my
circuit on their SPICE. So I just chucked it.
> I suspect this is too spendy for most of your applications -- $1.45 in reel
> quantities from DigiKey -- but if you ever feel like trying our an LDO again,
> there's one candidate...
>
For a 6.5V/150mA max regulator that is quite expensive.
Joel Koltner
I just thought I'd mention...
Some purchasing guy was talking with me about various parts for some project
yesterday and I noticed that we're using Linear Tech's LTC1844 in numerous
different products and at least so far no one's reporting seeing one
misbehave. This is a 150mA LDO (90mV headroom required at 150mA, 30mV at
50mA). The data sheet has a chart that claims it's guaranteed stable so long
as the output capacitance is >=1uF and the cap's ESR is <=300milliohms.
I suspect this is too spendy for most of your applications -- $1.45 in reel
quantities from DigiKey -- but if you ever feel like trying our an LDO again,
there's one candidate...
---Joel
Joel Koltner
"Joerg" <notthis...@removethispacbell.net> wrote in message
news:Ao%sj.7280$5K1....@newssvr12.news.prodigy.net...
> ...If the input supply then suddenly rises, the output may briefly overshoot
> the intended output voltage ..."
Thanks for pointing that out. Wouldn't a lot of regulators (LDO and non-LDO
alike) do this, though? Once you saturate your pass device, if it's slower to
respond than the feedback loop, you're sunk, aren't you? (Seems like to
op-amps where, once you hit the rails, pretty much all bets are off on how it
behaves... although things like the outputs reversing polarity like some old
op-amps did is a bit much...)
> For a 6.5V/150mA max regulator that is quite expensive.
You're paying for the "very" LDO in VLDO there, I think.
---Joel
Joerg
> Hi Joerg,
>
> "Joerg" <notthis...@removethispacbell.net> wrote in message
> news:Ao%sj.7280$5K1....@newssvr12.news.prodigy.net...
>> ...If the input supply then suddenly rises, the output may briefly overshoot
>> the intended output voltage ..."
>
> Thanks for pointing that out. Wouldn't a lot of regulators (LDO and non-LDO
> alike) do this, though? Once you saturate your pass device, if it's slower to
> respond than the feedback loop, you're sunk, aren't you? (Seems like to
> op-amps where, once you hit the rails, pretty much all bets are off on how it
> behaves... although things like the outputs reversing polarity like some old
> op-amps did is a bit much...)
>
I have never seen a LM317 do that. Usually when they approach dropout
LDOs steer their pass device into full conduction and naturally it'll
take a while to swing back when Vin jumps back up. OTOH non-LDOs are
different, they typically run out of base drive and that happens kind of
gracefully. Their pass device doesn't really leave the linear range, at
least not by much.
It's a bit like jet engines versus regular gas engines. When you pull
the throttle on a jet engine it takes a while until it reduces thrust
while a (non-turbocharged) gas engine reacts immediately.
Take a look at the right-most part of the schematic for the LM317, about
in the middle of this datasheet:
http://www.national.com/ds/LM/LM117.pdf
People have done rather strange things with the LM317. One example is in
the above datasheet on page 18 where they built a switcher with it. I've
used them as AM modulators and such, usually producing some forehead
wrinkles and coughing in design reviews.
>> For a 6.5V/150mA max regulator that is quite expensive.
>
> You're paying for the "very" LDO in VLDO there, I think.
>
When I am that tight on voltage overhead I usually build a switcher.
Joel Koltner
> I have never seen a LM317 do that. Usually when they approach dropout LDOs
> steer their pass device into full conduction and naturally it'll take a
> while to swing back when Vin jumps back up. OTOH non-LDOs are different,
> they typically run out of base drive and that happens kind of gracefully.
> Their pass device doesn't really leave the linear range, at least not by
> much.
Ah, gotcha.
Do I get a 'pass' in that the class I took in college was called "CMOS IC
Analog Design" and therefore I don't have to know what a BJT is?
Just kidding. :-)
> People have done rather strange things with the LM317. One example is in the
> above datasheet on page 18 where they built a switcher with it. I've used
> them as AM modulators and such, usually producing some forehead wrinkles and
> coughing in design reviews.
Nice. Very clever.
---Joel
Joerg
> "Joerg" <notthis...@removethispacbell.net> wrote in message
> news:9M0tj.7335$R84....@newssvr25.news.prodigy.net...
>> I have never seen a LM317 do that. Usually when they approach dropout LDOs
>> steer their pass device into full conduction and naturally it'll take a
>> while to swing back when Vin jumps back up. OTOH non-LDOs are different,
>> they typically run out of base drive and that happens kind of gracefully.
>> Their pass device doesn't really leave the linear range, at least not by
>> much.
>
> Ah, gotcha.
>
> Do I get a 'pass' in that the class I took in college was called "CMOS IC
> Analog Design" and therefore I don't have to know what a BJT is?
>
> Just kidding. :-)
>
In our area BJT means BJ's "Tatonka Stout", from these guys:
http://bjsbrewhouse.com/home_page/home_main.html
>> People have done rather strange things with the LM317. One example is in the
>> above datasheet on page 18 where they built a switcher with it. I've used
>> them as AM modulators and such, usually producing some forehead wrinkles and
>> coughing in design reviews.
>
> Nice. Very clever.
>
> ---Joel
>
>
Jamie Morken
> namde...@gmail.com wrote:
>> On Feb 10, 7:43 pm, Joerg <notthisjoerg...@removethispacbell.net>
>> wrote:
>>> namdegue...@gmail.com wrote:
>>>> All,
>>>> I am designing amulti-channelboard and I have individual linear
>>>> regulators for each channel running off a common switcher.
>>>> Now, my concern is that due to the split power planes, which are like
>>>> thin fingers running through the board, I will cause potential
>>>> differences based on the current in different channels, especially,
>>>> because I have a common ground plane. Now, I was thinking my choices
>>>> are:
>>>> 1) Split the ground plane as well, so I would have thin finger like
>>>> power and ground plane separated by narrow notch but connected at the
>>>> linear regulator end and at the input end. (All the channels are w.r.t
>>>> common ground plane.)
>>> designs it can work but only until a strong RF field shows up, upon
>>> which all hell breaks loose.
>>>
>>> OTOH if nobody split planes anymore I'd have less work so it does have
>>> some upsides ;-)
>>>
>>>> runnning across my planes (though separated by 2 layers) and even if I
>>>> isolate the grounds, some noise is bound to take the signal lines.
>>>
>>>> between the power plane 'fingers' in between them on the same layer as
>>>> the power planes.
>>> through here and there. And no thermal reliefs for those vias unless
>>> something needs to be soldered in there.
>>>
>>>> intervals.
>>> little". I would not split the ground at all.
>>>
>>>> all analog..there are digital lines and a digital section, but they
>>>> come into the analog area through an isolator.
>>> Use shielded versions for the magnetics if it's close by. Design your
>>> stuff with reasonable PSRR or bypass extra good where you can't, like
>>> for transistor stages.
>>>
>>> --
>>> Regards, Joerg
>>>
>>>
>>> - Show quoted text -
>>
>> that they are connected at the same potential at all the linear
>> regulators and at the input. But with my control lines, crossing
>> across my parts, my split is not going to have much effect. I could
>> put de-coupling caps across the notch to allow high frequencies to
>> pass, but again they might go through my control lines and not bother
>> with the caps...basically it is big ?..so yeah, I think I am not going
>> to split the ground...best to do it when I have free time as a test
>> project.
>>
>> For my thin ground slice in between the power planes (on the same
>> layer0, I put vias every 50 mils or so and no themal reliefs and it is
>> a relatively unpopulated location, so that will help. My switcher and
>> linear regulator are LT parts 3430/1763, but I have been told by just
>> about every amplifier part vendor (and LT themselves!) that linear
>> regulators don't attenuate frequencies above 100 kHz very well, so I
>> might keep a 100 kHz switching frequency and maybe put ferrites at the
>> linear regulator input.
>>
>
> Easy fix if you need good rejection: Use a npn-Darlington with enough
> heat dissipation capability to drop around 1.5V at your maximum current.
> Collector to switcher output, emitter to linear regulator input, 1K
> resistor from base to collector, 47uF cap from base to ground. Plus a
> couple discharge diodes. That'll behave like having a shoe size 15
> electrolytic in there.
Hi,
Is this npn-darlington solution good at rejecting HF better than a
linear regulator like the LM317? Also where do the discharge diodes go
in this circuit?
cheers,
Jamie
>
> Also, consider using the LM317. With a cap at the adjust pin it'll
> maintain a low output impedance all the way up to a MHz:
> http://www.national.com/ds/LM/LM117.pdf
>
> Personally I'd resist the temptation to use an LDO. I've seen them bring
> a lot of grief.
>
>
>> I am also considering using ceramic 10 uF caps instead of the tantalum
>> datasheets reference. Also some 0508/0402 caps.
>
>
> Tantalums in bypass applications are quite evil IMHO. Seen too many
> orange-greenish clouds wafting up after the explosion and so on.
>
Joerg
Yes, for smaller transistors and small currents. Mostly I don't even use
Darlington, just a transistor with a good guaranteed minimum beta. One
diode goes from emitter to collector (cathode to collector which would
be at the input rail), the other from base to collector (also cathode to
collector). The diodes make sure the capacitors don't zap the
transistor(s) in case someone drops the pliers and abruptly shorts out
the input rail.
Remember that this drops 700-800mV or so, or twice that in case of a
Darlington. And it is not too useful if the output of this stage must be
very stable. Of course, then you could run a secondary loop back to the
switcher but that becomes esoteric and is not for the faint of heart.
[...]
Jamie Morken
I am making an isolated bipolar powersupply that needs to be quite
clean, for 12bit ADC/DAC's, here is what I have so far:
For the npn solution I put in a pnp on the negative rail, not sure if
this is correct, or if it should still be an npn?
Thanks for any comments on the circuit. I would like to get as clean
a supply as possible, maybe 0.5mV noise would be nice on the +-5V rails.
cheers,
Jamie
>
> [...]
>
Joerg
It's correct, pnp for the negative rail. But you need to flip both D117
and D118 around.
> Thanks for any comments on the circuit. I would like to get as clean
> a supply as possible, maybe 0.5mV noise would be nice on the +-5V rails.
>
I don't think you need those high PSRR regulators. The LM317 is actually
pretty good. Don't know what you are using for the negative rail though,
your schematic doesn't say. Also, you need ADJ pin dividers, else you'll
only get 1.2V out of it. Bypass that ADJ pin to GND with a 10uF MLCC cap
for even better noise muffling.
Add a 0.1uF each in parallel to C22 and C29 to make sure RF spikes won't
sail through. When picking a negative regulator watch out. Many of those
are LDOs and LDO usually means trouble. Make sure it's unconditionally
stable.
John Larkin
wrote:
I don't understand this. It looks like a pwm controlled forward
converter, but the secondary circuit (D110, L5, C29) looks strange.
There's usually another "catch" diode to keep the current circulating
in L5 when the pwm is off. L5 might ring like a banshee when the pwm
snaps off.
As far as 100 KHz noise is concerned, you have more normal-mode ripple
rejection than you need. The bigger problem might be common-mode noise
on the (floating?) common, coupled through the winding capacitances.
John
Jamie Morken
Thanks, I updated the circuit:
I used a LM337 negative voltage regulator on the negative rail. I saw
an AC regulator circuit in the LM317 datasheet but am not sure if you
can really use the LM317 for a negative regulator for this?
I left the second voltage regulators in the circuit in case 8V and 5V
are both needed :)
cheers,
Jamie
Jamie Morken
Thanks, I added those diodes.
>
> As far as 100 KHz noise is concerned, you have more normal-mode ripple
> rejection than you need. The bigger problem might be common-mode noise
> on the (floating?) common, coupled through the winding capacitances.
Would a common mode filter be the solution for this I guess? I've seen
some capacitors that can be used for this I think, they have 4
terminals. Also would a common mode choke from each rail to the
floating ground be enough to get rid of the common mode noise or are
there other techniques to use?
cheers,
Jamie
>
> John
>
>
John Larkin
wrote:
>> I don't understand this. It looks like a pwm controlled forward
>> converter, but the secondary circuit (D110, L5, C29) looks strange.
>> There's usually another "catch" diode to keep the current circulating
>> in L5 when the pwm is off. L5 might ring like a banshee when the pwm
>> snaps off.
>
>Thanks, I added those diodes.
>
>>
>> As far as 100 KHz noise is concerned, you have more normal-mode ripple
>> rejection than you need. The bigger problem might be common-mode noise
>> on the (floating?) common, coupled through the winding capacitances.
>
>Would a common mode filter be the solution for this I guess? I've seen
>some capacitors that can be used for this I think, they have 4
>terminals. Also would a common mode choke from each rail to the
>floating ground be enough to get rid of the common mode noise or are
>there other techniques to use?
>
>cheers,
>Jamie
Hi, Jamie,
I couldn't make suggestions without understanding the application. Is
the circuit that this thing powers floating? What does it do?
We recently did a 16-channel thermocouple input board. Each channel
has a dc/dc converter based on an ISDN line transformer, driven by
some small mosfets at 60 KHz. The tc input channel is fully floating,
with optoisolators for data i/o. The biggest noise problem is
common-mode 60 KHz coupled into the floating circuits through
transformer capacitance. It's only a problem in the sense that we
don't want to poke a lot of 60 KHz into the customer's thermocouple
leads. We added a 1 (or maybe 2?) nF cap from the floating ground to
real ground as a compromise... too big a cap has problems of its own.
John
Joerg
Not that way. I prefer the LM317 over the LM337 because the 337 has the
architecture of an LDO and is not as stable. But it would require
another transformer winding. Or maybe you could used the one at pins 9
and 10 with a bridge rectifier. Then you could build a separate +8V
supply and just tie the positive side to GND. But you should be able to
get the LM337 to work as well, it's a simpler circuit.
> I left the second voltage regulators in the circuit in case 8V and 5V
> are both needed :)
>
Ok. 3V ain't a lot of head room but for 100mA it's fine.
Joerg
Good point, I had only looked at the DC side. Also, the primary driver
side does not look right.
Jamie, the best way to prevent DC runaway of the core is to AC couple.
Then second best is to use bridge rectifiers (and then you could
actually use a LM317 for the neg rail as well).
Or at least connect the anode of D119 to xfmr pin 7, flip D120 around
and connect its cathode to xfmr pin 4.
> As far as 100 KHz noise is concerned, you have more normal-mode ripple
> rejection than you need. The bigger problem might be common-mode noise
> on the (floating?) common, coupled through the winding capacitances.
>
--
Regards, Joerg
John Larkin
<notthis...@removethispacbell.net> wrote:
>>>
>>> "http://rocketresearch.nekrom.com/new/isolated%20bipolar%20powersupply/isolated%20bipolar%20powersupply.jpg"
>>>
>>> For the npn solution I put in a pnp on the negative rail, not sure if
>>> this is correct, or if it should still be an npn?
>>>
>>> Thanks for any comments on the circuit. I would like to get as clean
>>> a supply as possible, maybe 0.5mV noise would be nice on the +-5V rails.
>>>
>>> cheers,
>>> Jamie
>>>
>>>
>>>
>>>
>>>> [...]
>>>>
>>
>> I don't understand this. It looks like a pwm controlled forward
>> converter, but the secondary circuit (D110, L5, C29) looks strange.
>> There's usually another "catch" diode to keep the current circulating
>> in L5 when the pwm is off. L5 might ring like a banshee when the pwm
>> snaps off.
>>
>
>Good point, I had only looked at the DC side. Also, the primary driver
>side does not look right.
>
>Jamie, the best way to prevent DC runaway of the core is to AC couple.
>Then second best is to use bridge rectifiers (and then you could
>actually use a LM317 for the neg rail as well).
That's actually a pretty slick configuration. When the fets turn off,
the flyback energy is dumped back into the power supply.
Looks like Q92 drives the gate of p-fet Q91, also slick. Interesting
things happen over the pwm range from 0 to 50%, but I think it's OK.
Why didn't I think of that?
John
Jamie Morken
Hi John,
Yes its a floating supply for powering ADC/DAC and opamps that are
interfaced to by opto from the microcontroller side.
The primary side is the isolated microcontroller supply, and the
secondary side is 120VAC generated by a truesine inverter.
Thanks I'll add the 2nF cap across grounds. Would a common mode
choke, maybe 3 coils in this bipolar supply case be effective as well?
What about these 4 terminal caps, they seem to be more effective than
a larger common mode choke.
http://www.johansondielectrics.com/x2y/
cheers,
Jamie
>
> John
>
>
>
Jamie Morken
Terry Given suggested that one in a thread earlier :)
cheers,
Jamie
>
> John
>
>
Joerg
If Terry gave it to you then it should work. But make sure to obtain
enough information regarding the core. A DC run-off situation or core
saturation is no fun.
Terry Given
high praise indeed.
>>
>>
>> Terry Given suggested that one in a thread earlier :)
>>
>
> If Terry gave it to you then it should work. But make sure to obtain
> enough information regarding the core. A DC run-off situation or core
> saturation is no fun.
>
Thanks Joerg!
I came up with that circuit in 1994 - I did a 2MHz modulated
smps/gatedriver circuit. initially I had a clamp zener, but I used the
same zener for all 6 gatedriver/smps, and although the leakage energy
was OK from one, 6 was too much. so I thunked a bit, and figured out the
self-driven PNP diagonal half-bridge. the switching speed of the
PNP/PFET is pretty much irrelevant too - as long as one device switches
fast its ok - at least until you muck about for so long that the next
switching cycle has been and gone.....
we built about 300,000 of these, with no problems. running from +24V,
using FMMT491A and FMMT591A SOT23 BJTs. We could probably have patented
it - though we did patent the toroid winding technique I came up with.
as far as the core is concerned, in theory all is well even at 50% duty
cycle, as Vclamp = Vcc + 2*Vd, but Von = Vcc - 2*Vce. ware remanent flux
though - that can bite quite a chunk out of the available flux
excursion, rendering the typical "hey, Bsat < 300mT, no worries"
approach somewhat fraught with peril
its really good for open-loop flybacks too, if you choose Vout =
Vin*Ns/Np, as load reductions just mean more energy spins round in
circles on the primary.
Cheers
Terry
John Larkin
wrote:
BJT's? With base resistors? It looks much nicer with fets.
>
>as far as the core is concerned, in theory all is well even at 50% duty
>cycle, as Vclamp = Vcc + 2*Vd, but Von = Vcc - 2*Vce. ware remanent flux
>though - that can bite quite a chunk out of the available flux
>excursion, rendering the typical "hey, Bsat < 300mT, no worries"
>approach somewhat fraught with peril
>
>its really good for open-loop flybacks too, if you choose Vout =
>Vin*Ns/Np, as load reductions just mean more energy spins round in
>circles on the primary.
>
>
>Cheers
>Terry
Nice circuit, a few parts doing a lot of stuff. The only warning I'd
give Jamie is that he may not want fast edges getting into his
isolated analog stuff, and it would be tricky to slow this one down.
Our thermocouple supply was a classic open-loop forward converter,
dual nfets driving a center-tapped transformer, and we shaped the gate
drives to soften things up. Efficiency didn't matter much here. And we
got a nice clean square wave out of the secondaries, which also
clocked the delta-sigma ADC!
John
Joerg
Hey, you re-invented the modern electro-cardiograph :-)
Except that we (usually) feed back the analog signals and do the AD
conversion on the system side.
Terry Given
they were cheaper. and the 6 npn base drives had to go a long way on a
single layer (no 0V plane) to get from the CPLD to the gatedrivers, so
Vbe being < Vt was a good thing.
>
>
>
>>as far as the core is concerned, in theory all is well even at 50% duty
>>cycle, as Vclamp = Vcc + 2*Vd, but Von = Vcc - 2*Vce. ware remanent flux
>>though - that can bite quite a chunk out of the available flux
>>excursion, rendering the typical "hey, Bsat < 300mT, no worries"
>>approach somewhat fraught with peril
>>
>>its really good for open-loop flybacks too, if you choose Vout =
>>Vin*Ns/Np, as load reductions just mean more energy spins round in
>>circles on the primary.
>>
>>
>>Cheers
>>Terry
>
>
> Nice circuit, a few parts doing a lot of stuff. The only warning I'd
> give Jamie is that he may not want fast edges getting into his
> isolated analog stuff, and it would be tricky to slow this one down.
> Our thermocouple supply was a classic open-loop forward converter,
> dual nfets driving a center-tapped transformer, and we shaped the gate
> drives to soften things up. Efficiency didn't matter much here. And we
> got a nice clean square wave out of the secondaries, which also
> clocked the delta-sigma ADC!
>
> John
Lovely!
Cheers
Terry
Jamie Morken
John Larkin wrote:
> Nice circuit, a few parts doing a lot of stuff. The only warning I'd
> give Jamie is that he may not want fast edges getting into his
> isolated analog stuff, and it would be tricky to slow this one down.
> Our thermocouple supply was a classic open-loop forward converter,
> dual nfets driving a center-tapped transformer, and we shaped the gate
> drives to soften things up. Efficiency didn't matter much here. And we
> got a nice clean square wave out of the secondaries, which also
> clocked the delta-sigma ADC!
If I disable the PWM to the transformer for a couple cycles during ADC
reading would that get rid of the fast edges and common mode noise
issues?
I fixed the backwards diodes/caps on the positive and negative rails
of the floating bipolar supply:
cheers,
Jamie
>
> John
>
>
Joerg
>
>
> John Larkin wrote:
>
>> Nice circuit, a few parts doing a lot of stuff. The only warning I'd
>> give Jamie is that he may not want fast edges getting into his
>> isolated analog stuff, and it would be tricky to slow this one down.
>> Our thermocouple supply was a classic open-loop forward converter,
>> dual nfets driving a center-tapped transformer, and we shaped the gate
>> drives to soften things up. Efficiency didn't matter much here. And we
>> got a nice clean square wave out of the secondaries, which also
>> clocked the delta-sigma ADC!
>
> If I disable the PWM to the transformer for a couple cycles during ADC
> reading would that get rid of the fast edges and common mode noise
> issues?
>
Yes. Add some big electrolytics in parallel to C22 and C29, the required
capacitance depending on how many cycles of quiet time you need and how
much each rail pulls. You may have to add current limiting to the input
where Q91/Q92 are. Else charging up the big electrolytics at power-up
could take a toll there.
But I wouldn't worry too much about noise if you do a good layout. On
ECG units we never stop the clock.
[...]
John Larkin
wrote:
>
>
>John Larkin wrote:
>
>> Nice circuit, a few parts doing a lot of stuff. The only warning I'd
>> give Jamie is that he may not want fast edges getting into his
>> isolated analog stuff, and it would be tricky to slow this one down.
>> Our thermocouple supply was a classic open-loop forward converter,
>> dual nfets driving a center-tapped transformer, and we shaped the gate
>> drives to soften things up. Efficiency didn't matter much here. And we
>> got a nice clean square wave out of the secondaries, which also
>> clocked the delta-sigma ADC!
>
>If I disable the PWM to the transformer for a couple cycles during ADC
>reading would that get rid of the fast edges and common mode noise
>issues?
Absolutely. It's good that you have the resources available to do
that.
>
>I fixed the backwards diodes/caps on the positive and negative rails
>of the floating bipolar supply:
>
>"http://rocketresearch.nekrom.com/new/isolated%20bipolar%20powersupply/isolated%20bipolar%20powersupply3.jpg"
Looks a little complex, especially if you can shut down the switching
during digitizing.
John
Jamie Morken
I guess it may make more sense to synchronize the switching with the ADC
sampling (200kHz ADC) since both of these interfaces come from the FPGA.
As long as the common mode/switching noise doesn't last more than a
microsecond or so around each switch point. I wonder if a common mode
choke would exacerbate this problem, may be better to use no common
mode filtering if the switching is synchronized?
cheers,
Jamie
>
> John
>
>
Terry Given
look up symmetric PWM - rather than a sawtooth, it uses an equilateral
triangle. this is done all the time in 3-phase PWM (AKA Space Vector
Modulation). The peak & trough of the triangle waveform *always*
correspond to the exact center of the high & low pulses. Syncing the ADC
S/H thusly means the sample is always as far as possible from the
switching edges.
and when trying to make, say, sinusoids (its also true for DC) you will
notice that the centre of the pulse is equal to the desired output
current - so sampling like this automagically removes the switching
ripple artefacts.
Cheers
Terry