Can a Network Analyzer be used in place of a signal generator
EEUndergrad
I am working on research(as an undergrad), and my project needs an
RF signal generator capable of a constant output of 2.4 GHz at about
+5 dBm, and I was wondering if there was any way that a network
analyzer can be used to output one frequency (2.4 GHz) instead of
doing sweeps? and if so would the signal be much weaker than I need?
I'm an EE working under an ME, and have used a network analyzer once
in my time at school, so my knowledge of how it works is limited, and
not many people around me know how this stuff works. Would hooking it
up to be used as a signal generator that I amplify and transmit the
signal wirelessly put the network analyzer in any risk of getting
damaged?
for reference, I'm thinking of using the network analyzer for the 2.4
GHz signal generator because we already have one, while it would be
around $3,000 for a used/questionable signal generator or $11,000 for
a new one, which eats a LOT of the budget.
Thanks for any help you can give me.
Chuck Olson
"EEUndergrad" <benjami...@gmail.com> wrote in message
news:d840c63a-26a6-4ad7...@q30g2000yqb.googlegroups.com...
I checked the HP and Tektronics gear I have on my workbench and found If it
has a Sweep Mode selection called Manual Sweep, it can put out a single
frequency with all the controllability and accuracy of the product
specification. Another possible choice would be if it has Span selection
where you can select Zero Span, and again it will put out a single
frequency. Plus five dBm should be within the range of most network
analyzers.
Chuck
EEUndergrad
tm
> Great, thank you for your help
Ebay.
HP 8664A to 3 GHz for $1700 and a day left.
Look up the specs on the Agilent.com site. Should do exactly what you want.
mi...@sushi.com
> Great, thank you for your help
Because network analyzers have filters on the input, sometimes they
don't put out the cleanest signals, where a signal generator might
have a cleaner signal. [Usually only an issue if testing high speed
ADCs.] Otherwise, I don't see a problem, other than it ties up a nice
piece of test gear. ;-)
I have a 8660c. A boat anchor, but a nice boat anchor. I had to do a
lot of lets make a deal to get it working, so I'd never suggest buying
one on ebay.
John Walliker
200616317161
or
150614694236
or
200616317159
You would need a 1A variable voltage bench power supply to drive the
tuning coil and a low-current psu to drive the oscillator. Frequency
stability will not be anywhere near as good as the network analyzer
unless you use a phase locked loop drive system which may be over-
complicated for your project. (Suitable pll modules such as
160589712611 appear regularly on eBay.) Google will find you
instructions for using YIG oscillators.
John
Michael A. Terrell
The brand & model number?
--
It's easy to think outside the box, when you have a cutting torch.
John Larkin
Probably so. Read the manual!
What's the 2.4 GHz going to do?
John
EEUndergrad
I will post the brand and model # when I find out, I have yet to see
this device, but I'm told we have it.
Stupid question time: Can the network analyzer be used to create a
2.4GHz signal on the output(like my previous question), and then, at
the same time, read another signal at the input, and produce a
frequency spectrum plot?
I feel like a total electrical noob with all of this stuff.
John Larkin
That's precisely what a network analyzer does!
>
>I feel like a total electrical noob with all of this stuff.
Well, it used to be that one person could understand most of
electronics. No more; it's just too complex. Nowadays it takes a bunch
of years to get exposed to things, and even then it's easy to get
stuck in a niche. I've been fortunate to always work for companies
with terrible marketing departments, so got to work on all sorts of
unorganized stuff.
I'd recommend you get the manual for the analyzer and read it. There's
probably some tutorial stuff.
Joerg
On some analyzers it is called "zero span". The screen still runs but it
won't change the frequency. This is how I once tuned a piano (no joke).
But be aware that some analyzers briefly interrupt the "transmission"
when the sweep restarts at the left of the screen. So you'll see a
modulation. You could turn the sweep time to the maximum possible to at
least reduce the number of those events.
--
Regards, Joerg
http://www.analogconsultants.com/
"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
John Devereux
I don't think so, the output signal is usually synchronous with the
sweep isn't it?
[...]
>
> What's the 2.4 GHz going to do?
>
> John
>
--
John Devereux
John Larkin
There's an oscillator or synthesizer that's swept, and a tracking
scalar or vector voltmeter. Usually you can set the sweep width to
very narrow or no sweep at all.
John
EEUndergrad
the 2.4 GHz is going to be used to encode another signal and transmit
it.
this may be a stupid question, but is there any way to convert a 15
MHz signal into a 2.45 GHz, or even a 868 MHz signal, using frequency
multipliers or anything?
I'm running into a lot of resistance getting any good equipment, my
adviser is pushing me to try and find a way of making a basic signal
generator (15 MHz max frequency) work for this project. would it be
possible to use a lot of frequency mixers to step up the frequency up
a lot? like use 8 frequency mixers starting with a signal generating
9.57 MHz, going int both the IF and LO inputs, and then the output
being 19.14 MHz going into another frequency mixers IF and LO inputs
and outputting 38.28125 MHz, and so on until you get 2.45GHz? It seems
to me like that wouldn't work due to the difference in the resistances
causing reflections and the output would need to be amplified quite a
bit if it wasn't distorted beyond recognition.
I'm still new to anything above 1 MHz, but I feel like i'm learning a
lot.
Let me know what you think,
Thanks
Joel Koltner
>this may be a stupid question, but is there any way to convert a 15
>MHz signal into a 2.45 GHz, or even a 868 MHz signal, using frequency
>multipliers or anything?
You most likely want a mixer: You feed your 15MHz signal along with a 2.3GHz
local oscillator (this can come from your VNA behaving as a source) into it,
and out pops your signal at 2.45GHz... as well as 2.15GHz, which may or may
not bother you (i.e., mixers produce the sum and difference of the frequence
at their ports... they actually produce other frequencies as well -- both
leakage of the LO frequency and your 15MHz, as well as various "mixer spurs,"
but for just testing something out at lower powers you can often ignore
them... but beware if you're going to be using a power amplifier and putting
this on-air).
Note that you could have also used 2.6GHz as the LO, in which case you'd have
gotten 2.45GHz and 2.75GHz out.
If the undesired output does bother you, it's better to start at 868MHz and
use an LO of either 1.582GHz or 3.318GHz and you'll end up with 2.45GHz as
well as a (potentially unwanted) signal at 714MHz or 4.186GHz -- those are
quite easy to filter out, if necessary.
A page like this one:
http://www.mini-circuits.com/products/fm_coax_level_7.html ... has lots of
mixers that would be appropriate. "Level 7" means that you're supposed to
feed the LO port with 7dBm (500mV RMS in a 50ohm system); it's probably the
most popular level for passive mixers out there (you can be off a bit and the
performance won't change that much).
> like use 8 frequency mixers starting with a signal generating
> 9.57 MHz, going int both the IF and LO inputs, and then the output
>being 19.14 MHz going into another frequency mixers IF and LO inputs
>and outputting 38.28125 MHz, and so on until you get 2.45GHz?
See above. I think you're confusing mixers with frequency multipliers (which
is really just a mixer with its "IF" and "LO" ports connected together). If
you've using a signal that doesn't rely on an amplitude (envelope) modulation,
you can use a bunch of multipliers as you've outlined. However, if you do
need to preserve amplitude information, it'll change going through each
multiplier, and -- as you allude to -- while in theory you can figure out how
to "pre-disort" the input amplitude to obtain the desired output, in practice
this is a non-trivial problem that, given your current background, I wouldn't
recommend tackling.
BTW, it sounds as those your adviser probably doesn't have much real-world
experience in RF design. If that's the case, I suggest you see if you can
find some local company that could provide you with someone who does have that
sort of experience and is willing to provide it as a form of public service to
the school (and to gain some recognition on your reports); this will
dramatically increase the chance of your project actually working, IMO. :-)
All the better if said company is able to loan you the sort of equipment
you're after! (Bigger places like a LeCroy or Agilent or Tektronix can pretty
much *always* find spare network analyzers, frequency sources, spectrum
analyzers, etc. to loan out for a project like this.)
---Joel
Joel Koltner
between frequency multipliers and mixers in terms of their function and
construction; please ignore my comment that you might not.
RST Engineering
<benjami...@gmail.com> wrote:
>this may be a stupid question, but is there any way to convert a 15
>MHz signal into a 2.45 GHz, or even a 868 MHz signal, using frequency
>multipliers or anything?
Lots of questions. How precisely does the 2.45 GHz signal have to be
in frequency tolerance?
To answer your question about frequency multipliers, remember that
they have to be done in integer multiples. If you require precisely
2.45 GHz., the 163rd harmonic of 15 MHz. is 2.445 GHz. and the 164th
is 2.46 GHz.
What the hell does 868 MHz. have to do with it?
Google "step recovery diode". HP had a whale of a lot of good
multiplier design information back in the late 60s and early 70s.
>I'm still new to anything above 1 MHz, but I feel like i'm learning a
>lot.
Understood. I put myself through undergrad school doing VHF. J-band
transceivers, and C-band radar repair for the airlines, so I had a
hell of a head start coming out of college. Lots of my friends had
zero experience on anything above a couple of MHz. coming out of
school.
Just remember that the transition from lumped constant (coils and
capacitors) starts to blur somewhere around a GHz. Below that you can
probably wind coils that will work and above that mostly sections of
transmission line.
Jim
EEUndergrad
> On Tue, 14 Jun 2011 14:10:47 -0700 (PDT), EEUndergrad
>
Thanks for the advice, and the 868 MHz was referring to another option
for my antenna, as opposed to the 2.4 GHz antenna I was planning on
using.
lang...@fonz.dk
what are you really trying to do?
> this may be a stupid question, but is there any way to convert a 15
> MHz signal into a 2.45 GHz, or even a 868 MHz signal, using frequency
> multipliers or anything?
> I'm running into a lot of resistance getting any good equipment, my
> adviser is pushing me to try and find a way of making a basic signal
> generator (15 MHz max frequency) work for this project. would it be
> possible to use a lot of frequency mixers to step up the frequency up
> a lot? like use 8 frequency mixers starting with a signal generating
> 9.57 MHz, going int both the IF and LO inputs, and then the output
> being 19.14 MHz going into another frequency mixers IF and LO inputs
> and outputting 38.28125 MHz, and so on until you get 2.45GHz? It seems
> to me like that wouldn't work due to the difference in the resistances
> causing reflections and the output would need to be amplified quite a
> bit if it wasn't distorted beyond recognition.
>
> I'm still new to anything above 1 MHz, but I feel like i'm learning a
> lot.
>
> Let me know what you think,
> Thanks
get an evaluation board like this: http://www.analog.com/static/imported-files/user_guides/UG-095.pdf
?
-Lasse
Joerg
It can't plot a spectrum because the receiver is essentially just a
glorified scalar RF voltmeter. It will only plot a frequency response
through the device under test but it won't even do that with the output
stuck on 2.45GHz.
It sound to me as if the OP wants it to display some sort of modulated
spectrum coming off of the other side, and that it won't do.
[...]
George Herold
wrote:
> "EEUndergrad" <benjamin.mo...@gmail.com> wrote in message
>
> news:67ac4582-5bbe-466a...@w10g2000yqh.googlegroups.com...
>
> >this may be a stupid question, but is there any way to convert a 15
> >MHz signal into a 2.45 GHz, or even a 868 MHz signal, using frequency
> >multipliers or anything?
>
> You most likely want a mixer: You feed your 15MHz signal along with a 2.3GHz
> local oscillator (this can come from your VNA behaving as a source) into it,
> and out pops your signal at 2.45GHz... as well as 2.15GHz, which may or may
> not bother you (i.e., mixers produce the sum and difference of the frequence
> at their ports... they actually produce other frequencies as well -- both
> leakage of the LO frequency and your 15MHz, as well as various "mixer spurs,"
> but for just testing something out at lower powers you can often ignore
> them... but beware if you're going to be using a power amplifier and putting
> this on-air).
>
> Note that you could have also used 2.6GHz as the LO, in which case you'd have
> gotten 2.45GHz and 2.75GHz out.
>
> If the undesired output does bother you, it's better to start at 868MHz and
> use an LO of either 1.582GHz or 3.318GHz and you'll end up with 2.45GHz as
> well as a (potentially unwanted) signal at 714MHz or 4.186GHz -- those are
> quite easy to filter out, if necessary.
>
That's great advice. One of those ?YIG? sphere oscillators might work
as the 2.3 GHz LO.
George
mi...@sushi.com
I'm wondering if a simple 2.4GHz video modulator can be the guts of
this project. You can buy them from Supercircuits for example. One
hack might be to just use the local oscillator and then slap on a mini-
circuits modulator. Or you just dig up the circu9itry for such a part
from the FCC product website. All you need is the FCC ID number.
Some random link:
http://www.active-robots.com/products/radio-solutions/av-modules.shtml
If you are going to test something, you use test equipment. If you are
going to design something, it is better to use components. It doesn't
have to be chip/transistor level. Off the shelf modules are fine to
get something off the ground. I buy stuff from minicircuits that I
know I could build, but don't feel like reinventing the wheel.