cheapish spectrum analyzer

Jitter on the sampling clock and the digitizer.

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

Yeah run with no phase...(or set to one phase.) I guess you need access to the AC output as you say. Way back when, the EGG 124A had an AC voltmeter setting. (Single point noise density.) Do modern lockins have that feature?

George H.

Once the signal is shifted down to, say, a 10 KHz window, the digitizing could be done to 16 or 18 bits, with picosecond jitter. That would surely result in milliHz spectral resolution.

The remaining problem would be dynamic range limits from images. A swept analyzer with a real 10 Hz crystal bandpass filter would be better than a wider-band system with an FFT back end, but with an enormous speed and cost penelty.

Yigs are tuned with magnetic fields. It's amazing how good that can be made to work.

--

John Larkin         Highland Technology, Inc 
picosecond timing   laser drivers and controllers 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

But the LO has to have super low jitter for that to work, which is where the YTO comes in.

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

Their minimum resolution bandwidths are generally fairly wide, AFAIK--not like the 1-Hz of a 70000-series HP, or the 10 Hz of the

8566B. It's nice not to blow up the first mixer if you put a few volts on it by mistake.

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

Lock-ins work fine measuring noise. Each phase measures half the noise power. A lock-in really acts just like a narrow filter and downconverter.

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

Audacity

will do FFTs; the lower limit is probably around 20 Hz and the upper limit (with most sound cards) is 22 KHz. It can't do them "live" like a real spectrum analyzer, but the price is right.

Matt Roberds

We mod some CM501 chipset 20$ USB sound cards for both DC coupled input and DC coupled output. Then use them as 7 channel data recorders. Cost 20$.

So that could cover your low end, as long as your software does not have a low end highpass characteristic.

Steve

I think you might mean the C-Media CM6501 chip. C-Media no longer lists this chip on their web pile:

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

Right, AC voltmeter after the low pass gives the noise. I've got a lockin that uses the AD630, switched gain. I think that down converts noise near the odd harmonics as well as the fundamental. (But with a factor that goes as 1/n.)

George H.

If it uses square wave switching with no preselection, then that's right.

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

9 kHz is where conducted immunity testing *starts*.

The Rhode FSUs in "our" lab go down to 20 Hz (and up to 26.5 or 40 GHz). L ow frequency performance is often not very good in modern SpecAns. Things like linearity and noise figure are often not comparable to the performance in RF & uWave bands. LO feedthough is likely poor at low frequencies, eve n if you don't see it on the screen.

If one is doing serious audio work, then I would say RF SpecAns are not a g reat choice unless something special was done to service the band. (That i s a cost adder.) High performance audio work and a box that demodulates co mplex modulation up to 6, 8, 14, 18, 26, or 40 GHz are not usually co-joine d needs.

I mean, you can probably do almost better for audio with a "cheap" 24-it si gma-delta audio card, plus some good amps & attenuators, than your $100k RF /uWave VSA.

So that is 30 dB better spur performance or would that be 60 dB? I'm curious about what limits the performance of the SDR type units. Would it be the NCO/DDS circuit? I took a hard look at that some time back and found that it should be feasible to easily generate a sine wave digitally with ball park 120 dB close in spur performance not counting additional artifacts generated by a DAC.

--

Rick

Depends on how bad the SDR is. ;)

My 8566B is about -80 dBc at 100 Hz from a 5-GHz carrier, -90 at 1 kHz,

-120 at 100 kHz.

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