Which Spectrum Analyzer Measures Very Low Frequencies?

[Ice-Tea]

Just got a Marconi 2383 to work. Supposed to be between 100Hz and 4.2GHz. With trackking gen 

[billfernandez]

If 1.6Hz to 90KHz is an acceptable range, look at the Quant Asylum QA401:  https://quantasylum.com/collections/frontpage/products/qa401-audio-analyzer

[precaud]

You don't really need a DSA for this task. DSA's have two discrete measurement channels phase-locked to the generator for transfer function analysis/FRA. With their added complexity, these units tend to be large, heavy, and loud (fans).

Any of the single-channel FFT analyzers from the 80's on that have selectable DC-coupled input will do the trick.

Ono Sokki made some reasonably-sized ones and were well-built. Size-wise the HP 3561A isn't too bad. There were some smaller units from Nicolet, Spectral Dynamics, AND aka A&D aka Zonic, and others, often marketed as "mini-FFT", "personal-FFT", and the like.

[radiogeek381]

Just saw this thread.

If the problem is low-level signals that are below (> 40dB) some other wanted signal in-band, you probably want a real-live spectrum analyzer.  DSO's and FFT digitizers might do the trick, but the former is going to have limited dynamic range and soundcard versions of the latter might be just the trick, or burdened with spurious responses and noise sources of their own.

Several posters have suggested various choices (R&S, and such) My choice would be an HP 3580A, principally because there's one on the bench.  It doesn't get you down to mHz, but the specs cover 5 Hz to 50kHz, resolution bandwidths down to 1Hz. Dynamic range is a carp shoot, but the advertised "sensitivity" (whatever that means) is 30nV. 

I wouldn't pay a lot for one (for some reason there are a bunch on eBay for wayyyy too much money) I got mine > 10 years ago at a swap meet for about $100 (US).

[bson]

For low frequencies like < 10Hz I use my 34465A DMM with DIGI+MEM options and capture to a USB stick. Then decimate and FFT the capture.  (A basic boxcar filter works fine for decimation prior to the FFT to get meaningful bin sizes.)  Because of the high resolution capture it doesn't need a particularly long data set.

[innkeeper]

Tiny, what did you end up doing?

[TinyMirrors]

Hey! Just a quick update. I ended up borrowing a LabOne Spectrum Analyzer made by Zurich Instruments from another company. I ended up finding out all sorts of neat things about my circuits that I didn't realize. For instance, the cutoff frequency for a basic rc filter changes when biasing at different voltages such as 5V vs 100V. I didn't realize the ceramic capacitors could change so much. I feel like I'm learning a lot more than I expected.

Thanks for the great advice guys! I'm still watching ebay for an HP3562. Hopefully budge and available line up just right one day.

[cdev]

Have you seen the free "Spectrum Lab" software?

https://www.qsl.net/dl4yhf/spectra1.html#download

The reason I ask is because many audio cards would qualify as capture devices at low frequencies, especially if coupling capacitors were removed or increased in value.

It also runs under Wine on a modern Linux OS (just verified)

https://www.qsl.net/dl4yhf/spectra1.html

Another similar program is Baudline.