Any advice on low cost Spectrum Analyser for EMC testing 100K-4GHz

[Old_but_Alive]

do any members have some adivice on this topic ?

I am looking for a cheap pre-compliance setup for a product I am developing

maybe 4GHz is too much, 2GHz might be OK.

I was thinking of an SDR or HackRF type cheap solution

[Xwing]

It really depends on your actual needs, but a decent spectrum analyzer of more than 2 GHz costs at least € 2000, e.g. Rigol DSA832E.

[Old_but_Alive]

the Rigol DSA832E looks nice, but expenive compared to £57 HackRF

a HackRF does 1MHz to 6GHz, with an upconverter, it would go down to 100KHz.

I was hoping for some good cheap help

[vern]

a HackRF is ok if you want to check frequencies for hobby purposes, but not for pre-compliance testing. Sensitivity, bandwidth and step size for scanning are not practical for this purpose. HackRF also generates a lot of EMI by itself, when you look a a signal you always see some ghost harmonics somewhere.
Even with a spectrum analyzer like the Rigol it is really tricky. You see a spectrum and you don't know what you see because you have no idea about your environment (how noisy is it?) and what your antenna and the setup does with the values you get. If you have a real peak somewhere you will see it, but mostly it's just some dB's above the limit and right between the steps of your analyzer.
It is also really hard to catch intermittent signals, a professional device will catch them easily.
We have our devices pre tested in a lab and can then concentrate on the faulty frequencies in the resulting diagrams, which also enables us to calibrate the values we see with our spectrum analyzer.

If you can see problems with your device with a HackRF then you are in real trouble.

[dennis78]

HackRF and similar devices can be used for extreme simple monitoring (compare data without and with DUT) but without big experience results are usual useless. You can use it maybe for basic detecting problematic peaks and nothing more.

If you measure radiation effect big problem can be because FM and similar stations which saturate HackRF input and reduce sensitivity.

Another big problem is calibration equipment to satisfy minimal requirements.

Bad thing is that results can produce wrong conclusions.

[PM3295]

You can look at https://signalhound.com/products/usb-sa44b/ as well.

[enzine]

I do not want to demoralize anyone but I have made the CE marking for about 20 years and I have had the opportunity to do it with instruments and equipment at the highest levels.

For precompliance I used ESS Rohde & Schwarz EMI Receiver in my lab with specific and calibrated antennas.

Formal compliance was carried out in certified anechoic chambers external to the company
which often used spectrum analyzers and not EMI receivers.

It happened several times to have heated discussions with the certifier who challenged me emission diagrams out of bounds, even by a lot, which I did not find in precompliance.
Almost always the fault was the input stage of the spectrum analyzer that saturated due to strong out-of-band signal (for example below 100kHz, my field of application was power converters)

What I mean?

Even with a decent spectrum analyzer, you also need one or more calibrated antennas to correct the emissions detected by the instrument.

My advice?

If the equipment under test justifies the costs, turn to specialized laboratories for good precompliance and implementation of remedies for mitigation of emissions.

Good luck!

[w41k0v3r]

Hello,

It is not low cost, but for your information, your can buy a spectrum analyser from Siglent ( the SSA3021X ) and convert it to a SVA1032X.
You can see all details on the eevblog forum.
At this price it's unbeatable