It's always fun to see analog and RF electronics. It brings back memories of college. For dB's our electronics teacher told us to remember 4 things and we could figure out everything from there. 0dBm = 1mW -3dB = x0.5 0dB = x1 3dB = x2 But I still feel like the sorcerer's apprentice trying to cast the incantations :-)
For dB-anything it may be more important to remember that 0 dB is the reference level for "anythings". dBm is decibels "milliwatt" so 0 dBm is the reference level of 1 milliwatt. Another common one is dBA for audio noise level, where 0 dBA is the threshold of human hearing. For any measure, decibels do apply the same as your other rules: every 3 dB up is a doubling, 3 dB down is a halving. This *should* be really important for car audio enthusiasts who might think 133 dBA is "only" 3 more than 130 dBA. Yeah, bro... it's "only" twice as loud ;)
When measuring the 60dB pad, maybe check under "Amplitude" whether the attenuator of the analyzer itself is enabled. On my Rigol, it usually is by default, unless I set the reference level to something like -50dB or so. If your SA has a preamplifier, then of course enabling that helps as well (and on mine at least also disables the attenuator, wouldn't make much sense otherwise), but that's only if it has it.
For RF power meters, if you don't have a wide variety of expensive sensors, you want alot of different attenuators to stay under the max power allowed. So measuring your attens is nice because if you got a 20 db pad and it is 20.1 or .2 at your freq, you often have to stack them and the errors accumulate. OF course HP power meters often have a reference source but it is fixed freq.
Greetings... Noticed the delta of the attenuator at the higher frequency sweep even with devices specd for very high frequency. Curious, have you ever used, like store trace at the noise floor etc, with and without tracking generator on / off and see how much, if any leakage occurs between the input and ouput ports ? It's possible that phase leakage could be causing the slight deviation of the attenation measurement.. thanks
this is not the best way to measure. yes there will be other factors that try to be calibrated out. the best would be a true open short load cal. but even a good VNA will not be as accurate as a good scalar network analyzer setup. the best is fixed frequency and a power meter
@IMSAI Guy can you please share what kind of coaxial cables are you using for your spectrum analyzer test setup and measurements? What kind of quality coaxial cables to you recommend for SA measurements in D.C. to 3.5GHz range for example? I just purchased the SSA 3021X and looking forward to acquire high quality coaxial for measurements. Thank you in advance! 73 de Andrei, YO6TJJ
That is difficult as good cables can cost as much as the SA. I would not worry much at 3.5GHz. I have used cables made by GORE and Pasternack and Huber+Suhner. these new would be in the $300 to $1000 range. and used I got them for 1/10th the price. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-cFx3DtX8Eao.htmlsi=wt46HLID0PyfgV_9
The wiggly waveform of the attenuators on the Siglent spectrum analyzer appeared all the same. Is it possible that it was caused by the instrument itself?
I have seen many spectrum analyzer videos talking about normalizing the tracking generator trace and using the "Store Ref" button. Some people say to press this button and others say don't press the button, even on the exact same spectrum analyzer. There seems to be a lot of confusion about this. I have a Rigol DSA815-TG Spectrum Analyzer.
It is not the same hardware. The 3021 and 3032 *are* the same hardware as each other, but the 3015 is not so it cannot be modified to go to a higher max frequency like the 3021 can.
the SA is a much cheaper device so most labs did not have a vna (pre nanovna). the SA is also more rugged and maybe better suited for measuring things like amplifiers.
@@IMSAIGuy Oh yes, your communications analyzer comes to mind for wattage capability. Perhaps you could dissect and analyze some faulty attenuators. I'm always curious about the $$$ divide between the components we can get.
@@jspencerg Newer model, the SSA3021X Plus spectrum analyzer can be easily hacked to SVA1032 (SA + VNA), more information can be found on the EEVblog forum. The SSA3021X Plus in only $255 more expensive than the SSA3021X.
@@jeremycyclist6269 The SSA3015 can use the same hack to become an SVA1015. I've done the first step, opening up the software features, but haven't had the courage to swap to the SVA firmware. Also one of IMSAI Guy's recent videos, this one or the previous SSA one, mentioned the SVA having an RF bridge built in... I don't recall seeing that difference mentioned in the EEVblog forums. Kind of important since that means a piece of hardware is missing in one's "hacked" SSA -> SVA
A 6 dB "pad" is used on signal generators outputs when checking radio receivers that are supposed to be nominally 50 ohm inputs, I think it has to do with isolation and also maintaining an actual 50 ohms to the generator. Then you add subtract 6 from the reading.
