Siglent SDG1032 Amplitude not constant when sweeping

[Aldo22]

What is this board??

That doesn't matter. I just wanted to show that SMA connectors are quite common for such devices.
You can get them cheaply: https://de.aliexpress.com/item/1005003311808070.html

The DUT in the example is this: https://de.aliexpress.com/item/1005005704247436.html

[nctnico]

Edit: Oh, and if you need to deliver that 1..30 MHz sweep to your device under test, you need to provide a proper input for it on the DUT. (BNC, 50 Ohm terminated.)

Or SMA with adapter, which is handier.
I always try to keep the distances as short as possible so that there is less noise etc. 

But do not get overworked about the frequency. Keep in mind that the wavelength of a 30MHz sine wave is around 10 meters. With a cable length of up to 2 meters, reflections aren't an issue. If I needed to bet money, I'd put it on a display artefact of the oscilloscope. The differences in amplitude as shown by screenshots are a tell-tale sign. This is why it is so important to get some images which show the waveforms at the frequencies where the amplitude is highest and lowest.

[eTobey]

I now zoomed it further. I deliberately picked a place where there was an artefact. It was a break in the sweep without a signals, but it is rather narrow, and it gets overlayed by the signals on the side. i could measure 3 break in about 13.4 ms interval after each other.

It is only visible in normal aqusition, and not in peak acq. Changing the acqs. mode does not change the sweep amplitude that is seen on the scope.

[Aldo22]

@eTobey: Somehow I'm lost.
Could you repeat the question?
Is it about Siglent SDG1032 sweeping, coax/wiring, oscilloscope settings, frequency response of your DUT or what exactly?

Maybe it would also be useful if you could show your DUT?

[eTobey]

Is it about Siglent SDG1032 sweeping, coax/wiring, oscilloscope settings, frequency response of your DUT or what exactly?

Maybe it would also be useful if you could show your DUT?

It was a bit of everthing, except scope settings. I just added that screendump, because someone wanted it (kind of).
But i am satisfied with the information here and my findings: https://www.eevblog.com/forum/testgear/siglent-sdg1032-amplitude-not-constant-when-sweeping/msg5506162/#msg5506162

[nctnico]

I now zoomed it further. I deliberately picked a place where there was an artefact. It was a break in the sweep without a signals, but it is rather narrow, and it gets overlayed by the signals on the side. i could measure 3 break in about 13.4 ms interval after each other.

It is only visible in normal aqusition, and not in peak acq. Changing the acqs. mode does not change the sweep amplitude that is seen on the scope.

To me this looks like a display artefact. You have to keep in mind that a digital oscilloscope (in general, not just Siglent) can't use the entire memory length to produce the image for a trace because that would be way too slow. So what the oscilloscope does is called sub-sampling. It takes the value for every Nth sample (1 in 100 for example) and use that to produce a trace envelope outline. Typically this works but for frequency sweeps, you can get intermodulation products which then start to show as valeys which don't make sense.

[TurboTom]

Please deepen your knowledge in sampling theory! If you sample a signal swept to 30MHz at 10MSa/s, these aliasing patterns are exactly what to expect. Please make sure you have sampling memory on your scope set large enough to match the Nyquist criteria, i.e. Fsample > 2*Fsignal! This is digital scope basics!

@nctnico beat me to it   

P.S: I've got the impression your DG812 wasn't that much "rubbish" after all... 

[eTobey]

To me this looks like a display artefact.

Please deepen your knowledge in sampling theory! If you sample a signal swept to 30MHz at 10MSa/s, these aliasing patterns are exactly what to expect.

Indeed it is, but it does not really matter here, as those valleys are very very narrow, and zoomed out you can barely see them. Using peakmode, those narrow valleys even completely dissappear.

[nctnico]

Indeed it is, but it does not really matter here, as those valleys are very very narrow, and zoomed out you can barely see them. Using peakmode, those narrow valleys even completely dissappear.

That is because peak-mode runs the ADC at full speed before doing the decimation so you won't get sampling related intermodulation products. The post-processing (to make the trace display) intermodulation will still be there though.

[eTobey]

I have now made a final measurement, using proper connection, and using the maxhold function.

It shows clearly a asymmetry and the amplitude varation over the sweep in great detail. Too bad, that the math dont work with zoom mode (all of them?).

[newbrain]

I have now made a final measurement, using proper connection, and using the maxhold function.

It shows clearly a asymmetry and the amplitude varation over the sweep in great detail. Too bad, that the math dont work with zoom mode (all of them?).

Do I read it correctly that now, with proper cabling, the variation is about 50 mV from end to end of the sweep?
And the signal level is about 2 Vp (no measure, and graticule very faint)?
So ≈0.2 dB of change.
That's not bad in itself, and also in spec, from the datasheet:

Amplitude flatness -0.3 +0.3 dB 50 Ω load , 2.5 Vpp, compare to 10 kHz sine

[eTobey]

I have now made a final measurement, using proper connection, and using the maxhold function.

It shows clearly a asymmetry and the amplitude varation over the sweep in great detail. Too bad, that the math dont work with zoom mode (all of them?).

Do I read it correctly that now, with proper cabling, the variation is about 50 mV from end to end of the sweep?
And the signal level is about 2 Vp (no measure, and graticule very faint)?
So ≈0.2 dB of change.
That's not bad in itself, and also in spec, from the datasheet:

Amplitude flatness -0.3 +0.3 dB 50 Ω load , 2.5 Vpp, compare to 10 kHz sine

It is a 5.05 Vpp "amplitude", but this is at a limit, what i did not notice then. Have you read the details on the picture name? (I guess its a bad place for that?)

Amplitude flattness is measured on a fixed frequency? If so, it could be better, but i think its probably a differend rail that is used for that (2.5V vs. 5V), and that could be different.

[newbrain]

Have you read the details on the picture name?

OOPS! Totally missed the picture name.

Then I stand corrected. It's a 50 mV variation from 1 to 30 MHz on a ~2.5 Vp signal.
I.e, 20 x (log10(2.525 - 0.05) - log10(2.525) ~= -0.174 dB , a bit less than my estimation.

Kudos to the output stage designers in Siglent (or FW designers for the calibration table, or both).

But I think this is a good lesson for better probing, connections, and termination techniques.

Don't worry, we all were wet behind the ears, once: just three days ago a dear and usually really smart friend of mine was probing a small RF power amplifier prototype.
He sent me some screen capture - I could not understand what was happening - for me it had worked with no issue...
Then he casually mentioned that he was not using the probe spring.
"Oh, definitely do not use the alligator for RF or fast edged signals!"
"No worries, I'm not using that either!"
 

[eTobey]

But I think this is a good lesson for better probing, connections, and termination techniques.

Don't worry, we all were wet behind the ears, once: just three days ago a dear and usually really smart friend of mine was probing a small RF power amplifier prototype.

Indeed it is. Never thought that those kind of slow signal changes could have such a dramatic effect. Thought it would only be relevant for higher frequencies (> 60MHz) or digital signals. I also discovered today, that compensation is as important, while looking with 2 probes at a triangular wave and plotting the difference. If i remember right, it was even at just 1kHz.

I do not worry, i do learn (while trying it out mostly) ;-) There are some good videos about the gnd spring on youtube. It seems to me, more than for proper termination.

[thm_w]

I would say many of the issues were scope/gen settings and not probing technique.
Remove the terminator and it should be flat up until ~20MHz.

Compensation is important at low frequencies, yes, it is called LF compensation sometimes https://www.electronics-notes.com/articles/test-methods/oscilloscope/scope-probe-compensation.php