Defective Siglent SDG-1032X? or firmware issue?

[cds333]

Good evening friends,

So I ordered a "Like New" SDG-1032X and discovered it has a bug that causes both outputs to fail; the UI still functions fine but both outputs just go dead (also no internal switching noises where there should be). Restarting the unit always fixes the issue.

With firmware 33R3 installed, the malfunction is caused consistently by switching CH1 or CH2 to ARB while an output is active. (pressing waveforms > pg 1/2 > ARB)

The default ARB waveform is StairUp, however my 1032X does not show the correct screen for the default StairUp, instead it just shows a flatline.


If I select (ArbType > BuiltIn > StairUp) it will show the correct screen:


And if I activate either output then, everything works as it should.

I'm guessing the code has an issue with that waveform descriptor and its throwing a fatal error of some kind that bricks the outputs. No idea why the UI is unaffected.

But... then I upgraded the fw to 33R8 and the problem is fixed! (Both the flatline ARB screen and the output issue.)

I know there are things they don't list in the changelog (like "changed annoying-AF keypress to a more agreeable volume and frequency" - which they did in R8) but I would imagine if there was a bug that completely shut off both outputs and required a restart it wouldn't have gone unnoticed for ~10 months and there would be a mention of it in the changelog.

And clearly thousands of other people did not get this same bug or there would be a much lower score on Amazon and posts on here about it.

So I assume I got a defective model... but it seems to be a firmware issue with 33R3?

Questions:

Can someone explain this mystery? Or is that @tautech Siglent guy around?

Should I return it? Assuming it is a hardware issue that presents as a firmware issue?

Is there any good way to test it? It is working fine now but I want to make sure whatever issue caused this doesn't cause some other error or errors I won't find for months... (And of course it does not have a self-test feature like the comparable Siglent Oscilloscope)


Thanks!

[tautech]

Or is that @tautech Siglent guy around?

No....always.

Consider that changing to ARB while outputs are active is NOT normal operation.
However that this unit now operates like you might expect is a win for you.
Enjoy.

[2N3055]

What's up with the fonts ?
It is not usual thing here on EEVBlog. I don't it mind personally, just saying.

You bought device used with no warranty. From someone who did God knows what to it.
For instance, someone might have tried to update FW and it lost power during process and some data inside got scrambled.
Or it was customer return that underwent who knows what kind of abuse...

You didn't get defective model. That phrase means something else.
You got broken device. Which is exactly what you bought. A cat in a bag.
This is exact reason why I recommend people to buy new with warranty, from reputable local dealer..
Sometimes they have "demo" units for very good price, and they will still give warranty for it.

As for the test, it is you who need to test it. Use scope, 50 Ω BNC loads (or scope with 50Ω inputs), and some coax cables.
Self tests cannot test everything. For instance, input connectors and such.

Go through waveforms on both channels and verify they all produce correct signal.
Then go through levels on both channels, and verify that signal levels are correct.

100 Mhz scope should have flat enough frequency response for this test.
It shouldn't take you more than half an hour to go through enough combinations to verify all the waveforms and to verify all attenuator ranges on output work  correctly.

https://www.siglenteu.com/wp-content/uploads/dlm_uploads/2022/06/SDG1000X_ServiceManual_SM0201X-E01B.pdf
Page 35 Performance verification.

Make note that on page 38, amplitude specifications are in dBm (power).
Those numbers translated to voltages mean that when AWG is set to 1 V P-P you should read anywhere from 0,966 to 1,035V P-P and if you set generator for 2.5V P-P then output amplitude needs to be between 2,416 and 2,588 V.
Across the whole frequency band of AWG. If it is inside that interval the you are good.
As I said, use good (and short) BNC coax cables, and on measuring device side (scope for instance) you need to have 50 Ω termination (good external pass through terminator or internal to scope).

[cds333]

Sorry if I mislabeled you; you literally have "Siglent" in your signature...

Thank you for letting me know that. It would be nice if the manual included such a warning.

I would assume most people would expect that a dual channel generator would allow you to change one channel's waveform without deactivating the other channel.

So you're saying that setting CH2 to ARB while CH1 is outputting a sinewave is not recommended?

If so, why?

Thanks

[Fgrir]

I have a SDG2042X, but from what I have seen these two generators seem to share similar firmware quirks.  I have found that these generators will interrupt their output signals just about any time any parameter is changed, so I really wouldn't recommend changing any settings while you have an active output that you need to be continuously generated.  For example, when I have CH1 generating a 1KHz sine and CH2 set to OFF, changing the waveform on CH2 will cause the CH1 output to go to zero for about 8ms.  I agree that this is not desirable behavior, but I think it is part of their design.  I can't say whether you have a bad unit, but if the newer firmware fixed your problem I'd probably assume you just stumbled on a firmware bug.

[cds333]

I am hoping someone else has the 1032X and can confirm that 33R3 firmware also gives them the flatline waveform for the StairUp pattern when first entering the ARB screen. At least then I would know that mine is not unique.

[2N3055]

I am hoping someone else has the 1032X and can confirm that 33R3 firmware also gives them the flatline waveform for the StairUp pattern when first entering the ARB screen. At least then I would know that mine is not unique.

 It does not. That is the point.

[cds333]

Very good. Thank you.

I am still trying to decipher how exactly a malfunction like this happens. It is capable of displaying the correct screen, and generating the correct waveform, I just have to go into the menu to get it. It's only the default screen that is affected. So somehow a hardware problem is causing the code to glitch... Very odd.

[2N3055]

Very good. Thank you.

I am still trying to decipher how exactly a malfunction like this happens. It is capable of displaying the correct screen, and generating the correct waveform, I just have to go into the menu to get it. It's only the default screen that is affected. So somehow a hardware problem is causing the code to glitch... Very odd.

I told you some examples. If you start FW upgrade and yank a power cable (or loose power) during process file system can get scrambled. Same as your PC.
It is an OS with a disk inside..

But after you reflashed it all is OK now? This is what you said..

[cds333]

Correct, but then when I reflashed the previous version the bug came back. Flash the new version it's gone. The flashes are going fine- I've gone back and forth 3 times now.

But I know for a fact that the first owner did not flash the firmware because the startup count was at 2 when I first turned it on. (Also means Amazon Warehouse didn't turn it on to verify it even worked (unless the first guy never even turned it on)).

Anyway when I discovered I could turn the 2042X into a 2122X in under 10 minutes I made up my mind about this one's fate. Hopefully Amazon will not resell it to another unsuspecting customer. I plan on leaving a hidden note explaining it is defective.