Rigol DHO800/900 Oscilloscope Bug Reports + Firmware

[Mechatrommer]

i think its a bug. no worse signal will trigger like this (attached) iirc i encountered such trigger anomaly while using DS1054Z but didnt care too much as i concentrated on debugging my circuit. dont remember what caused such trigger event... tinysa based on PLL if i'm not mistaken, not some jittery DDS IC.. i dont think PLL (square signal) or filter after that will cause such nasty noises/spikes. its very unlikely to be in that place. ymmv.

[Fungus]

There's clear lumps and bumps in that signal, wouldn't surprise me at all if you're just catching points where the signal goes backwards.

Again: Can we press STOP on a bad trigger (use record mode), zoom in on the triggered area, and take a close look? Use the 12 bits.



And... can we do it in the Test and compare thread:

https://www.eevblog.com/forum/testgear/rigol-dho804-test-and-compare-thread/

The first post in this thread says to keep it clean and tidy and this is already devolving

[ebastler]

There's clear lumps and bumps in that signal, wouldn't surprise me at all if you're just catching points where the signal goes backwards.
Again: Can we press STOP, zoom in on the triggered area, and take a close look? Use the 12 bits.

I appreciate that you like your DHO800. But please don't let that cloud your judgement.

Look at how far away the level of the mis-triggered signal is from the trigger threshold in some of the demonstrations. What level of "noise" would you expect on the input signal? "Edge trigger" is not just about the signal having a positive or negative edge, but about passing the threshold in that direction!

Also, please consider that most of the demonstration was done with signals >= 125 MHz, at very fast timebase settings. Would you expect the scope to resolve and show tiny, much faster spikes in STOP mode?

[csuhi17]

It is the same for Auto and Normal settings.

I couldn't catch the event in record mode, it's as if it pulls the wave into a good position when I scroll frame by frame.
https://youtu.be/MWDAznqXl3g
The event is visible in the background if you look at it, but when I play it again frame by frame, it doesn't appear.
If I play the recorded wave several times, the error does not always appear around the same frame.
I also checked the signal with the STO2202C and also with infinite endurance setting, I didn't see a single hair of signal on it.

I'm still trying, if there's another idea I can try.

[bobxyz]

It sure looks like you caught a display glitch in that video at time ~2:48 https://youtu.be/MWDAznqXl3g?t=167

Is there an easy way of going frame-by-frame in youtube videos?

[Martin72]

I can reproduce this effect from 90Mhz and then only in the negative trigger range, see pictures.
And:
If you reduce the sample rate (e.g. add a second channel), the spook is over.

[Fungus]

How is this possible? That's not the trigger level...



Video:

[ebastler]

How is this possible? That's not the trigger level...

That's very far off. Acquisition stopped and trigger level changed afterwards, by any chance?

[csuhi17]

in the video I changed the trigger.
with the rising edge setting, the error is in the negative range.
with the falling edge setting, the error is in the positive range.

Fungus
If I see correctly, your trigger is outside the trigger zone, the trigger line is set very low. Your wave is not calm though. Does the Stop/Run button correspond to the forced trigger?

[Fungus]

How is this possible? That's not the trigger level...

That's very far off. Acquisition stopped and trigger level changed afterwards, by any chance?

Nope. See the video... 

[Antonio90]

But, is it triggering? Or just stopped?

[ebastler]

How is this possible? That's not the trigger level...

That's very far off. Acquisition stopped and trigger level changed afterwards, by any chance?

Nope. See the video... 

In the video there is no trigger line in the picture, so I don't know what you did before the snapshot.

In the video, it is also obvious that you let the scope run in Auto trigger mode, with the trigger threshold set below the signal. So it triggered at random time points, when the auto-trigger timeout struck. And then you stopped the acquisition, which of course shows the signal at a random phase. 

I don't normally use that emoji on others since I find it impolite, but your stupid popcorn thing really got under my skin. Learn how to use that scope, would you?

[Fungus]

In the video there is no trigger line in the picture, so I don't know what you did before the snapshot.

It's in the same place, see the marker on the right.

[Fungus]

Learn how to use that scope, would you?

I know how to use it. I'm just saving the second video for after everybody tells me the sky is falling.

[ebastler]

I know how to use it. I'm just saving the second video for after everybody tells me the sky is falling.

It is buggy behaviour that the scope cannot trigger even when the trigger threshold is still within the signal amplitude range. I just tried on my DHO1000 and it triggers nicely all the way, with the trigger threshold set anywhere between the maxium and the minimum of the signal swing.

Your little "stop-the-acquisition" game does not contribute anything to illustrating or explaining the problem, I'm afraid. csuhi17 has also confirmed that the erroneous triggers occur in normal triggering mode too, so they have nothing to do with the auto-triggering timeout. So what are you trying to tell us?

[Fungus]

In the video, it is also obvious that you let the scope run in Auto trigger mode

Yep. All the ghostly "bad" triggers being discussed above are just AUTO triggers.

If you turn off AUTO mode it either triggers correctly or stays in WAIT mode indefinitely.

See video:


Should it trigger down at the bottom of the wave? That's a matter of preference. Me? I prefer that it doesn't.

The video also shows what happens if you turn on "noise reject", i.e. it moves the accepted trigger point further up the waveform.

(presumably looking for a steeper slope)

[ebastler]

If you turn off AUTO mode it either triggers correctly or stays in WAIT mode indefinitely.

It is the same for Auto and Normal settings.

Seems that you are not testing under the same conditions, or your scopes are behaving differently.

[Martin72]

I think I have found a little bug.
As you know, I find it a bit unfortunate that you can't change the function of a measurement window, e.g. if you have selected frequency, you can't change it to something else but either select another measurement window or delete the existing one.
This is not yet a bug, just a bit annoying.
The bug:
In this window you can call up the "setting" and set various parameters, including the source.
And that doesn't work.
For example, if you change the source from ch1 to ch2, ch2 is displayed in the setting, but the values remain for ch1.
Then you can also omit the source selection, that would probably be the easiest fix.

[Fungus]

I haven't seen any erroneous triggering at all.

[Fungus]

I think I have found a little bug.
if you change the source from ch1 to ch2, ch2 is displayed in the setting, but the values remain for ch1.

Yep, confirmed.

[ebastler]

Maybe someone could try the following trigger-related thing. I found this on the DHO1000 and wonder whether the DHO800 behaves the same:

• Look at a reasonably fast sine wave (say 50 MHz) with a fast time base (5 ns/div). Set normal DC-coupled trigger, trigger threshold around zero. Stable triggering as expected.
  
• Switch trigger coupling to either AC or LFR (low frequency rejection). Trigger time suddenly has a jitter of approx. 1.5 ns peak-peak.

What gives? One can argue that there is no spec for the trigger jitter at all -- except for external trigger, in case of the DHO1000 -- so it's not a bug in the sense of "violating a spec". But I don't see any technical reason why applying a high-pass filter should induce such jitter, so it seems to clearly violate reasonable expectations. Does the DHO800 do the same?

[Martin72]

Just switched the units on....

[Martin72]

Confirmed.
AC or LFR, you get these jittering waveform at 50Mhz.
The higher the worser it gets.
It starts already at appx 5Mhz (barely visible).

[ebastler]

Confirmed.
AC or LFR, you get these jittering waveform at 50Mhz.
The higher the worser it gets.
It starts already at appx 5Mhz (barely visible).

Many thanks for checking and confirming this. On the DHO1000 I was under the impression that it is a constant jitter time -- it just becomes more visible at the fast time bases one uses with higher signal frequencies.

The effect is even more pronounced on your DHO800 screenshot, approx. 2.5 ns peak-peak instead of 1.5 ns on the 1000. That's proportional to the difference in sampling rate, so we may be seeing an effect of the discrete time steps on an FIR filter here.

[Martin72]

Here pics from 10Mhz, 25Mhz, 100Mhz.
Try this with other scope I have here.