Rigol DHO800/900 Oscilloscope Bug Reports + Firmware

[shapirus]

Level of trigger is not saved when turning off....

This is even worse than saving nothing at all, as you would not expect this, and then you may end up looking at the board... 

Lol what a mess.

It does save trigger level (any type) for channels 1 and 2, and does not (any type) for channels 3 and 4.

Besides, it always boots with channel 1 turned ON, even if it was off before the power-off, if one of 2, 3, 4 is turned on.

Rigol's software engineering is absolutely terrible. No, it's ok to have bugs in your code. It's not ok to fail so miserably at QA before releasing the product, and even less ok to fail to release bugfixes in a timely manner.

I think it's my first and last Rigol product. No, it's a nice little scope, I do like it, even more so because of the fact that it's very hacking-friendly. I'm keeping it. But I think I've already had enough of dumb bugs that a junior QA engineer would catch minutes after a development release was built, and of the attitude that makes this possible.

[Randy222]

Perhaps I missed the chatter in last page or so around remembering settings at power-off.
Isn't there an on/off setting for just that, to "start with last settings"?

[shapirus]

Perhaps I missed the chatter in last page or so around remembering settings at power-off.
Isn't there an on/off setting for just that, to "start with last settings"?

Yes there is. It doesn't help to get those trigger levels remembered.

[eTobey]

Now i know why they dont let the scope reload settings after restart by default: Its a complete MESS! 

As not loading single value wasnt bad enough, it gets even WORSE: 
The trigger ID format is not only not remembered (sets to binary), the right value is shown, but not triggering. It triggers on any ID it seems! 

Cant wait to get the sds814 today to get things done, and not to get confused!

[ebastler]

The trigger ID format is not only not remembered (sets to binary), the right value is shown, but not triggering. It triggers on any ID it seems! 

That's a likely type of software bug, and I think we saw at least one of a similar flavor, where a threshold setting (?) was remembered in the UI but was not actually functional. It's an easy error to make for the firmware developers: Store and recall the parameters in some set of variables on the CPU side, but forget that something also needs to be actively set up in the FPGA or other hardware component.

But because it is so easy to get wrong, it should of course also be verified as part of a systematic software QC... 

[saridis]

hello,  i have recently acquired  a DHO924S and created some bode plots for various passive components with success and no issues at all.
Untill today that i tried to do the same but using ch1 as input and ch4 as output (untill now i only used ch1 and ch2 respectively).
In short it doesnt work    . It auto ranges the vertical scale of ch4 to minimum voltage of 200 microvolts and sets the bandwidth limit at 20Mhz for that channel also, bode plot results of course were way off normal. My testing bandwidth was 100hz to 25Mhz with 5vpp output.

I checked the manual and it doesnt mention that bode plot feature can be only used with ch1 and ch2.
Any other combination i tried (ch1 and ch3, ch2 and ch3, ch3 and ch4, ch2 and ch4 ) didnt work as well.

I also tried to default the settings, restart etc but with no luck.

Maybe i do something wrong as a newbie i dont know 

(firmware version is 00.01.02)
* and the reason that i tried to use ch1 and ch4 is that i pulled from my bag the ch4 compensated probe and didnt want to adjust it again for ch2

[Houseman]

I am having trigger problems for any channel even single waveforms below 100us timebase.
I cannot trigger it correctly since for a simple 0-5Vp square waveform appears flickering as shown in attacment.
I have reduced the Horizontal Acquisition, the Memory depth. nothing.
The problem appears for any waveforms also in the 100us timebase.
What am I doing wrong?
Thank You in advance.

[electronics hobbyist]

I am having trigger problems for any channel even single waveforms below 100us timebase.
I cannot trigger it correctly since for a simple 0-5Vp square waveform appears flickering as shown in attacment.
I have reduced the Horizontal Acquisition, the Memory depth. nothing.
The problem appears for any waveforms also in the 100us timebase.
What am I doing wrong?
Thank You in advance.

You can try "Normal" trigger, it may be caused by "Auto" trigger timeout.

[shapirus]

I am having trigger problems for any channel even single waveforms below 100us timebase.
I cannot trigger it correctly since for a simple 0-5Vp square waveform appears flickering as shown in attacment.
I have reduced the Horizontal Acquisition, the Memory depth. nothing.
The problem appears for any waveforms also in the 100us timebase.
What am I doing wrong?
Thank You in advance.

Try to trigger on rising edge OR falling edge, not *both*, that you have in the screenshot.

[ebastler]

I cannot trigger it correctly since for a simple 0-5Vp square waveform appears flickering as shown in attacment.
What am I doing wrong?

The screenshot and video show that you have selected the trigger to act on both, positive and negative edges. So it will sometimes trigger at the beginning, sometimes at the end of the "high" phase of the signal. Select either positive or negative edge in the trigger setting.

In the video, you have also set the trigger threshold pretty high -- so the trigger could also be caused by the small oscillation after the signal's positive edge. Better to set the trigger threshold about halfway between the 0V and 5V levels, for best noise immunity.

Finally, this appears to be a signal with fast pulses (800 ns long) but long gaps inbetween (assuming it comes from the internal generator, which is set to 1 kHz). So the scope may "lose patience" while waiting for the next trigger event, and start a capture on its own. That's what it does in "Auto trigger" mode, to avoid a situation where you see no signal at all. As suggested by electronics hobbyist, set the scope to Normal trigger instead. 

[Mechatrommer]

playing with 16bit data out of DHO800 to PC and plot FFT... i'm sensing 156.25MHz -30dB artefact when no signal on input (50 ohm terminated), attached also different artefact profile from 8bit DS1054Z, maybe it matters? or something i did wrong 

[Mechatrommer]

wishlist: please rigol, enable single window overlayed display on DHO800/900 like in DHO1000, its just a simple enable/disable toggle switch in FW? what do you care if we want overlayed display on small screen?

https://www.eevblog.com/forum/testgear/choosing-between-entry-level-12-bit-dsos/msg5432216/#msg5432216
https://www.eevblog.com/forum/testgear/choosing-between-entry-level-12-bit-dsos/msg5432306/#msg5432306
https://www.eevblog.com/forum/testgear/choosing-between-entry-level-12-bit-dsos/msg5432441/#msg5432441
https://www.eevblog.com/forum/testgear/choosing-between-entry-level-12-bit-dsos/msg5432615/#msg5432615

DHO1000:


DHO800/900 (no "Display Area" option):

[core]

playing with 16bit data out of DHO800 to PC and plot FFT... i'm sensing 156.25MHz -30dB artefact when no signal on input (50 ohm terminated), attached also different artefact profile from 8bit DS1054Z, maybe it matters? or something i did wrong 

I can try to take out the bin file from DHO1074, just tell more about the settings.
Also, I am interested regarding the software you have used. It's open source or 

[AndyBig]

playing with 16bit data out of DHO800 to PC and plot FFT... i'm sensing 156.25MHz -30dB artefact when no signal on input (50 ohm terminated), attached also different artefact profile from 8bit DS1054Z, maybe it matters? or something i did wrong 

This peak at a frequency of 156-odd MHz (and its harmonics) is clearly visible even on the oscilloscope itself in the FFT window. Some kind of internal interference from the operation of the circuit.

[ebastler]

This peak at a frequency of 156-odd MHz (and its harmonics) is clearly visible even on the oscilloscope itself in the FFT window. Some kind of internal interference from the operation of the circuit.

It's 1/8 of the main 1.25 GHz sampling clock, right?

[AndyBig]

This peak at a frequency of 156-odd MHz (and its harmonics) is clearly visible even on the oscilloscope itself in the FFT window. Some kind of internal interference from the operation of the circuit.

It's 1/8 of the main 1.25 GHz sampling clock, right?

Very similar to that. Maybe the interface between the ADC and the FPGA operates at this frequency?...

[Mechatrommer]

This peak at a frequency of 156-odd MHz (and its harmonics) is clearly visible even on the oscilloscope itself in the FFT window. Some kind of internal interference from the operation of the circuit.

It's 1/8 of the main 1.25 GHz sampling clock, right?

my concern if this is another riglol scenario, but since ds1000z also has it probably it is normal or not easily fixable? i'll leave that open to more expert people, and i will just enjoying my scope and try to compensate at that frequency (or avoid that region) if a task requires it. ymmv.

I can try to take out the bin file from DHO1074, just tell more about the settings.

my setting was 625MSps because i was watching a 20MHz signal, full sample rate 1.25GSps will make that signal goes further to the left, less undesirable. but that FFT plot i linked above is when no signal at 1X dso input setting and 50 ohm terminator in place. anything greater than 1Mpts points will make a descent FFT resolution plot (at least when using my app). as AndyBig mentioned you should be able to see it on your scope's display as well if the artefact presents or not. ymmv.

Also, I am interested regarding the software you have used. It's open source or 

you can google VisaDSO in soasystem.com for setup file download. i've given up promoting it since i guess new kids on the block dont like older VB app, let the lucky persons found it by themselves. sometime stranger from unknown emailed me for the code, but once i sent it, they never come back to give feedback, so its like a waste of time, and probably nobody like nor can comprehend the old VB syntaxes anyway... i saw some bugs on latest upgrade to support DHO800/900, since afaik i'm the only one who's using it, i dont bother to fix asap, unless when necessity arrived. ymmv.

[gf]

This peak at a frequency of 156-odd MHz (and its harmonics) is clearly visible even on the oscilloscope itself in the FFT window. Some kind of internal interference from the operation of the circuit.

It's 1/8 of the main 1.25 GHz sampling clock, right?

Looks like typical ADC interleaving spurs, i.e. small mismatch between the interleaved ADC branches. The spurs here¹⁾ look particularly like offset spurs. Self-calibration may help, but the accuracy and resolution of the calibration may still not be sufficient to eliminate them completely (note that offset mismatch in the amount of only a fraction of an LSB are sufficient to show up as spurs in a FFT with a large number of points).

¹⁾ located at sample_rate/N*n, where N is the number of ADC branches, and n is 1...N/2.

[core]

I can try to take out the bin file from DHO1074, just tell more about the settings.

my setting was 625MSps because i was watching a 20MHz signal, full sample rate 1.25GSps will make that signal goes further to the left, less undesirable. but that FFT plot i linked above is when no signal at 1X dso input setting and 50 ohm terminator in place. anything greater than 1Mpts points will make a descent FFT resolution plot (at least when using my app). as AndyBig mentioned you should be able to see it on your scope's display as well if the artefact presents or not. ymmv.

Also, I am interested regarding the software you have used. It's open source or 

you can google VisaDSO in soasystem.com for setup file download. i've given up promoting it since i guess new kids on the block dont like older VB app, let the lucky persons found it by themselves. sometime stranger from unknown emailed me for the code, but once i sent it, they never come back to give feedback, so its like a waste of time, and probably nobody like nor can comprehend the old VB syntaxes anyway... i saw some bugs on latest upgrade to support DHO800/900, since afaik i'm the only one who's using it, i dont bother to fix asap, unless when necessity arrived. ymmv.

Very interesting information on this little site. I am a Linux user, but I think I will try VisaDSO in a virtual machine.
Thanks for sharing!

P.S. Your "natural habitat" is incredibly small and crowded!

[ceut]

Hello !
I have a Siglent SDS1104X-E , no issue with it, but this DHO with Android inside and this user interface seems to be really great 
But I have a question about the 256 level intensity grading.
I have seen some reviews, starting by Dave one's  and it seems to me that the grading is not very good, compare with my Siglent.
Most of the time, I see big yellow color on waveform, without any grading 

Here are some pictures I have made with my Siglent, compared to the Dave's opening video of this topic 
What do you think ? Is there a kind of bug on the graphical output display of the Rigol ?

Also, while I was making the screenshots, I have found that the Rigol has only 10 divisions on the time axes , whereas the Siglent shows 14.
So on the 2 first screenshot, we can see more of the signal on the Siglent than on the Rigol

[tautech]

I have seen some reviews, starting by Dave one's  .......

You've been not here for a bit you said so may have missed......Dave just received 2 of the new Siglent 12bit scopes to review....

[shapirus]

I'm not sure. In practice, I find it doing its job very well at intensity grading to help discern noise from useful signal, catch infrequent events etc.

There's a lot of room for possible adjustment using the trace brightness and the persistence timeout settings.

One caveat: intensity grading works only when the trigger is triggering. Otherwise, the trace is drawn at 100%.

[Fungus]

Also, while I was making the screenshots, I have found that the Rigol has only 10 divisions on the time axes , whereas the Siglent shows 14.
So on the 2 first screenshot, we can see more of the signal on the Siglent than on the Rigol

Rigol grid squares are wider...

So on the 2 first screenshot, we can see more of the signal on the Siglent than on the Rigol

I can see more detail on the Rigol.

[Lathe26]

Seems like Rigol has slowed down firmware releases for the DHO800/900 family.  They were averaging ~1 release per month but that change after December 2023.

• v00.01.02.00.02  2023/12/28 (latest release)
• v00.01.02.00.01  2023/12/5
• v00.01.02.00.00  2023/11/2
• v00.01.01.00.02  2023/09/12
• v00.01.01.00.01  2023/08/10
• v00.01.00.00.19  2023/07/24

[Howardlong]

Seems like Rigol has slowed down firmware releases for the DHO800/900 family.  They were averaging ~1 release per month but that change after December 2023.

• v00.01.02.00.02  2023/12/28 (latest release)
• v00.01.02.00.01  2023/12/5
• v00.01.02.00.00  2023/11/2
• v00.01.01.00.02  2023/09/12
• v00.01.01.00.01  2023/08/10
• v00.01.00.00.19  2023/07/24

Much of it needs rewriting TBH!

It's difficult to know whether they're knuckling down to make it into a usable tool, or whether they've just given up with it.