Rigol DHO800/900 Oscilloscope Bug Reports + Firmware

[Fungus]

When I adjust the CH1 voltage scale, between 50mV and 100mV, the signal visibly and measurably shifts by -2mV.

What firmware is that?

Try the latest one:

https://www.eevblog.com/forum/testgear/rigols-new-dho800-oscilloscope-unbox-teardown/msg5272446/#msg5272446

[csuhi17]

What firmware is that?

00.01.02.00.02
I updated the last firmware today, then calibrated it after 1 hour of heating.

It happens on all four channels.
I checked the AWG signal with another oscilloscope (Micsig) and when I changed the vertical sensitivity on the Rigol's active channel, the AWG signal shifted.
I don't know if anyone else has noticed this or if it's just me.
I don't remember if anyone mentioned it.
I have not tried with older firmware.

[pakakezu]

I can reproduce the problem.
Setup is a bit jank for some reason the noise is terrible even with BW limit.
I think this is an ARB issue. Because feeding on any other channel and changing CH1 does the same thing.
Reduced freq to 1khz. Measuring with a multimeter changing ch1 does affect output offset.  20mv signal 10mv offset Not termintated with 50ohm, only multimeter connected.
Terminating into an attenuator produces the same thing.
Also latest fw. 00.01.02 build 24/01/03
It is not because of the relay current, because changes even when the relay is not clicking.
Also for the life of me cant upload an arb waveform. Support is silent.

[pakakezu]

Outputting DC from arb. Same thing. Hopefully copy paste problem with calibration, and not hw.

[Randy222]

Since I don't know I have to ask.
When doing FW upgrade, does it replace FFT cal file? I do assume any edits I have done to Rigol files would be replaced from files that come from the new FW gel upgrade?

[AndyBig]

Since I don't know I have to ask.
When doing FW upgrade, does it replace FFT cal file? I do assume any edits I have done to Rigol files would be replaced from files that come from the new FW gel upgrade?

When updating the firmware from 00.01.01.xx.xx to 00.01.02.xx.xx, the calibration files change, yes. More precisely, one of the files changes - cal_vertical.hex. When updating the firmware from 00.01.02.00.00 to 00.01.02.00.02, the calibration files do not change.

[AndyBig]

00.01.02.00.02
I updated the last firmware today, then calibrated it after 1 hour of heating.

It happens on all four channels.
I checked the AWG signal with another oscilloscope (Micsig) and when I changed the vertical sensitivity on the Rigol's active channel, the AWG signal shifted.
I don't know if anyone else has noticed this or if it's just me.
I don't remember if anyone mentioned it.
I have not tried with older firmware.

Yes, there is something strange with the offsets on the latest firmware 00.01.02.00.02. I don’t know if this was the same on older firmwares. The offsets change when you switch the vertical scale, they change when you turn on/off the 20 MHz band limit... And the offsets are different on different channels, for me this is most visible on channel 4. I'll now try to record a video demonstrating this.

[AndyBig]

[Robaroni]

I think you need a termination on that channel.

[csuhi17]

I have no offset when I turn on the BW limit on its own channel.

I made a video to make it easier to understand.
After uploading it, I played with it a little more and noticed that if I change one of the following values, the AWG signal also slips or starts shaking.
When I change or turn on the "BW Limit", "Bias", "Offset" and "Scale" parameters of another channel.

[AndyBig]

I think you need a termination on that channel.

There is a 50 Ohm terminator there

[Robaroni]

Okay, thanks missed it.

[JDW]

The DHO800 manual says the Probes default to 1x, and I confirmed in the CH menus that is true.  But why default to 1x? They ship with the mechanical switch set to 10x.  And of course, most users will use the 10x setting much of the time anyway.  So it would make sense if the Probe Ratio defaulted to 10x.

I also am not seeing many posts about this.  I guess everyone is either silently changing their setting to 10x without posting about it, or DHO800 users are simply leaving it set to 1x, even though the mechanical switch is set to 10x.

[ebastler]

The DHO800 manual says the Probes default to 1x, and I confirmed in the CH menus that is true.  But why default to 1x? They ship with the mechanical switch set to 10x.  And of course, most users will use the 10x setting much of the time anyway.  So it would make sense if the Probe Ratio defaulted to 10x.

I also am not seeing many posts about this.  I guess everyone is either silently changing their setting to 10x without posting about it, or DHO800 users are simply leaving it set to 1x, even though the mechanical switch is set to 10x.

As long as the scope remembers the settings when powered off, I don't see this as a problem. Having to change the probe settings during use is not uncommon -- not so much because you want to use the passive probes at 1x, but because you may want to use direct BNC cable connections to some signal source instead of the 10x probes sometimes. So it's not a big chore for the user to set up his preferred attenuation during the initial use.

Shipping the scope with 10x as the default setting would probably cause another problem: Users complaining that their new scope does not go down to 500 µV/div as advertised, but stops at 5 mV/div.

What I do find unfortunate is that the DHO scopes do not show the current probe setting in the channel info boxes. All other information is there (sensitivity, bandwidth, coupling etc.), but the probe attenuation factor is not displayed, and is actually a bit hidden, two dialogs away.

[JDW]

As long as the scope remembers the settings when powered off, I don't see this as a problem.

It does NOT remember the change to 10x when you power off and then on again.  It resets to 1x, even if you do not press the AUTO button after restarting. Therefore, it is a problem that I would classify as a BUG.  But more than that, it's a huge headache to have to manually change all 4 channels to 10x every time you power on the scope!

[ebastler]

It does NOT remember the change to 10x when you power off and then on again.  It resets to 1x, even if you do not press the AUTO button after restarting. Therefore, it is a problem that I would classify as a BUG.  But more than that, it's a huge headache to have to manually change all 4 channels to 10x every time you power on the scope!

Have you set the power-on state to "Load last"? It's a switch in the Utility > Setup menu.

[csuhi17]

it's a huge headache to have to manually change all 4 channels to 10x every time you power on the scope!

What is "Utility->Setup->Load last" set to for you?
Default or Last?

[JDW]

I appreciate the excellent tip!  Mine was set to Default.  I changed it to Load Last, then changed all 4 probes to 10x, then restarted, and now I see it saves the 10x setting.  Great!  Thank you!

[Randy222]

With 01.02.xx.xx FW, if you just have a channel w/o connection, or terminated, and you start to dial up the horizontal time base (more time across the window), do you come across a few selections where the "flat line" paints the whole bottom half of the screen?
That's what I get.

[AndyBig]

With 01.02.xx.xx FW, if you just have a channel w/o connection, or terminated, and you start to dial up the horizontal time base (more time across the window), do you come across a few selections where the "flat line" paints the whole bottom half of the screen?
That's what I get.

Hmm, I haven't encountered such a bug.
Maybe some specific conditions are needed for its appearance?

[calabr]

Hello all!
 New bug found - first call for cursors after startup always show 100% cursor brightness. No matter what is set in display settings.
After opening Display menu - brightness become as it set, even if nothing changed.
 DHO924 all firmware, including 0.1.2.0.2

[Randy222]

Either more bugs coming out from 01.01, or a bunch news bugs created in 01.02

If they are new from 01.02, that's very frustrating.

The odd display offset seen on an open channel (or terminated) while changing vertical scale (volt) does not appear if I set channel BW(limit) to 20.
Disable the BW limit and that odd display offset is seen. Then at some point playing around with the channel BW limit from OFF to 20, the option for OFF went missing.

This is the odd painting I get, why would it paint the whole bottom half like this? I only changed time scale one click wider than +-100ms.

[Lathe26]

Bug: NTSC trigger is non-functional

On a DHO924S, the Video NTSC trigger is not working (i.e., displays random time points using Auto while Normal and Single never trigger).  This is using firmware 00.01.01.  Different composite NTSC signal sources have been attempted but the main one has been a Tektronix Pathfinder TSG 95 which generates a known good signal that other oscilloscopes, vectorscopes, and TVs are fine with.  I have not experimented with other video formats such as PAL, 480p/60, etc.

I've experimented with different triggering settings.  The first value for each setting is the value I used most frequently across different experiments.

• Polarity of Positive and Negative.  The image for Positive shows a normal composite NTSC signal with a negative voltage sync pulse.
• Trigger level set to various voltages, though most online videos for other Rigol scopes show it is best to set the level to be in the sync pulse's voltage range
• Using with and without 75 Ohm termination.  Using with termination shows the expected normal voltage ranges.
• Syncing on All Lines, Line, Odd, or Even
• Noise Reject off and on
• Using DC or AC coupling
• Set probe input to 1X and 10X (yeah, was getting desperate with trying different values)
• Confirmed that triggering set to CH1 where the signal was attached

I'm still willing to concede that the error could be occurring between the chair and the touchscreen, though it is looking unlikely.

Edit: Load Last is currently set to Default to avoid leaving the oscilloscope in a goofy state.

[TurboTom]

I use the DHO924S's own AWG, with a 20mV signal.
two channels are active CH1 and CH3, the AWG is connected to CH3 (5mV/div), the other is empty.
When I adjust the CH1 voltage scale, between 50mV and 100mV, the signal visibly and measurably shifts by -2mV.

When tested with the other channels, they do not relate to each other in the same way.
In my case, for example, if the signal is on CH3, by adjusting CH1 or CH2, it goes down by 2-3mV, but CH4 does not suppress CH3 as much.

Red REF is when CH1 is active and set to 100mV/div.

Might actually be related to this finding: https://www.eevblog.com/forum/testgear/rigol-dho804-test-and-compare-thread/msg5112108/#msg5112108

My take on this is that the AWG module's high output level branch (THS3095) draws its power from (locally) unregulated rails. Since the analog frontends are pretty power intensive and the balance of this power may teeter between positive and negative rails, depending on the settings, the rails powering the AFE's local regulator(s) may very well fluctuate considerably. Rigol should have added local regulation to all AWG module supplies. Either greed, negligence or the lack of voltage headroom pushed them to do without those and voila.... At least, at low AWG level settings, the problem is absent.

[ebastler]

Bug: NTSC trigger is non-functional
[...]
I'm still willing to concede that the error could be occurring between the chair and the touchscreen, though it is looking unlikely.

I don't think this is a user error. Problems with video triggering have been reported months ago for the DHO1000, https://www.eevblog.com/forum/testgear/new-rigol-hdo1000-12-bit-dso-bugs/msg5007853/#msg5007853, and then for the DHO800 as well, https://www.eevblog.com/forum/testgear/rigol-dho800900-oscilloscope-bug-reports-firmware/msg5223633/#msg5223633. Looks like Rigol has not been able to fix this yet?