New Rigol HDO1000 12-bit DSO - BUGs

[markone]

Spotted right now another issue with Trigger with low level signal, i.e Sine 1MHz 50mVpp does not trigger with zero and negative thresholds (see attached image).
Rising threshold to 5mV and / or signal amplitude to 60mVpp scope starts to trigger ...

Recalled default setting, trigger problem disappeared ... 

[Fungus]

The same with i2c, I have to set the threshold to 4.4V for the decoder to decode correctly (The i2c signal on the Arduino has a slow rise time due to internal resistor value).

FYI, you probably just need to add an appropriately sized pull-up resistor for your bus capacitance.

There's nothing "probable" about it. You're supposed to add external pullups. It's I2C 101.

[TK]

The same with i2c, I have to set the threshold to 4.4V for the decoder to decode correctly (The i2c signal on the Arduino has a slow rise time due to internal resistor value).

FYI, you probably just need to add an appropriately sized pull-up resistor for your bus capacitance.

There's nothing "probable" about it. You're supposed to add external pullups. It's I2C 101.

Not sure, AVR microcontrollers have internal pull-up resistors and Arduino is using them on I2C lines and the values are not good for I2C, so it is I2C 101 to have pull-up resistors on the BUS, just not easy to have the right value when using Arduino boards.

[Fungus]

AVR microcontrollers have internal pull-up resistors

Yes, in the 30-40k Ohms range - for minimal low power consumption in low-power devices.

I2C signals needs fast rise times so you need a lot less resistance than that for it to work properly.

(or you can limit your bus speed to something slow that works with the Arduino's pullups)

[Marchello]

DHO1074, firmware - 2.04

I have a problem with disabling the display of scales at the edges of the screen. If I turn them off in the menu (Display->Show Scale "Off") they turn off. But if I press Start or Stop Waveform Capture, the scales turns on by itself again!
I think it's not normal.

Can someone repeat this?

[dominicc]

There's a bug with the SPI trigger thresholds in FW 00.02.07, and probably earlier:

Steps to repeat:

1)   Setup 4  channels on SPI CS/CLK/MOSI/MISO using a 3.3v p-p SPI signal.
2)   Adjust each channel to a non-zero offset.
3)   Change the trigger to SPI trigger, using CS to trigger with a 1.5V.
4)   Configure SPI decoding, Set the SPI decoding levels to 1.5V on each channel.
5)   Ensuring the decode event view is visible and displaying correct data.
6)   Now go back into the trigger and simply change the CS channel to a different channel and back again.
7)   Observe the event view is now empty, this is because the SPI decode trigger levels have changed (!!!)
   Now go back into the SPI decode setup, observe all the trigger levels still show 1.5V (!!!)
9)   Now click one, observe the UI now shows it’s current value, which is different to 1.5V (!!!)
10)   Adjust them all back to 1.5V, and observe that the event view is again correct.

See video recorded from the webui:



Other things, other than the Step 6 above, also cause the issues with incorrect or reset trigger levels.  For instance, turning off and on and on any channel will reset one or more to 0v.

I've reported this to my supplier and will follow up with Rigol based on their response.

[dominicc]

The rotary encoders on my DHO4204 aren't reliable, as this video demonstrates:



It feels like the light isn't passing through some of the holes in the encoder wheel as the issue is present at the exact same spot on the dial.  I've not taken mine apart, just guessing based on what I know about how MCUs process the signals from a rotary encoder (example attached).

[dominicc]

Has anyone been able to get the video triggering working on a PAL or NTSC signal?

I can't get it to work here...  (see images attached).

It works OK on my old DS1074Z with the same signal... Did I miss something?

[maxwell3e10]

When saving data to .csv file in full memory format the x-scale tInc is wrong. Sometimes it is the same as for "screen" format- which only has 1000 points, but other times seems to be off by another factor. So often one cannot tell the time axis of the data saved.

When the scope is running in average acquisition mode, it stops triggering randomly and one needs to press "run" again.

[hfiennes]

Just got my DH1074 today, and I'm having issues with i2c decode.

Basically, the scope is always assuming 10 bit addressing. There is nowhere to set 10 bit (I guess the "without/with" sets 7/8 bit) so I can't clear it. I've tried setting 7 bit addressing in the I2C trigger and copying this across with the "copy trig" button in the decoder setup, but no luck.



It came with 2.00 and have updated to 2.11, but the bug persists. Anyone else see this? I was going to try a factory reset in case there's something randomly stale, but couldn't find one of those either.

[Fungus]

Basically, the scope is always assuming 10 bit addressing. There is nowhere to set 10 bit (I guess the "without/with" sets 7/8 bit) so I can't clear it. I've tried setting 7 bit addressing in the I2C trigger and copying this across with the "copy trig" button in the decoder setup, but no luck.

(Attachment Link)

It came with 2.00 and have updated to 2.11, but the bug persists. Anyone else see this? I was going to try a factory reset in case there's something randomly stale, but couldn't find one of those either.

I counted the clock pulses between the first start and stop in that image and there were 18 of them. Is the 'scope wrong in assuming 10 address and 8 data?

[skander36]

...
It came with 2.00 and have updated to 2.11, but the bug persists.
...

First mention that the Fw. 2.11 is out. I missed that.

From release notes:
00.02.11 2023/08/17

     - PassFail function optimization
     - Ref function optimization
     - Read screen data, fix the abnormal number of return values
     - Waveform distortion before triggering with large time base
     - CAN trigger related issues resolved
     - Fixed the wrong format of saved memory bin wave file
     - Fix low storage depth, after ZOOM is turned on, the square wave will be stabbed
     - Fix the data abnormality at the end of the high-resolution mode waveform
     - Fixed the inconsistent behavior of XY window switch operation
     - Fix the invalid initialization of webcontrol network settings
     - Fix Pass/fail cannot be terminated
     - Fixed the problem that SCPI could not get the measurement value
     - Fixed the problem of loss of test mask in special cases

Anyone is still encounter trivial bugs like scope crashing or Android stop messages (like in Dave's tests with DHO 800)?
From 2.07 version I did not encounetr this kind of problems but maybe I'm not using the scope enough.
 

[tooki]

The rotary encoders on my DHO4204 aren't reliable, as this video demonstrates:

[...]
It feels like the light isn't passing through some of the holes in the encoder wheel as the issue is present at the exact same spot on the dial.  I've not taken mine apart, just guessing based on what I know about how MCUs process the signals from a rotary encoder (example attached).

FYI, practically all rotary encoders used as human-machine interface (HMI) elements are electromechanical, basically a type of rotary switch, using flat contacts on one half and slightly offset leaf spring contacts on the other half. (Which contact type is on which part -- the rotor and the base -- varies by model.) Optical rotary encoders are very expensive and are used only on high-reliability HMIs like aircraft cockpits, and as the absolute-position encoders on rotary shafts in things like robotics.

The only optical rotary encoders found commonly in consumer devices are the ones used on mouse scroll wheels. (And long ago, when mice had balls, as the ball encoders, too.) I suspect it's simply because the wheel needs to be clickable, and the simple slot-type photointerrupter allows the use of a slotted wheel that can move up and down a bit.

See e.g.
https://newscrewdriver.com/2022/06/08/quadrature-encoder-rotary-knob-with-detent/
and

[Valentin72]

I recently received my Rigol DHO1072 from Aliexpress. Wow, big & beautiful. Almost the first time I turned it on and I found a bug, recipe:
- enable 10x probes on both channels (note, without 10x it is not repeated)
- trigger rs232 115.2k
- decoder channel 1 rs232 115.2k bus status on
- turn multipurpose button 1 until full screen "RIGOL.SCOPE has stopped"

ps many thanks to this forum, opened my rigol up to DHO1202

[Fungus]

Almost the first time I turned it on and I found a bug

Update your firmware and try it again. 

[skander36]

Almost the first time I turned it on and I found a bug

Update your firmware and try it again. 

Yes. I can't confirm this behaviour with the last ver. 00.02.11.

L.E. I remember that in one of the early fw. versions I've had a problem with the multifunctional knob(nr.1) where this does not taking any action. This behaviour has disapeared suddenly after few days.

[Valentin72]

You are right, I had version 02.07, I updated to 02.11 and the bug is not reproduced, thanks

[Fungus]

Out of interest: Does the reference voltage for serial decoding work correctly when the probe attenuation is set to 10x? Or is it 10x smaller than the value you set?

There's a bug in the DHO800 series at the moment and I was wondering is it's in the DHO1000.

[skander36]

Out of interest: Does the reference voltage for serial decoding work correctly when the probe attenuation is set to 10x? Or is it 10x smaller than the value you set?

There's a bug in the DHO800 series at the moment and I was wondering is it's in the DHO1000.

On 10X it is ten times bigger than on 1X, if I understand correctly your question.

[Fungus]

On 10X it is ten times bigger than on 1X, if I understand correctly your question.

Yes, that sounds like the same bug.

The problem is that the marker for threshold voltage is shown in the wrong place on screen so it can lead to it not working properly. To get it in the correct place you have to set the probe attenuation to 1x mode.

[skander36]

Yes. I always need to "drag" it from the bottom ...
I forgot about this. Usually I'm using Rigol 5k for protocol decoding.

[Alex-lab]

Some new bugs and "features":
a) If you select save both PNG and CSV on Quick, after restart it will save PNG and BIN.
b) File names are still like Rigol1, Rigol2... quite annoying, I would prefer to have simple date-time stamp, as we have for smartphone photos.
c) Appearance of waveforms in PNG files depends on the Run/Stop state. - Not sure, if this is a bag. In running mode it looks much smoother (High-res).
d) Real bug. If you stop the waveform, then change x-scale (s/div) for more details, after that varying the x-position breaks x-scale factor. So you have the original appearance of the waveform (as it was at stop pressed), but with a wrong X-scale. Doing the same in reverse order can bring it back into correct original view.

[Fungus]

a) If you select save both PNG and CSV on Quick, after restart it will save PNG and BIN.

Might be related to the whole "state isn't saved correctly" bug on the DHO800. There's a few things that don't save there.

b) File names are still like Rigol1, Rigol2... quite annoying, I would prefer to have simple date-time stamp, as we have for smartphone photos.

Does it have a RTC with battery? If not, it's going to be worse.

c) Appearance of waveforms in PNG files depends on the Run/Stop state. - Not sure, if this is a bag. In running mode it looks much smoother (High-res).

On the DHO800 there's a display setting called "Waveform freeze" to control this.

[NE666]

Does it have a RTC with battery?

Yes, it does.  So an option for timestamp based filenames would be nice.

[Alex-lab]

On the DHO800 there's a display setting called "Waveform freeze" to control this.

Indeed, but it doesn't change anything.

Some more bugs:
a) Adjusting Y-position makes incorrect rounding for zero value on the scale. In fact it is not a zero, but instead a very small value (like 26.048uV).
b) Incorrect evaluation of Vbase value. Here is just example, same issue is for many periods of the same signal.