-
I discovered two bugs in the burst mode of Rigol signal generators DG4062 and DG4102 and DG4162.
BUG #1:
Summary: When bursting 1-cycle square waveforms, the duty cycle of the output waveform suddenly changes from 0.01% to 99.99% between pulse widths 70µs and 200µs.
Steps to reproduce:
1) Set CH1 of your signal generator to the Square waveform (Period: 70µS, StartPhase: 0º, Duty Cycle: 50%)
2) Set the Burst mode of CH1 to (Type: N_Cycle, Cycles: 1, Burst Period: 1sec., StartPhase: 0º, Source: Internal, Delay: 0 ).
This will set your signal generator to output a 35µs positive pulse every 1 second (very low duty cycle).
(the pulse is 35µs wide because the duty cycle of the square wave was set to 50% and ½ of 70µs is 35µs)
Now, slowly increase the period from 70µs to 200µs using the Square waveform menu of CH1 (which I have highlighted above in green color) and observe on a scope how the polarity of that positive output pulse suddenly changes to a negative going pulse (the CH1 waveform inverts)
Consequences: If you had a PWM amplifier/load connected to the CH1 output of that generator, then it would blow up as the driving signal suddenly inverted its duty cycle.
BUG #2: Several days later I have discovered another bug in the Burst mode.
Summary: When bursting 1-cycle square waveforms, the output signal suddenly disappears when the Burst mode Delay is rapidly decreased.
Steps to reproduce:
1) Set CH1 of your signal generator to the Square waveform (Period: 10mS, StartPhase: 0º, Duty Cycle: 50%)
2) Set the Burst mode of CH1 to (Type: N_Cycle, Cycles: 1, Burst Period: 100mS, StartPhase: 0º, Source: Internal, Delay: 89ms ).
3) Set CH2 of your signal generator to the same settings as CH1 was set above (it does not matter whether the output of CH2 is enabled).
4) Rapidly decrease the Burst mode Delay of CH1 from 89ms to 0ns (marked above in brown color) by using the numerical keypad + Enter and observe how the CH1 output pulses completely disappear for over 2 minutes. Rapidly decreasing this Delay with the knob can have the same effect.
5) Watch how the CH1 output signal stays completely disabled, (even if the CH1 Burst Delay is increased to the previous value) until the CH1 output is cycled OFF and ON again, ...or the Burst mode is cycled OFF and ON again, ...or after ~138 seconds elapse.
Models affected (both bugs):
All DG4xxxx models running firmware up to v1.08 and v1.10.
I did not test these bugs on firmware v1.09 and above but other people did.
I would appreciate a confirmation of this bug from people running firmware 1.09 and above. -
I am on 00.01.09, and can not reproduce the bug you described.
-
This bug was not tested on firmware v1.09 and above, because these firmwares "downgrade" the 200MHz model DG4202 to the 60MHz model DG4062.
Please note that if you change the model to DG4162 before upgrading you will be fine. So it's either a DG4202 with the bug or a DG4162 without it.
-
I discovered a bug in the burst mode of Rigol signal generators DG4062 and DG4102 and DG4162.
I can confirm this Bug on Software 1.08 Hard version 1.01 S/N <2000
Summary: When bursting 1-cycle square waveforms, the duty cycle of the output waveform suddenly changes from 0.01% to 99.99% between pulse widths 70us and 110us.
Error occurs with any Number of Burst Cycles.
Error occurs with Burst period at least down to 400uS , could be lower.
Error occurs on Squarewave Period 73us or greater.
If on Square wave Frequency:
No error Freq. >= 13.87160718 KHz
Error Freq <= 13.87160717 KHz
Any Ideas why the errors start at this boundary
Below I show No error then flip at 73uS Square wave period.
Could this be related to the Npulse Bug reported Here
Shoud I go to FW 1.09??
-
If on Square wave Frequency:
I don't know why the bug starts at this boundary but I know that it happens at different boundaries for different models.
No error Freq. >= 13.87160718 KHz
Error Freq <= 13.87160717 KHz
Any Ideas why the errors start at this boundary
For example in the DG4062 model, the error starts around the 100us boundary. -
After my first reply to 'GonzoTheGreat' yesterday, I upgraded my DG4062 to firmware version 00.01.10 (previously 00.01.09), and tried again to reproduce this bug.
In my case, this occurs exactly when I go from square wave period of 196.6080µs to 196.6081µs.
This doesn't occur when using pulse as source signal, but rising and falling time of pulse on the DG4062 are much longer than square. -
I upgraded my DG4062 to firmware version 00.01.10 (previously 00.01.09), and tried again to reproduce this bug.
I appreciate your vigilance despite the initial negative result.
In my case, this occurs exactly when I go from square wave period of 196.6080µs to 196.6081µs.
I have edited my initial description of the bug to cover that pulse width. -
@GonzoTheGreat
Could you try to setup phase in the burst mode to 359.999° and let me know your findings?
In my case, from 40ns to 200ms time period, no change in polarity.
For sure, there is a bug, but if this could help ... -
Could you try to setup phase in the burst mode to 359.999° and let me know your findings?
Yes, it is a workaround.
Using the SYNC output on the front panel is a workaround, too - if you do not care about the lack of amplitude adjustment. -
I have discovered a second bug in Burst mode and I have added it in this post.
-
I have discovered a second bug in Burst mode and I have added it in this post.
I received an e-mail this morning, from Rigol U.S. telling me that a new firmware release for the DG4XXX series in coming soon ...
By the way, I tried to reproduce this 'second' bug on my DG4062, FW 00.01.10, and cannot get it.
May be it's there, but will occur at different conditions that yours. -
By the way, I tried to reproduce this 'second' bug on my DG4062, FW 00.01.10, and cannot get it.
I've updated the steps needed to reproduce the Bug #2. Please try it again with the new values.
May be it's there, but will occur at different conditions that yours. -
By the way, I tried to reproduce this 'second' bug on my DG4062, FW 00.01.10, and cannot get it.
I've updated the steps needed to reproduce the Bug #2. Please try it again with the new values.
May be it's there, but will occur at different conditions that yours.
I tried again with your conditions and various other conditions for wave period, burst period and delay.
Can't reproduce the bug... -
I tried again with your conditions and various other conditions for wave period, burst period and delay.
Maybe it was fixed in your firmware version. Let's wait and see if the user Teneyes can reproduce it.
Can't reproduce the bug... -
I tried again with your conditions and various other conditions for wave period, burst period and delay.
Maybe it was fixed in your firmware version. Let's wait and see if the user Teneyes can reproduce it.
Can't reproduce the bug...
Here is a copy of Rigol DG4xxxx firmware revision list (latest one).
I don't see for the latest version anything related to the bug you are facing.
-
I have amended the steps to replicate Bug #2 again here. Please verify it.
-
I have amended the steps to replicate Bug #2 again here. Please verify it.
Bingo !!!
Same here with firmware 00.01.10.
I also discovered another bug in the AM, FM, ... mode. I will come back later with the description but I think that this one was not present before I updated the FW some days ago.
Do you intend to report that to Rigol?
Edit: Not sure if it's a bug, but feel it's annoying. Try this:
- Preset -> Factory defaults
- Output channel 1 ON. You should have at this stage a 1kHz, 5 Vpp signal on your scope.
- Press Mod (default values are AM modulation, 100Hz, Sine, 100% modulation.
- Change Shape to any other shape (Square, Triangle, etc ...). The signal goes OFF.
- Press Mod 2 times, or press Output twice to get the signal back with the new shape.
-
Bingo !!!
I applaud your patience.
Same here with firmware 00.01.10.
Could you edit your older messages so new users skimming through this thread will immediately see that these bugs are replicateable ?Do you intend to report that to Rigol?
No, I have difficulty communicating with them.I also discovered another bug in the AM, FM, ... mode.
I'd say it's a bug, too.
- Preset -> Factory defaults
- Output channel 1 ON. You should have at this stage a 1kHz, 5 Vpp signal on your scope.
- Press Mod (default values are AM modulation, 100Hz, Sine, 100% modulation.
- Change Shape to any other shape (Square, Triangle, etc ...). The signal goes OFF.
- Press Mod 2 times, or press Output twice to get the signal back with the new shape.
My unit also behaves the same way as you described above. -
I have amended the steps to replicate Bug #2 again here. Please verify it.
Finally testing #2 bug
In order to better understand the Delay function in Burst Mode , I show below the Sync Out pulse on Ch. 2 of DSO
Pix are:
1. with 10ms Squarewave period and 100ms Burst Period
2. with 10ms Sawtooth period and 100ms Burst Period 00 delay
3. with 10ms Sawtooth period and 100ms Burst Period 20ms delay
4. with 10ms Sawtooth period and 100ms Burst Period 50ms delay
5. with 10ms Sawtooth period and 100ms Burst Period Maximum delay = 89.998ms
6. with 10ms Squarewave period and 100ms Burst Period Maximum delay = 89.998ms,
with Zoom to show 2ms reset time between bursts
7. with 10ms Squarewave period and 100ms Burst Period 00 delay
NO output on jump to 00 delay
Yes I can Confirmation the Bug #2 on FW 1.08
-
I think it is not possible to get any response on a report of a bug in FW that is not the Latest version of released FWDo you intend to report that to Rigol?
No, I have difficulty communicating with them. -
Same here with firmware 00.01.10.
No confirmation for this Bug report on FW 1.08 , DG4202
I also discovered another bug in the AM, FM, ... mode.
Not sure if it's a bug, but feel it's annoying. Try this:
- Preset -> Factory defaults
- Output channel 1 ON. You should have at this stage a 1kHz, 5 Vpp signal on your scope.
- Press Mod (default values are AM modulation, 100Hz, Sine, 100% modulation.
- Change Shape to any other shape (Square, Triangle, etc ...). The signal goes OFF.
- Press Mod 2 times, or press Output twice to get the signal back with the new shape.
-
Finally testing #2 bug
Why is CH2 even monitored?
In order to better understand the Delay function in Burst Mode , I show below the Sync Out pulse on Ch. 2 of DSO
It does not matter if CH2 output is enabled - see pt.3 in the replication instructions ...it only matters if the CH2 Burst mode is enabled. -
In order to be describe the second (blue) trace in the Displaytesting #2 bug
Why is CH2 even monitored?
In order to better understand the Delay function in Burst Mode , I show below the Sync Out pulse on Ch. 2 of DSO
It does not matter if CH2 output is enabled - see pt.3 in the replication instructions ...it only matters if the CH2 Burst mode is enabled.
"I show below the DG4000 Chan.1 Sync Out pulse on Ch. 2 of DSO. in order to show the delay time from sync output until the single pulse is Generated"
Yes I missed the #3 point, thanks
Yes I can Confirm the #2 Bug in the display below I show the loss of output as I set the delay to 0
I will modify my previous post
I also observed that the Output comes back On after about 140 Seconds without any intervention
See display 2.
But again this is with Old Firmware 1.08
-
I also observed that the Output comes back On after about 140 Seconds without any intervention
Interesting, over 2 minutes....
See display 2.
But again this is with Old Firmware 1.08
I have modified the original description of the Bug #2 to reflect this.
I confirm this on my unit with firmware v1.08 but we still need to know if the output comes back after ~140sec with other firmware versions. -
I also observed that the Output comes back On after about 140 Seconds without any intervention
Interesting, over 2 minutes....
See display 2.
But again this is with Old Firmware 1.08
I have modified the original description of the Bug #2 to reflect this.
I confirm this on my unit with firmware v1.08 but we still need to know if the output comes back after ~140sec with other firmware versions.
Confirmed with firmware version 00.01.10.
Output comes back after a bit less than138s.