Rigol DS1054Z Software or Hardware Bug?

[bitshift]

Hey guys,

I have a Rigol DS1054Z and I've been playing around with some filters today. I was comparing the output of the first filter on CH1 versus the output of the second filter on CH2. I was getting "strange" output on the second filter, so I made the second filter equal to the first filter and to my surprise, the output was not equal on the scope.

Scope version information:



CH1 @ 1V per division and CH2 @ 1V per division triggering off CH3. Math A-B is on at 1V per division.



CH1 @ 2V per division and CH2 @ 2V per division triggering off CH3.



CH1 @ 500mV per division and CH2 @ 500mV per division triggering off CH3. I forgot to turn the math function on here but the A-B operation reflects the difference between to the two wave forms.



Note the difference in wave form. The probes are physically connected to the same point in the circuit:



I first suspected probe compensation so I checked both probes:

CH1 compensation:



CH2 compensation:



CH1 & CH2 & CH3 compensation side by side @ 1V per division:




CH1 & CH2 compensation side by side @ 500mV per division:



Note the difference in amplitude of the square wave. I tried two different probes connected to channel 2 but the same issue was present. I tried the self calibration feature of the scope but again the problem persisted. Just to be sure, I then tried switching CH2 with CH3 and the results were correct.

CH1 @ 1V per division and CH3 @ 1V per division triggering off CH2.



CH1 @ 500mV per division and CH3 @ 500mV per division triggering off CH2.



This issue only happens when switching from 1V per division to 500mV per division and lower on CH2. I can't decide whether this is a hardware or software bug since there are no relay clicks when you change from 1V to 500mV per division. Any ideas on what the issue could be?

[Fungus]

This issue only happens when switching from 1V per division to 500mV per division and lower on CH2. I can't decide whether this is a hardware or software bug since there are no relay clicks when you change from 1V to 500mV per division. Any ideas on what the issue could be?

Might be something to do with the fact that the DS1054Z hardware doesn't support 500mV per division. You're not supposed to enable that 'feature'.

[bitshift]

Might be something to do with the fact that the DS1054Z hardware doesn't support 500mV per division. You're not supposed to enable that 'feature'.

I believe that is 500uV not 500mV?

[Fungus]

Might be something to do with the fact that the DS1054Z hardware doesn't support 500mV per division. You're not supposed to enable that 'feature'.

I believe that is 500uV not 500mV?

Oops. 

[Fungus]

Channel 2 is AC coupled so I expect to see a capacitor charge/discharge waveform like that. What is channel 1 set to?

[bitshift]

Both AC coupled. I didn't mention it but on both AC and DC coupling the result is the same.

[Fungus]

What happens if you turn the 20MHz bandwidth limiter off?

[Mark]

What happens when you connect all the scope ground leads to ground and repeat the test? 

[bitshift]

What happens if you turn the 20MHz bandwidth limiter off?

The same response is acquired. This is with no bandwidth limit and both channels DC coupled.

[bitshift]

What happens when you connect all the scope ground leads to ground and repeat the test?

No difference

[Orange]

Interesting puzzle.

It could be a faulty CH2 on your scope. It could also be a faulty CH2 probe.

Have you tried swapping the CH1 <> CH2 probe at the scope BNC connector ?

[bitshift]

Interesting puzzle.

It could be a faulty CH2 on your scope. It could also be a faulty CH2 probe.

Have you tried swapping the CH1 <> CH2 probe at the scope BNC connector ?

I tried swapping the probe from CH2 with CH1 and the again with CH3 but in both cases the problem remained.

[bitshift]

Channel 2 is AC coupled so I expect to see a capacitor charge/discharge waveform like that. What is channel 1 set to?

Just for more context, CH1 and CH2 are the outputs of a filtered PWM signal. I'm trying to visualize the response of two different filters (but in the screenshots it's the same filter) to a step in the PWM duty cycle. CH3 is only a reference output so I can cleanly see when the PWM duty cycle changes. CH3 is not connected to CH1 and CH2 in any way.

[Orange]

Interesting puzzle.

It could be a faulty CH2 on your scope. It could also be a faulty CH2 probe.

Have you tried swapping the CH1 <> CH2 probe at the scope BNC connector ?

I tried swapping the probe from CH2 with CH1 and the again with CH3 but in both cases the problem remained.

I think your scopes CH2 hardware is damaged on the lower range. Probably a compensation C disconnected/open in the front-end.

[Fungus]

I think your scopes CH2 hardware is damaged on the lower range. Probably a compensation C disconnected/open in the front-end.

I agree.

It's a capacitance curve, something has a lot more capacitance than it's supposed to have. If it's only channel 2 then channel 2 is broken.

Do a factory reset to make sure it's not some obscure configuration. If it's still there then send it back.

[bitshift]

Did a factory reset and the issue persists. I guess its time to test if Rigol will honor the warranty with the hacked firmware.

Thank you all for your help 

[alsetalokin4017]

Remove the "hack" before you send it back !!  (And just for completeness sake, retest for the problem after you've removed the hack.)

:SYSTem:OPTion:UNINSTall

sent to the scope using your favorite method of sending SCPI commands.

[Fungus]

Did a factory reset and the issue persists. I guess its time to test if Rigol will honor the warranty with the hacked firmware.

You can remove the hack... use an ethernet cable and install the Rigol software on the CD.

There's a command-line thingy in there somewhere. Open it and do ":SYSTem:OPTion:UNINSTall"

[bitshift]

Thanks all 

Just did a reset so back to a 1054Z. I tested again and the problem is still there. I will get hold of the dealer tomorrow and arrange for the unit to get sent back.

[bitshift]

I took the scope to the dealer and they checked it out and let me know that Rigol said they would give me a brand new unit. I received a unit today but the box was a little beat up/opened already. I'm skeptical that it is indeed a new unit. Is there a way I can lookup when the unit was manufactured or released based on the serial number?

[bitshift]

Further to the above, would it be wrong for me to insist on a sealed unit? I haven't had to time to inspect the instrument yet. I will do so thoroughly once I get home from work.

[Fungus]

Further to the above, would it be wrong for me to insist on a sealed unit?

Rigol is under no obligation so long as it's in at least as good condition as the one you returned.

(and the alternative might be to wait months for your personal 'scope to be fixed).

[bitshift]

Understood, thanks. My inner skeptic took over for a while Busy testing the scope now and all seems perfect.

[fivefish]

Good to hear.

I guess its time to test if Rigol will honor the warranty with the hacked firmware.

And to answer your question, YES. 

When you get back the scope, the hack isn't there though, but you'll have the latest firmware.
Just re-apply the hack.