**Solved** New Rigol DS1054z - First Oscilloscope Is this normal?

[tech5940]

Hello all, so I just received my new scope this week. Now this is my first scope I have ever owned/used but I've been playing around with it and seem to be getting the hang of some things. I've been able to get SPI data to trigger and decode without issue.  One weird thing I've noticed is when I try and view the scopes own sample waveform (probe and ground clip hooked up to the terminals beside the BNC inputs) it seems to jump around all over the place. I figured it was just some settings I had fooled around with,so I pressed "Storage" and set back to "default", then hooked up the probes on CH1 and pressed auto. 90% of the time it doesn't trigger properly and jumps all over the place. The auto function sets it to 500mV, If I change this to 1V it usually stabilizes. I turned on the hardware frequency counter and its all over the place as well, anywhere from 1kHz to 1.4kHz or so. The second I change it to 1V everything settles down to 1kHz, sometimes 999.999Hz but the waveform stays triggered.

Am I missing something here?

Just to save  a lot of questions I've tried the following:

I've tried all 4 probes
I've tried all 4 channels
I've performed the scopes Auto calibrate (the 20min procedure)
Trigger defaults after "Auto" are:
Type:Edge
Source:CH1
Slope:Rising edge
Sweep:Auto
Coupling:DC
Holdoff:16.0ns
Noise reject:OFF

I've wiggled the connection / taken the probe hook off and pressed the sharp point to the terminal to see if it was just a bad connection.
When I'm using 10X probe setting on the probe and the scope the waveform won't trigger properly after "Auto" at 500mV
When I change to 1X on the probe/scope and press auto its stable at 500mV
Turning on noise reduction USUALLY stabilizes the triggering, but AUTO defaults this to off so I'm assuming it should work without it.


I'f I'm missing something really obvious here go easy on me! Like I said its my first scope.

Thanks

I should add:
Firmware is latest: 00.04.02.SP4
Scope has NOT been "upgraded" to 100Mhz

[kwass]

What version of the firmware is installed?  If it's not the latest and you have the trigger coupling set to AC you'll get the results you describe.

Set the trigger coupling (not the channel coupling) to DC to see if this solves the problem.  If so, upgrade to version 00.04.02.SP4 and you can use AC coupling too.

[tech5940]

What version of the firmware is installed?  If it's not the latest and you have the trigger coupling set to AC you'll get the results you describe.

Set the trigger coupling (not the channel coupling) to DC to see if this solves the problem.  If so, upgrade to version 00.04.02.SP4 and you can use AC coupling too.

Hey, thanks for the reply,

Firmware is the latest 00.04.02.SP4

Trigger coupling was set to DC, I just tried AC and have the same problem. HFR seems to make it stable though but I don't  think I should have to use HFR just to stabilize the scopes test signal.

[AndyC_772]

When it's in the 'unstable' condition, try speeding up the horizontal time base until you can clearly see what shape the edges actually are. It's possible that, around the point halfway up (or down) there's a slight step or kink.

If that coincides with the trigger threshold, it's possible that even on a rising edge, the signal does occasionally cross the trigger threshold in a negative-going direction, which would trigger the scope. Such edges are referred to as 'non-monotonic', and they're a problem in applications where signal integrity is important.

[miguelvp]

When you calibrated the scope did you have all the probes disconnected? If not you should recalibrate with all the BNCs unattached from the probes.

At 1KHz test signal I don't see how it could have problems triggering.

[AndyC_772]

A high speed edge is a high speed edge, regardless of whether there's 1000 of them per second, or many millions.

It's an important distinction which even many professional engineers don't really appreciate. If an edge has a rise time which is short compared to the length of transmission line to which it's connected, then signal integrity must be considered. The fact that these edges are relatively infrequent makes no difference at all.

[miguelvp]

True, but the rise time on my DS2072 test signal is about 2.9us maybe it's way faster on the DS1054z.

[tech5940]

When it's in the 'unstable' condition, try speeding up the horizontal time base until you can clearly see what shape the edges actually are. It's possible that, around the point halfway up (or down) there's a slight step or kink.

If that coincides with the trigger threshold, it's possible that even on a rising edge, the signal does occasionally cross the trigger threshold in a negative-going direction, which would trigger the scope. Such edges are referred to as 'non-monotonic', and they're a problem in applications where signal integrity is important.

The first picture I attached is what I think you were looking for. If not let me know, the second picture I went the other way with the time base in case anyway asks for that.  I don't see any kinks or steps on the rise.

[tech5940]

When you calibrated the scope did you have all the probes disconnected? If not you should recalibrate with all the BNCs unattached from the probes.

At 1KHz test signal I don't see how it could have problems triggering.

Yes all probes were disconnected.

[Karel]

Try to set the trigger type (sweep) to "normal" instead of "auto".

[tech5940]

Try to set the trigger type (sweep) to "normal" instead of "auto".

It does the same thing.

What I'm really trying to determine here is if this is normal for a new scope to be unable to trigger properly off its own test signal after pressing "auto". I know the "auto" setting on a scope if far from perfect but if it was ever supposed to work correctly I'd assume it would be with its own test signal.

Also when I turn on the hardware frequency counter, it's not a stable 1kHz jumps all over as high as 1.5 kHz which seems odd to me.

My question is less about how do I get the signal to trigger and more about is it normal for it not to trigger in default settings and "auto" tuned. 

Thanks for the suggestion.

[miguelvp]

That signal looks very noisy when it doesn't trigger.

Can to test it in someone else's place?

[AndyC_772]

I agree it looks noisy, though I don't see anything much wrong with the shape of the edge itself. Maybe that's all it is?

It would be nice to know where all that noise is coming from, though. Not quite sure how you might do that without other scopes or signal sources to compare to.

[Stupid Beard]

For what it's worth: I turned my scope on, reset it to defaults, hooked up ch1 probe to cal, hit the auto button, selected no options, turned on the frequency counter and this is what I get. Pretty much what you are expecting, i think. The rise time is ~2.9uS, same as what miguelvp said.



Mine is from January, has been upgraded to 100mhz and is the same firmware version as yours. I have previously run the cal thing after the scope had been on for about an hour, but I don't remember when so it was probably a month or more ago.

This is probably not that useful, but figured I'd post it anyway just in case.

[AG6QR]

What I'm really trying to determine here is if this is normal for a new scope to be unable to trigger properly off its own test signal after pressing "auto". I know the "auto" setting on a scope if far from perfect but if it was ever supposed to work correctly I'd assume it would be with its own test signal.

No, it is not normal.  It does not necessarily mean the scope is broken, but something is wrong, either in the test signal, the probing, the scope, or perhaps the nearby RF environment.

Compare your pictures to Stupid Beard's.  Notice especially that your traces show much "fatter" width on the flat portions of the waveforms.  That indicates that you've got some high frequency noise superimposed on the 1 kHz test signal.  You can see the noise better when you used the faster timebase.  Just eyeballing the fatness of your traces, it looks like around 1/4 volt or so.   

The task becomes figuring out where the noise is coming from.

You might try setting the trigger to "single" mode, and running once, at a very fast timebase, fast enough to clearly determine the frequency of the noise.  Does the noise even have a regular and consistent frequency?  Are you very near a powerful radio transmitter?  Is the frequency of the noise close to the frequency of the radio transmitter?

It also might be interesting to try probing a known smooth source, like a battery.  Do you see that same sort of fat trace with noise?

Another experiment might be to take the scope to a different location.  Maybe try a friend's house, or at least a spot as far away from your normal workbench as practical in your home.  If the noise disappears when you move the scope away, the noise could be coming from other equipment in your lab.

The noise could be coming from something inside the scope, in which case you might have an issue to be dealt with by sending it in for warranty repair.

[miguelvp]

The noise could come from your refrigerator and affecting the whole house, or the air con, or anything that cycles and goes on then off. That's why I also suggested taking it to someone else's place.

That might be affecting the counter. When it's quiet and the trigger locks, press the clear button and see if the frequency counter tells the truth. If you have a DMM with frequency measurement then measure the signal to see if it's off as well.

[tech5940]

For what it's worth: I turned my scope on, reset it to defaults, hooked up ch1 probe to cal, hit the auto button, selected no options, turned on the frequency counter and this is what I get. Pretty much what you are expecting, i think. The rise time is ~2.9uS, same as what miguelvp said.



Mine is from January, has been upgraded to 100mhz and is the same firmware version as yours. I have previously run the cal thing after the scope had been on for about an hour, but I don't remember when so it was probably a month or more ago.

This is probably not that useful, but figured I'd post it anyway just in case.

Thanks for posting a shot of yours. Definitely helpful as I now see the noise everyone is talking about.


Sent from my iPhone

[tech5940]

What I'm really trying to determine here is if this is normal for a new scope to be unable to trigger properly off its own test signal after pressing "auto". I know the "auto" setting on a scope if far from perfect but if it was ever supposed to work correctly I'd assume it would be with its own test signal.

No, it is not normal.  It does not necessarily mean the scope is broken, but something is wrong, either in the test signal, the probing, the scope, or perhaps the nearby RF environment.

Compare your pictures to Stupid Beard's.  Notice especially that your traces show much "fatter" width on the flat portions of the waveforms.  That indicates that you've got some high frequency noise superimposed on the 1 kHz test signal.  You can see the noise better when you used the faster timebase.  Just eyeballing the fatness of your traces, it looks like around 1/4 volt or so.   

The task becomes figuring out where the noise is coming from.

You might try setting the trigger to "single" mode, and running once, at a very fast timebase, fast enough to clearly determine the frequency of the noise.  Does the noise even have a regular and consistent frequency?  Are you very near a powerful radio transmitter?  Is the frequency of the noise close to the frequency of the radio transmitter?

It also might be interesting to try probing a known smooth source, like a battery.  Do you see that same sort of fat trace with noise?

Another experiment might be to take the scope to a different location.  Maybe try a friend's house, or at least a spot as far away from your normal workbench as practical in your home.  If the noise disappears when you move the scope away, the noise could be coming from other equipment in your lab.

The noise could be coming from something inside the scope, in which case you might have an issue to be dealt with by sending it in for warranty repair.

I'm stuck at work for the next two days but thanks for the great suggestions. I'll see if I can track down the source of noise on Wednesday and report back.


Sent from my iPhone

[tech5940]

The noise could come from your refrigerator and affecting the whole house, or the air con, or anything that cycles and goes on then off. That's why I also suggested taking it to someone else's place.

That might be affecting the counter. When it's quiet and the trigger locks, press the clear button and see if the frequency counter tells the truth. If you have a DMM with frequency measurement then measure the signal to see if it's off as well.

Maybe I'll start with turning stuff off around the house and see it the noise settles down. And yes I have a DMM which measures frequency so that's a great suggestion I'll try that and see what it shows.


Sent from my iPhone

[tech5940]

The noise could come from your refrigerator and affecting the whole house, or the air con, or anything that cycles and goes on then off. That's why I also suggested taking it to someone else's place.

That might be affecting the counter. When it's quiet and the trigger locks, press the clear button and see if the frequency counter tells the truth. If you have a DMM with frequency measurement then measure the signal to see if it's off as well.

Maybe I'll start with turning stuff off around the house and see it the noise settles down. And yes I have a DMM which measures frequency so that's a great suggestion I'll try that and see what it shows.


Sent from my iPhone

So I hooked up my multi-meter to the scopes test output, and it reads a steady 999.6Hz even when the scope is not triggering properly. Makes me wonder if that frequency counter on the scope is actually hardware based because its currently reporting that same signal as around 3.5kHz.  But at the bottom of the screen the other frequency measurement shows dead on 1.00kHz.

[dadler]

The overlay frequency counter on the 1054z is hardware based. The one at the bottom of the screen is software and just computes the frequency from the on-screen waveform data.

[AG6QR]

So I hooked up my multi-meter to the scopes test output, and it reads a steady 999.6Hz even when the scope is not triggering properly. Makes me wonder if that frequency counter on the scope is actually hardware based because its currently reporting that same signal as around 3.5kHz.  But at the bottom of the screen the other frequency measurement shows dead on 1.00kHz.

Just as background, the scope's frequency counter is a hardware-based counter that counts how often the trigger circuit is triggered.  In my experience, it's a very accurate measure of how frequently the trigger is happening, but of course, whether that represents an accurate measurement of the input waveform depends on how triggering is set up.

In your case, we already knew that the triggering is flaky, probably because of some noise superimposed on the signal.  If you use a low pass filter to clean up the triggering, I'm guessing the frequency counter will look better.

The other frequency measurement at the bottom of the screen is based on measuring the width of the waveforms on the display.  It's not particularly precise.

[tech5940]

The overlay frequency counter on the 1054z is hardware based. The one at the bottom of the screen is software and just computes the frequency from the on-screen waveform data.

Well that's what I thought, I was wondering why in my screenshot its showing 3.5kHz when its hooked up to the 1kHz test signal.

[tech5940]

So I hooked up my multi-meter to the scopes test output, and it reads a steady 999.6Hz even when the scope is not triggering properly. Makes me wonder if that frequency counter on the scope is actually hardware based because its currently reporting that same signal as around 3.5kHz.  But at the bottom of the screen the other frequency measurement shows dead on 1.00kHz.

Just as background, the scope's frequency counter is a hardware-based counter that counts how often the trigger circuit is triggered.  In my experience, it's a very accurate measure of how frequently the trigger is happening, but of course, whether that represents an accurate measurement of the input waveform depends on how triggering is set up.

In your case, we already knew that the triggering is flaky, probably because of some noise superimposed on the signal.  If you use a low pass filter to clean up the triggering, I'm guessing the frequency counter will look better.

The other frequency measurement at the bottom of the screen is based on measuring the width of the waveforms on the display.  It's not particularly precise.

Oh ok thanks, that makes sense then that its measuring the trigger frequency and since the triggering is messed up its reporting a false frequency.

Now I have to figure out were all the noise is coming from.

I was just messing around with the scope had the probe tip attached to the test signal's ground lug I was getting around 260mv of noise peak to peak.  While on the same settings I was attaching a small length of coax to the Aux trigger out and the noise goes from about 260mv pp down to 60mv pp.

Anyone ever probe the GROUND connection on their scope? I'm really not sure where to start as far as tracking down the noise. Turning off almost everything (computers/fridge/wifi/heater etc) made no difference.

[AG6QR]

I was just messing around with the scope had the probe tip attached to the test signal's ground lug I was getting around 260mv of noise peak to peak.  While on the same settings I was attaching a small length of coax to the Aux trigger out and the noise goes from about 260mv pp down to 60mv pp.

Anyone ever probe the GROUND connection on their scope? I'm really not sure where to start as far as tracking down the noise. Turning off almost everything (computers/fridge/wifi/heater etc) made no difference.

I don't know what to make of the coax on the aux trigger out phenomenon, but here are some ideas to help figure out whether it's coming from inside the scope or outside.

When you see the 260mV noise, does it have a regular frequency?  Or does it appear random?   Can you trigger on it?  If it seems stable, what is its frequency?  Can you do a FFT display on it and get its spectrum?

Do you have a BNC terminator that you can place across the scope input (probe) jack?.  That should effectively short the scope input out, and produce a steady 0V signal, maybe varying by an occasional pixel or two, but no more.  Does it?

If you connect a probe to the scope, and clip the ground lead of the probe to the tip of the probe, that also should produce a steady 0V, although the loop formed by the ground lead and the tip can pick up some stray RF, if there is any such RF nearby.  What do you see?

If you see noise when using the scope with the ground probe clipped to the tip, what happens if you shrink the loop by using the grounding spring instead of the wire ground lead?  Put the spring on the probe, and touch both the grounding spring and the tip together to a plate of conductive metal.  If you were picking up radio frequency signals in the earlier step when using the longer wire ground lead, this should lessen them significantly.  Does it?