Weird behavior of my oscilloscope

[Noise0]

Hi
My oscilloscope shift the signal when i change the vertical scale.
Please check this short  video that show the issue:

Is my oscilloscope broken? or is there any reason for this behavior ?

thanx a lot

[rstofer]

I think your signal has a DC offset.  Try turning on AC Coupling (I'm assuming you have it set for DC Coupling).  The triangle in the lower left corner shows where 0V is positioned.

[pigrew]

I've seen similar behavior on KeySight DSO-X 3000 scopes.

What's happening is that the input amplifiers are saturating, and they take time to recover. For accurate measurements on scopes, I've found the entire signal must be within the vertical range. (Digital zoom doesn't have this problem, but won't give you any extra resolution.)

You might be able to do an experiment where you watch the min(V) change as the frequency of your signal (preferably a square wave) changes (so that you can determine how long it takes for the scope's front-end to recover from out-of-range inputs...)

[Noise0]

I've seen similar behavior on KeySight DSO-X 3000 scopes.

What's happening is that the input amplifiers are saturating, and they take time to recover. For accurate measurements on scopes, I've found the entire signal must be within the vertical range. (Digital zoom doesn't have this problem, but won't give you any extra resolution.)

You might be able to do an experiment where you watch the min(V) change as the frequency of your signal (preferably a square wave) changes (so that you can determine how long it takes for the scope's front-end to recover from out-of-range inputs...)

Thank you for your help!

I have done this experiment:


The first pic shows my signal: 1Hz,  8vpp,  640mv offset



Then I zoom in and the second image shows what happens


It takes 20ms to recover form that "saturation" ...
It seems that rocking the signal frequency dont chance very much those 20ms to recovery..

So is this a normal behavior for a Rigol DS1054?
At least can I assume that my oscilloscope is fine?
thx

[TimFox]

I am not expert on the front-end electronics of modern DSOs, but in traditional CROs there is a "balance" control on the input differential amplifier. 
This is to be adjusted so that with zero input voltage (DC coupled), changing the scale factor (V/div) does not shift the horizontal line.

[JDubU]

Have you tried doing a self calibration (Utility -> Self-Cal)?

[Noise0]

Have you tried doing a self calibration (Utility -> Self-Cal)?

thx
im doing it just now.
It seems it takes a long time to complete..
let's see if things improve

[bdunham7]

So is this a normal behavior for a Rigol DS1054?
At least can I assume that my oscilloscope is fine?

Yes.  The simplest solution is to keep your signal on the screen.  If you must run it off for some reason, consult the manual to see how far you can go before you overload the input amplifier and get that delayed recovery.

[pigrew]

So is this a normal behavior for a Rigol DS1054?
At least can I assume that my oscilloscope is fine?
thx

It's normal. There is nothing to indicate your oscilloscope is damaged.

[Noise0]

It's not that uncommon for me to want to examine the underside of a large signal in detail. So today i learned that the only way is to freeze the signal and use digital zoom. I'm just curious: is the reason for this limitation due to the fact that my beloved Rigol is an entry level oscilloscope or is it generally something that cannot be done with digital oscilloscopes? Is there something wrong with me or is this actually a serious limitation?

[bdunham7]

It's not that uncommon for me to want to examine the underside of a large signal in detail. So today i learned that the only way is to freeze the signal and use digital zoom. I'm just curious: is the reason for this limitation due to the fact that my beloved Rigol is an entry level oscilloscope or is it generally something that cannot be done with digital oscilloscopes? Is there something wrong with me or is this actually a serious limitation?

There has been discussion about this issue before where I pointed out that one shortcoming of modern, entry-level DSOs is this overload recovery.  Apparently quite a few non-entry-level DSOs also behave this way, but some do not.  Here is the previous thread:

https://www.eevblog.com/forum/testgear/siglent-sds1104x-e-anomaly-on-what-should-be-a-simple-task/msg3450928/#msg3450928

Note that the expen$ive Tek TPS2024 as well as the ancient Tek 2221 and Fluke Scopemeter all have no issue recovering quickly from an overload.

[RoGeorge]

First trace setup:
- generator is a DG4102 (upgraded to DG4202), 8V max, 640mV min, 1Hz, square
- oscilloscope DS1054z (upgraded to DS1104z)
- probe at 10x, calibrated, tip is plugged straight into the generator's BNC, without any retractable hook or alligator wire

Second trace setup:
- generator is a DG4102 (upgraded to DG4202), 1.6V max (8V/2 because of 50 ohms terminator, x 10 because of direct cable), 128mV (640mV/10x2) min, 1Hz, square
- oscilloscope DS1054z (upgraded to DS1104z)
- direct 50 ohms BNC cable terminated with 50 ohms near the oscilloscope connector

It seems the culprit is the oscilloscope, not the generator, not the probe.

[Noise0]

Thanks to all, I always find useful help on this forum 

[David Hess]

It's not that uncommon for me to want to examine the underside of a large signal in detail. So today i learned that the only way is to freeze the signal and use digital zoom. I'm just curious: is the reason for this limitation due to the fact that my beloved Rigol is an entry level oscilloscope or is it generally something that cannot be done with digital oscilloscopes? Is there something wrong with me or is this actually a serious limitation?

Some oscilloscope inputs are designed to gracefully handle overload, but most are not.  I do not think it has anything to do with price, and most manufacturers do not guarantee performance or behavior.

[Doctorandus_P]

Also:

Your scope switches between showing values of 720mV, 800mV, 840mV 880mV. that shows the resolution of the measurement is very low. Much lower then you'd expect and I wonder what's happening there.