SDS1104x-e AC Coupling offset

[Steamer]

Hello,

On my just recently aquired SDS1104 I noticed something strange on channel 3 & 4 when using 1X probe and AC coupling.

Up to about a 24 volt signal the AC coupling does it job correctly, but after increasing the voltage further it gets an offset and it goes completely offscreen at about 32volt. This happens only on channel 3 & 4, I've tested channel 1 & 2 up to 36v and they work as I would have expected.

On the pictures the RMS voltage is displayed on the channel 1 measurements.

Greetings,
Emile

[StillTrying]

It's possible there's negative going stuff off screen, try with the trigger point at the center of the screen and X on 2ms or 5ms/div so that there's a lot more than half a mains cycle on screen. you might also need peak-detect. and then panic.

[tautech]

Welcome to the forum Steamer.

Please tell us the firmware version as there's been a couple of recent updates.

Also have you run the Self Cal after a good warm up period ?
This is necessary after installing FW updates to correct any possible channel offset.

[Steamer]

Hello tautech,

The firmware is at version 6.1.25R2 and yes I did run several self cal routines after a good warmup.

But that didn't improve the offset.

Greetings,

Emile

[tautech]

I don't quite understand the use case and why you need AC Trigger coupling and why the Trigger point is off to the left of the display. 
I'm not making excuses just asking why you'd need to use a DSO in this way ?

I'll have a look at some of this behavior later today.

[Steamer]

I just noticed the different behaviour of channel 1 & 2 and 3 & 4 when running some noise measurements on a switching psu.
Tried to setup a poor mans differential probe on channel 1 and 3, but then ran into this problem that I don't understand.

AC Trigger coupling and trigger point where just randomly chosen. They don't influence the problem that I have.
On the 200mV scale and above channel 3 & 4 also work as expected, on the 100mV scale and below it offsets with the higher DC voltages. (>24V)

I'm looking forward to your results.

[FlyingHacker]

Because anyone who is used to a CRO naturally wants the trigger off to the left of the screen.  I still prefer it there.

[tautech]

I just noticed the different behaviour of channel 1 & 2 and 3 & 4 when running some noise measurements on a switching psu.
Tried to setup a poor mans differential probe on channel 1 and 3, but then ran into this problem that I don't understand.

AC Trigger coupling and trigger point where just randomly chosen. They don't influence the problem that I have.
On the 200mV scale and above channel 3 & 4 also work as expected, on the 100mV scale and below it offsets with the higher DC voltages. (>24V)

I'm looking forward to your results.

Through improper use of the instrument I can get some offset on channel 3.....but it's negative !

I don't quite understand the use case and why you need AC Trigger coupling and why the Trigger point is off to the left of the display. 
I'm not making excuses just asking why you'd need to use a DSO in this way ?

I'll have a look at some of this behavior later today.

In summary, garbage in garbage out !
And some user error.

So let's step back and examine what you are doing.
AC Line triggering on ch 1 and examining ch3 but not at line frequency related timebase settings. Why on earth would you even attempt to ? Garbage will always result.

So if we want to properly examine ch3 we'd set the triggering to do so and it's nothing to do with mains line frequency at all. If there is a standing voltage on this rail we'd select channel input AC coupling and Trigger DC coupling and then we can adjust input attenuation to magnify any waveform riding on the rail.

Whenever you set a scope for examination of anything, the trigger source and type is of utmost importance and to get this wrong delivers only garbage to the display. The use case for AC trigger coupling is quite rare and it's a good idea to not even consider using it.
For some study see Daves vid:
https://www.eevblog.com/2014/11/20/eevblog-685-what-is-oscilloscope-ac-trigger-coupling/

Hope this helps.

[Steamer]

So I've changed the setup, DC triggering on channel 3 and AC coupling. Exact same input to channel 1 and channel 3.

I want to see the AC component that is riding on the DC voltage, when the dc voltage is 24 volt channel 3 starts to offset and at about 32 volt channel 3 is offset totally off screen.
Channel 1 functions as I would expect.

First picture is at DC voltage 24 volt and second picture at about 28.24 volt. On 32 volt channel 3 is totally off screen, and the offset >2 volt. Channel 1 and channel 3 both are measuring the same signal.

[tautech]

Thanks, I can't see any restriction in the datasheet that explains why we see this offset behavior so I'll shoot a memo to tech support at the factory for them to have a look at this.
We'll see what they have to say about it.

[rf-loop]

Here is roughly 30V DC (29.4530 - 29.4540 (Keithley2015THD)) from some PS3005 output using Banana-BNC adapter and splitted using BNC T's and connected to all inputs using coaxial cables.

This is how it works when all is ok.

First image all inputs AC coupled.
Ripple observed using 5mV/div
Trig DC rising edge CH1. (also tested trig AC coupling and works also rock solid of course)

Next image exactly same input signal.
All channels DC coupled and now 10V/div.


(Sidenote but not for this OP's case here: For randon readers it need note that Siglent AC coupling HPF corner frequency is extremely low (look Performa01 made high quality depth review) so if signal have some even rare unsymmetric components it of course may produce "offset" like effect and this may happen also so that in worst case this part of signal is not visible on the screen because this happen in analog front end where signal affect all time independent of acquisition periods.)

[Steamer]

Here are my results connecting 4 probe's together, I didn't have BNC T couplers.

First at 25 volt, second at 27 volt, third at 28 volt, and the last at 29 volt dc.

All is not ok, I'm going to recalibrate it again, but I don't think that will help. (Edit: it didn't help) 

[StillTrying]

I think the last few screen shots are enough to prove Ch 3 & 4 are faulty at least on AC coupling. As it doesn't seem to be there on 200mV/Div, is it worth using the fine Y control between 200mV to 100mV to see where it actually starts.

[Steamer]

The problem starts right at the change from 120mv/div to 118mv/div, and that is when you hear a relay clicking.

[StillTrying]

So what was it ?

Leakage on the PCB from the raw input to the <120mV range components is about all I can think of!

[BrianHG]

Have you done a scope self-calibration recently?  Many DSOs have one buried in the utility menu somewhere.

On my old Tektronix, I've done a scope self-calibration once with an active probe in one channel, and, there was a DC offset error on my channel 1 until I redone the scope self-calibration.  The self calibration was originally needed since the DC offset drifted out of alignment by around 0.3v over a year of use.

The DC offset would still show in AC coupled mode since this drift is after the DC probe filter in the scope electronics.

[Steamer]

Calibration was done several times, and that didn't make a difference.

The problem isn't resolved yet. As far as I know Siglent is looking into it, I've opened a support case.

Emile

[Steamer]

They replaced my faulty SDS1104X-E with a new one. (I only had it for about a week)

This one is working as expected. Thank you for the service Batronix/Siglent.

Emile