Question about Vertical Accuracy of Siglent SDS 1202X-E

[mio83]

Hello,

I have just bought a Siglent SDS 1202X-E, my first digital oscilloscope.
It looks great and after some testing it seems to work well! The measured bandwidth (-3db) is actually at 250Mhz

However I noticed that when nothing is attached (either to CH1 or CH2) the voltage read is, on average, not 0V but something slighly above or below (it varies over time, kind of randomly). The following is a screenshot taken after doing the calibration procedure, with Acquisition set to average (1024 averages) for reducing noise:


Here you can see the official (datasheet) vertical specification of the SDS 1202X-E.

Is what I am seeing compatible with the "offset accuracy"? Can somebody please help me decode what it is meant by: ±(1%* Offset+1.5%*8*div+500 uV): ≤1 mv/div

thanks!
 

[tautech]

Vertical accuracy changes as the scope warms up and accuracy is specified after ~30 mins run time.
While the scope is getting up to temp you can set Quick Cal to ON where it will automatically correct accuracy as the scope warms up.
Offset accuracy spec is related to measurements while the 0V position of a waveform is off the display.

[bdunham7]

Offset accuracy spec is related to measurements while the 0V position of a waveform is off the display.

Are you sure?  And why would it have to be all the way off the screen?  I would interpret that figure at zero offset (after pressing the Vpos control, for example) on the 500uV range as being (.01 x 0) + (.015 x 4uV) + (500uV) = 500.06uV.  Or just a maximum zero signal offset of about 500uV, which would be pretty reasonable for an absolute measurement.  Also, I'd expect this to be valid with the input shorted, although it probably doesn't make much difference.

[mio83]

I would interpret that figure at zero offset (after pressing the Vpos control, for example) on the 500uV range as being (.01 x 0) + (.015 x 4uV) + (500uV) = 500.06uV.  Or just a maximum zero signal offset of about 500uV, which would be pretty reasonable for an absolute measurement.  Also, I'd expect this to be valid with the input shorted, although it probably doesn't make much difference.

Are you sure your calculation is right (specifically, the .015 x 4uV part)? Shouldn't it be:

(.01 x 0) + (.015 x 8div x 500uV) + (500uV) = 0 + (60 uV ) +  500uV =  560uV ?


Does this mean that values are supposed to be taken +/- 560uV and so the screenshot I posted shows a behaviour within specification?

Vertical accuracy changes as the scope warms up and accuracy is specified after ~30 mins run time.
Offset accuracy spec is related to measurements while the 0V position of a waveform is off the display.

Yes, I did the calibration procedure after 30 minutes. The temperature in my room is quite stable.

@bdunham7 @tautech thank you a lot for your answers.
I am still quite unsure though, if the screenshot I posted earlier shows a behaviour within specs or not.


Thanks!

[tautech]

Does this mean that values are supposed to be taken +/- 560uV and so the screenshot I posted shows a behaviour within specification?

Some of the noise is system noise and a 5ns setting for a 50Hz signal doesn't seem appropriate for any meaningful measurement.
IMO you need use DC triggering for most tasks where you have not only a visual level setting but also adjustment of the trigger level. Pressing the level control will set the trigger to 50% of the waveforms amplitude.

Vertical accuracy changes as the scope warms up and accuracy is specified after ~30 mins run time.
Offset accuracy spec is related to measurements while the 0V position of a waveform is off the display.

Yes, I did the calibration procedure after 30 minutes. The temperature in my room is quite stable.

Yet the Self Cal will not be valid until the scope reaches full operational temp.
This is what Quick-Cal is for. The scope will regularly quickly check its accuracy as it warms up.

[tautech]

More info on verification of accuracy is in the Service manual:
https://int.siglent.com/upload_file/user/SDS1000X-E/SDS1000X-E_ServiceManual_SM0101E-E01A.pdf

[StillTrying]

Yes lower the time base speed to around 1ms to 5ms/div, it allows you to see any 50Hz hum on the DC level.

The probes on X1 and with the GND clipped to the point is good enough for a DC 0V input.

Just 8 or 16 waveform averaging should be enough to get a thin trace even on 1mV and 500uV/div.

To trigger on a flat line, trigger source AC LINE should do, or set the trigger to AUTO and don't worry about the trigger level.

Both of the channels 0V markers should be visible on the left of the screen.

Are the DC levels correct on lower sensitivities 10mV to 200mV/div, they should be.

Is the fan working.

[bdunham7]

Does this mean that values are supposed to be taken +/- 560uV and so the screenshot I posted shows a behaviour within specification?
I am still quite unsure though, if the screenshot I posted earlier shows a behaviour within specs or not.

I miseyeballed a decimal, 560uV looks right.  The service manual Tautech posted doesn't even specify a test for the 500uV range, so I don't know how official an answer you are going to get, but from experience with a variety of scopes, what you have appears to be good.  I would expect the DC level to wander around a few hundred uV or so over time. One of the things that Siglent (at least the SDS1xxx X-x lines) excel at is low front end noise, linearity and accuracy.  The fact that they have a properly functioning 500uV range is itself a standout achievement.

You might get a slightly better result with the input shorted by terminator or if you have none, a 1X probe with the tip grounded.  Triggering on one of the channels in the AUTO mode with the trigger point set near the top of the screen will give you a totally random triggering which would eliminate any AC fields that might be correlated to the AC line trigger.  Slowing the trace down to at least 2ms/div and turning off the AVERAGE function will give you an idea of how noisy the actual signal is so you can see if there is any significant interference on the inputs.  And use the 20MHz BW limit at settings below 2mV/div.

[mio83]

Hello!

Thank everybody for the answers and feedbacks.

I have made four pictures:
They are all made with CH1 only (CH2 is disabled) and nothing attached to CH1 (nor CH2).
This has been done after about 30min of operation and a "calibration" procedure, including "quick cal = ON".

1) Noise with Vertical Res = 0.5mv and Horizontal = 2ms.  The statistics shows that the average is 93uV.
2) Noise with Vertical Res = 0.5mv and Horizontal = 2us.  The statistics shows that the average is 95uV, which agrees with (1) !
3) 1Mhz 0.5mV Rms source, Res = 0.5mv and Horizontal = 2us. The signal is quite noise, the average is 60uV.
4) 1Mhz 0.5mV Rms source, Res = 0.5mv and Horizontal = 2us. Here I have set CH1 bandwidth to 20Mhz and Trigger "Reject HF". The signal is much less noisy. Average is 31uV.


So in all scenarios, I get a deviation from the ideal "0V average" of some 100s of uV (or less).

I am very happy with the scope overall. I just want to be sure if this type of deviation from the ideal 0V is within spec.
From what I understood from your comments/help, it looks like it is working within specifications.

Is the fan working.

yes.


EDIT: The 0.5mV rms signal has been generated by a Siglet SDG 1032X, which is working fine.

[tautech]

Good work. 

Press the blue Print button only briefly to take a USB screenshot. 

[mio83]

Good work. 

Press the blue Print button only briefly to take a USB screenshot. 

The scope actually froze 2 times after pressing Print!
Is this a known issue?

For me what worked best is:
1) push it for a long time, until a message appears saying "click once to take a screenshot"
2) then push briefly.

Do you have any recommendations?
Any other "bugs" or known "issues" of the SDS 1202X-E ?

Thanks!!

[tautech]

Good work. 

Press the blue Print button only briefly to take a USB screenshot. 

The scope actually froze 2 times after pressing Print!
Is this a known issue?

No.

For me what worked best is:
1) push it for a long time, until a message appears saying "click once to take a screenshot"
2) then push briefly.

Another member had issues with Print freezes so I got a new X-E out and took some 40 consecutive Print screenshots without any issues.
There is a USB stick recommendation/requirement of 8 GB max and FAT32 format which especially for uploading files into the scope should be adhered to and if not is possibly responsible for freezes.
Try another USB stick.
A short press of the Print button should always work and capture a screenshot.

Other processes for USB saves are explained here but for most work the Print button is the most convenient:
https://siglentna.com/operating-tip/save-external-usb-sds-series-oscilloscopes/

Do you have any recommendations?

Always use the power ON/OFF button for a nice soft powering down where the DSO will remember previous settings. A hard OFF at the wall does not remember setting and at the next boot some older settings will return.
If you have issues with the scope misbehaving power OFF and ON a couple of times so it can clear its brain and tidy its operating system.

[CharlotteSwiss]

Hello,

I have just bought a Siglent SDS 1202X-E, my first digital oscilloscope.

which firmware mounts your siglent? haven't you noticed any problems when you press the print key to save the image on a usb stick? thank you 

[mio83]

Hello,

I have just bought a Siglent SDS 1202X-E, my first digital oscilloscope.

which firmware mounts your siglent? haven't you noticed any problems when you press the print key to save the image on a usb stick? thank you 

It has the software version 1.3.26.

Yes, it did freeze a couple of times after pushing "PRINT" but it works most of the time.
Do you know what's going on?

thanks

[CharlotteSwiss]

Do you know what's going on?

 
it is a fairly common and known bug, except for Siglent engineers
 

[bdunham7]

Get a USB2 drive of 8GB or less with a flashing operation indicator light.  Do a full format to FAT32 and use the drive only with the scope.  Press the print button once and don't hold it in at all and everything should work.  If it slows down or locks up, see what the flashing light is doing during that time and report that to Siglent.

[CharlotteSwiss]

Get a USB2 drive of 8GB or less with a flashing operation indicator light.  Do a full format to FAT32 and use the drive only with the scope.  Press the print button once and don't hold it in at all and everything should work.  If it slows down or locks up, see what the flashing light is doing during that time and report that to Siglent.

i tried 10 usb-2 keys (from 500mb to 8gb), any fat ... the only thing i didn't take into consideration is maybe to quickly press the print key .. i should try..
I never looked at what the key light did, I'll try

[CharlotteSwiss]

I have just bought a Siglent SDS 1202X-E, my first digital oscilloscope.

if you are interested, here in the forum they helped me to understand the Siglent 1202 in the smallest details, progressively following the lineup of the manual; avoid the first pages ... the best comes later

https://www.eevblog.com/forum/beginners/what-an-oscilloscope-recommended-for-a-woman-passionate-about-electronics/

 

[StillTrying]

Whatever the device, I usually make sure there's no vital data on the memory stick and let the device format it, rather than a PC.

[tautech]

Do you know what's going on?

 
it is a fairly common and known bug, except for Siglent engineers
 

Yet of the dozens I have sold not one customer has contacted me about it.
You can be sure it has been reported to Siglent from your reports of Print freezes but maybe it is so rare and hard to reproduce they have not been able to isolate and fix it.

I have mentioned to you before try a Default and reinstall the firmware however I can't remember if you tried that.

[CharlotteSwiss]

Whatever the device, I usually make sure there's no vital data on the memory stick and let the device format it, rather than a PC.

I am not able to format the key from the siglent 

[CharlotteSwiss]

Yet of the dozens I have sold not one customer has contacted me about it.
You can be sure it has been reported to Siglent from your reports of Print freezes but maybe it is so rare and hard to reproduce they have not been able to isolate and fix it.

I have mentioned to you before try a Default and reinstall the firmware however I can't remember if you tried that.

your buyers likely were luckier than us 

in the meantime I would wait to install the SAME firmware, it is still a risky operation (and I think it does not solve anything); a user had said that with the previous firmware the problem was not there, but the improvements of the latest firmware would be lost ...

[tautech]

Yet of the dozens I have sold not one customer has contacted me about it.
You can be sure it has been reported to Siglent from your reports of Print freezes but maybe it is so rare and hard to reproduce they have not been able to isolate and fix it.

I have mentioned to you before try a Default and reinstall the firmware however I can't remember if you tried that.

your buyers likely were luckier than us 

in the meantime I would wait to install the SAME firmware, it is still a risky operation (and I think it does not solve anything); a user had said that with the previous firmware the problem was not there, but the improvements of the latest firmware would be lost ...

No it is not.

Every year I install dozens of new firmware versions in many different models without any issues.
Today just one in a SDS1104X-U that have just arrived and new firmware was released while they were shipping.
7 more X-U to go but always new equipment we sell has the latest firmware installed.

Download, unpack and study the PDF install instructions and you can't go wrong.
This is the current version for you to reinstall to see if it fixes your Print freeze issue:
https://int.siglent.com/upload_file/zip/firmware/Oscilloscope/SDS1202X-E_1.3.26_EN.zip

[CharlotteSwiss]

thanks tautech
later maybe I'll try, without lighting a few candles first   

[MathWizard]

Hello,

I have just bought a Siglent SDS 1202X-E, my first digital oscilloscope.
It looks great and after some testing it seems to work well! The measured bandwidth (-3db) is actually at 250Mhz

However I noticed that when nothing is attached (either to CH1 or CH2) the voltage read is, on average, not 0V but something slighly above or below (it varies over time, kind of randomly). The following is a screenshot taken after doing the calibration procedure, with Acquisition set to average (1024 averages) for reducing noise: (Attachment Link)


Here (Attachment Link) you can see the official (datasheet) vertical specification of the SDS 1202X-E.

Is what I am seeing compatible with the "offset accuracy"? Can somebody please help me decode what it is meant by: ±(1%* Offset+1.5%*8*div+500 uV): ≤1 mv/div

thanks!

I'm pretty new to Oscopes but when I find my siglent having some non-zero value when I think it should be zero, it's usually just a matter of zooming back the time scale some, and then seeing it's just waveform voltages from a lower frequency signal.

So maybe it's just the way the trigger decides to do that sometimes , and it's showing a part of a waveform a bit off-zero for a particular time base, if I'm putting that into words as I mean.