Siglent SDS1202X-E compensation issue (and other manufacturers ?)

[engreric]

I was wondering if anyone knew if this issue also applies to the sds1104x-e. I'm seeing a similar issue between 1v and 500mv div. My serial is SDSMMDBQ2R and I also purchased it from amazon. I'm thinking of returning it and buying one from saelig to hopefully get one from a different, maybe newer batch. With this unit from amazon, I also had one bad startup up on the first day where it hung on the siglent boot screen. what do you guys think?
Thanks,
Eric

[Performa01]

Welcome to the Forum!

The SDS1004X-E (i.e. the 4-channel units) have never had that significant impedance change with different attenuator ranges (aka "Compensation Issue").

So it appears like you just got a lemon and sending your current unit back and getting a replacement (even from somewhere else) sounds like a good plan to me.

[rf-loop]

I was wondering if anyone knew if this issue also applies to the sds1104x-e. I'm seeing a similar issue between 1v and 500mv div.

No.

And more, between 500mV/div and 1V/div it do not switch between front end attenuator circuits.

There is three bands. (coarse steps)
1. 500uV - 118mV/div (500uV, 1, 2, 5, 10, 20, 50, 100mV/div)
2. 120mV - 1.18V/div (200mV, 500mV, 1V/div)
3. 1.20 - 10V/div  (2, 5, 10V/div)

Old problen in some SDS1202X-E was between band 1 and 2. (using fine adjust it can hear BW change (relay click) just between 118mV/div and 120mV/div)

Also I have tested several individual SDS1004X-E models and never seen any this kind of compensation issue what was in 1202X-E model when there was missing capacitors.

In some OLD manufacturing lot(s) there was hardware mistake (missing (3.9pF 500V) capacitors) in SDS1202X-E. 

If you think there is problem in your individual scope please show images and perfect explanation how it was tested.

ETA:

I also had one bad startup up on the first day where it hung on the siglent boot screen. what do you guys think?

It feels bit strange. If you find any problem it is possible (but rare) that you have get "monday morning" unit.
 

[Performa01]

Old problen in some SDS1202X-E was between band 1 and 2. (using fine adjust it can hear BW change (relay click) just between 118mV/div and 120mV/div)

Doh - How could I miss that?

Of course there cannot be any change between 500mV/div and 1V/div. So this must be something completely different.

If you think there is problem in your individual scope please show images and perfect explanation how it was tested.

Yes, indeed. I think we should rule out other error sources before you send your unit back just to find out that the next one behaves exactly the same...

[exe]

I was wondering if anyone knew if this issue also applies to the sds1104x-e. I'm seeing a similar issue between 1v and 500mv div. My serial is SDSMMDBQ2R and I also purchased it from amazon. I'm thinking of returning it and buying one from saelig to hopefully get one from a different, maybe newer batch. With this unit from amazon, I also had one bad startup up on the first day where it hung on the siglent boot screen. what do you guys think?
Thanks,
Eric

Can be a calibration issue. I'd first do autocal carefully following the manual (e.g., warm up for at least 30mins, detach leads, etc). I think mr. tautech can add more.

Also be sure you do measurements when probes are in 10x position.

[engreric]

Regardless, it failed to boot once on the first day I had it. That's disconcerning and it's going back while I still have the 30 day return.

[tautech]

Regardless, it failed to boot once on the first day I had it. That's disconcerning and it's going back while I still have the 30 day return.

Probably not necessary but do what gives you confidence in the product.
What you have from buying from Amazon is a seller that does not check products before dispatch or install latest firmware that might have fixed any rare/random boot failure.
There has been new firmware and you didn't say which version is in your unit.
The latest for 4ch X-E's is 6.1.26.
https://www.siglentamerica.com/service-and-support/firmware-software/digital-oscilloscopes/#sds1000x-e-series


The compensation issue is long remedied unless you have a NOS unit and rf-loop has described how to check yours.
It must be done with probe set to 10x and input menu probe setting set to 10x too.
There are several screenshots earlier in this thread that show the issue clearly for you to compare against.

Just to confirm what others have said; there is no input compensation issues with SDS1*04X-E DSO's.

[engreric]

Performa01: Thank you for the welcome!, This is my first time writing on the forum though I've been reading and searching on here for a while. I'm a computer engineer trying to build back up my analog skills and this forum has been the best resource I've seen on line for building up practical skills and is a great expansion to all the youtube videos out there.

exe: thanks for the recommendation of self-cal after 30min of warmup, I neglected to wait enough time to warmup, my digital mind overlooks these things.

tautech: I did the firmware upgrade, it came shipped with 6.1.25 R2 and I just upgraded last night to 6.1.26 and then did a self-cal after waiting 30min.

rf-loop: I really appreciate the thorough details and the explanation of the band changes and the old issue between ch1 and ch2 on the 1202x-e, that's another thing for me to check, though my unit is an 1104x-e

I do have another question, and I really apologize if it seems like I'm hijacking a thread. I should probably just start another thread (let me know if I'm in the wrong). I'm just trying to put my scope through all the paces and see if I can find any issues/quirks before I'm committed to this unit. I just got the software and screenshots working, siglent could do a better job at describing where to get the driver and using the easyscope software. Now I just need to get the ethernet working.

Is it normal to have a 5 to 10mv offset in the AC coupling mode, no probe connected, x10 setting, 5mv vertical, all timebasese?

[rf-loop]

Performa01: Thank you for the welcome!, This is my first time writing on the forum though I've been reading and searching on here for a while. I'm a computer engineer trying to build back up my analog skills and this forum has been the best resource I've seen on line for building up practical skills and is a great expansion to all the youtube videos out there.

exe: thanks for the recommendation of self-cal after 30min of warmup, I neglected to wait enough time to warmup, my digital mind overlooks these things.

tautech: I did the firmware upgrade, it came shipped with 6.1.25 R2 and I just upgraded last night to 6.1.26 and then did a self-cal after waiting 30min.

rf-loop: I really appreciate the thorough details and the explanation of the band changes and the old issue between ch1 and ch2 on the 1202x-e, that's another thing for me to check, though my unit is an 1104x-e

I do have another question, and I really apologize if it seems like I'm hijacking a thread. I should probably just start another thread (let me know if I'm in the wrong). I'm just trying to put my scope through all the paces and see if I can find any issues/quirks before I'm committed to this unit. I just got the software and screenshots working, siglent could do a better job at describing where to get the driver and using the easyscope software. Now I just need to get the ethernet working.

Is it normal to have a 5 to 10mv offset in the AC coupling mode, no probe connected, x10 setting, 5mv vertical, all timebasese?

It is not normal (if all other things are normal) but before it can really say:

Just for sidenote:
Due to 10X setting your scope real offset is 0.5 - 1mV  (1 - 2 division)
Probe multiplier setting in scope input menu do not change anything in front end. Only displayed level numbers change.

Have you done Self Cal after 30min (30min is really minimum)
And it need do so that nothing is connected to input BNC's (Nothing means that really nothing, example if you keep probe connected but probe is not connected to anything, it is error. Input BNC's need be empty and open (exept if you have BNC metal caps without center connection)

Also in Utility menu there is Quick-Cal. Keep it normally ON.


Is it really normal or not, it is more easy to answer if you follow next.

I will ask you do this next - exactly:

1.
Independent of if you have previously done selfcal or not, do it now. Turn your scope on and wait least over 30-45min (without notable environment temperature changes) (waiting  time you can still use it if you like)
After continuously power on least 30-45min disconnect all, if there is anything, from the input terminals so input BNC's  are just empty.

2.
Press Default button.

3.
Start SelfCal and leave it alone running it. During selfcal  let it run alone without touching it (and specially do not touch inputs).

4.
After selfcal. Scope still running.
Check that Quick-Cal is On


5.
Still do not connect anything to input BNC's.
Then without any other changes do next things only.
Push Ch4 selection (channel turns on) and turn Vertical  for  500uV/div. Do not adjust vertical position.  Set BandWidth (BW) to 20M then do same for Ch3, then Ch2 and then Ch1.  (and of course all they are now DC coupled and BW and ) If they are over each others or not just do not care and do not adjust anything)
Turn trigger level to top, 2.25mV.


6.
Do not keep your fingers or anything near inputs (and of course nothing connected)
Save screen image to USB

7.
After 6. then do nothing else but push Ch4 selection and set CH4 coupling to AC, Then Ch3, then Ch2, then Ch1  (other things just stay still like step 5 and 6.)
Save screen image to USB

Show these both images.

[engreric]

rf-loop, Thanks so much for the help. I followed your directions exactly. Here are the results.

[rf-loop]

Weird.

How your t/div is 200us/div if you have done position 2. "push default"
1us/div is  factory default. 

Have you stored your own defaults for "Default" button?  (yes it can do)
If you have not done this, then good question is: Who have done it?

When you first time turn it on and look "system status" for check what FW there is,  do you remember what was Startup Times counter. Normally after factory it is 0 and when you star it there is then just 1.


For return Default button (and other things) back to factory state
(under display these rectangle buttons can think F1 - F6)
Push "Save/Recall" button
Then F2 (Recall)
then  F1 (Type Setups)  and select "Factory Default"
Then F5 (Press To Recall)


After then your scope is in known state and Default button have factory defaults.



But then this main issue. Specially this amount of Ch1 offset change between input coupling  DC and AC (there can be some because if there is example unsymmetric noise, trace is in AC center what may differe from zero. (I mean something like this: Example if you connect 10% +1V / 90% -1V duty square wave and coupling  AC you see center is near -1V )
I have never seen this and I have tested many SDS1kX-E scopes. Now if there is some



If it is in my hands I do bit more some tests for analyze bit more  what is going on there.
Also you told that first start up fails (Stuck with "Siglent" bootup screen)  .

Together with these things
Perhaps it is not wise to loose more time with analyzing it, specially if you do not have experience and equipment for more deep tests.

I will recommend you return it back to this carton box broker "Amazon" whose only expertise is money reception and packet delivery.

Attached image show how it is typically (fresh selfcal and inputs open) Of course individual scopes may differ some but imho not like your Ch1
DC Coupled
AC Coupled

Note that when QuickCal is ON it traces may some times adjust offset and it may move trace some amount.
True offset hardware step is bit rough because sure there is not over 20bit DAC for it.
Offset full range is 4V and if (just example) step is one 500uV div it mean 13bit and sure there is more resolution but not for 20uV resolution what is vertical position adjustment resolution on the screen using position adjustment knob.
Quite normal is that with open input and 500uV/div  trace average may differ around 200 - 500uV. Least I think it is normal. But you have clearly more. Specially Ch1

[Performa01]

The main difference between DC/AC is a capacitor in series with the input of the low frequency path of the input buffer, which typically is some OpAmp.

Ideally, the voltage across this capacitor should be zero, but in reality there will always build up some charge because of funny physical effects and - depending on the actual circuit - also through the input current of the OpAmp (which should be very low, because it's usually JFET inputs).

Because of this, we will always see some minor offset shift as can also be seen in the screenshots of rf-loop.

Engreric, your screenshots clearly show an unexpected behavior, hence this unit should be returned. This sample either has bad capacitors or OpAmps with excessive input current.

[Roger Need]

There is an option to turn Quick-Cal on or off in the Utility menu.  There is no mention of this in the SDS1202X-E manual.  Does anyone know exactly what this feature does??

[tautech]

There is an option to turn Quick-Cal on or off in the Utility menu.  There is no mention of this in the SDS1202X-E manual.  Does anyone know exactly what this feature does??

Quick-Cal or should we say ‘hidden cal’ was an undocumented feature that predominantly ran while the DSO was warming up.
Member rf-loop bought it to our attention a few months back and asked Siglent for a way to control its behaviour and subsequently they added control for it in a recent firmware release.
So this was applied to 2 and 4 ch models.
See here:
https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/msg1852547/#msg1852547

[Rerouter]

I should ask as its at least somewhat on topic, For these scopes, it seems only the DC calibration is stored?

for the 500uV-100mV attenuator range, I have a slightly misbehaving channel 3, DC is dead on, AC is 1.2mV high, and its been consistent over a few days of testing.

In the same boat, 3 of the channels are 300uV under when set to ground coupling. I can understand this is beyond the specifications of the scope, just curious why calibrating AC coupling for the lower ranges wasn't a thing?

[cjm]

I see a similar de-compensation issue with my SDS1202X (not the -E version) and wondered if anyone had similar before and done rework on the front-end compensation components of this older model?

The specific problem in my SDS1202X is essentially a shelving of the response, a drop of ~0.6db between ~300kHz and ~3MHz. The response is acceptably flat from DC to ~280kHz and again flat from ~3.3MHz to 100MHz. Testing with a signal generator sweep shows what looks like a simple single pole shelf response in the affected range, so I'm guessing replacing a single cap or resistor could fix the issue - (if I knew which one!). Btw. the problem is very similar on both channels and at all vertical gains.

Below capture from the scope is an example of how the issue appears in the time domain - a rather droopy looking rendition of a 1 MHz square wave. A 1kHz or a >4 MHz square wave looks fine, as expected given the range of affected frequencies.

Test conditions for below plot: Scope warmed up for 1 hour. Two different signal generators used on channel 1 and channel 2. Signal generators first tested with two other scopes with no issues – 1 MHz square wave seen as "square". 50 Ohm coax connections with both channels properly 50 Ohm terminated at scope.

[tautech]

I see a similar de-compensation issue with my SDS1202X (not the -E version) and wondered if anyone had similar before and done rework on the front-end compensation components of this older model?

The specific problem in my SDS1202X is essentially a shelving of the response, a drop of ~0.6db between ~300kHz and ~3MHz. The response is acceptably flat from DC to ~280kHz and again flat from ~3.3MHz to 100MHz. Testing with a signal generator sweep shows what looks like a simple single pole shelf response in the affected range, so I'm guessing replacing a single cap or resistor could fix the issue - (if I knew which one!). Btw. the problem is very similar on both channels and at all vertical gains.

Below capture from the scope is an example of how the issue appears in the time domain - a rather droopy looking rendition of a 1 MHz square wave. A 1kHz or a >4 MHz square wave looks fine, as expected given the range of affected frequencies.

Test conditions for below plot: Scope warmed up for 1 hour. Two different signal generators used on channel 1 and channel 2. Signal generators first tested with two other scopes with no issues – 1 MHz square wave seen as "square". 50 Ohm coax connections with both channels properly 50 Ohm terminated at scope.

I can't say for sure if the 1202X-E input stage alignment instructions attached in Reply #57 are also correct for your SDS1202X but for the 1M inputs they should be close.
You can run with those and see what the result is without even dismantling the scope.

You'll need 1.2Vp-p and 12Vp-p square wave at 10 KHz.

Let us know how they look based on Siglent's 1202X-E alignment instructions and we can get further info from the factory if required.

[cjm]

Thanks tautech. The 10kHz signals (attached) look ok. The problem area on my SDS1202X is more in the 300kHz-3MHz range, so maybe different to the SDS1202X-E issue.

[tautech]

Thanks tautech. The 10kHz signals (attached) look ok.

OK they are but certainly not perfect, that is when we were doing the X-E rework your displayed result would not have been acceptable.

The problem area on my SDS1202X is more in the 300kHz-3MHz range, so maybe different to the SDS1202X-E issue.

Maybe but I suspect not so much, in that the range where you see a problem could be some relic of a drifted front end adjustment. However I'll drop tech support an email and see if they can offer advice or hopefully a remedy.

[cjm]

Many thanks tautech. Very much appreciated.