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

[borjam]

I hope they don't do some kind of software filtering. That would be bad for people with different probes! 

[StillTrying]

Perhaps they could toggle the relay very fast so that it sounds like a buzzer, to act as a reminder to the user to re-compensate the probes every time the relay changes range. 

[exe]

Interesting, I have something similar on my micsig TO1104 and  SDS1202X-E, although not to such a big extend. So, when I tune compensation I check squareness at different V/div _and_ timebase levels. Guess what, what is good for one level may show a bit of overshooting/undershooting on another level (a pixel or two). (I dunno why timebase affects picture, may be it's a rendering issue/feature).

I concluded all scopes to some extend have it. Even when people say "I don't have it" they may actually have it. Just look at this post: https://www.eevblog.com/forum/testgear/siglent-sds1202x-e-compensation-issue/msg1328565/#msg1328565 . Check the second to the bottom picture. See? .

Furthermore, different scopes show square wave... differently. Or at least I think so. I think no scope has absolutely flat response (or is it bad probes?), at least not on scopes I can afford.

Also, it may be related to probes as well. I see different models of probes show different response. BTW, I found siglent probes to be nice and easy to compensate (they worked slightly better than my testec TT-HF-212).

BTW, did you let your scope to warm-up for at least 30mins?

[pantelei4]

Rigol DS2072A everything is fine.

[Hydron]

Just checked my RTB2004, the only thing i could see was a small change in noise and maybe a slightly larger than normal step in attenuation change.

Would have tested a DS1054Z too but I just sold mine - might try some rigol and tek scopes at work next week though.

[TK]

Rigol DS2072A everything is fine.

Picture 5 does not look fine to me

[pantelei4]

Picture 5 does not look fine to me

There are no perfect devices.

[tautech]

We have a resolution to this issue from the factory.

Please do not attempt these modifications unless you have all the tools and rework ability.
It involves adding one 4 pF 1206 (3216 in metric) 500V rated C0G (NP0) capacitor onto unpopulated pads in each channel's input stages and maybe a small adjustment of one or both of the input trimmers.

In a few days when I have the capacitors in hand, member Defpom has indicated he will do a vid for us of the whole procedure.
Please find attached the disassembly, rework and adjustment instructions from the team at Siglent.


Rework video containing adjustments and dis/reassembly guidance:
https://www.eevblog.com/forum/testgear/siglent-sds1202x-e-compensation-issue/msg1456838/#msg1456838

[Orange]

Is there also a fix available for the SDS1102X ?

This issue is even worse as on the E model. See picture below (picture is taken from other user in this forum in a previous posting).
https://www.eevblog.com/forum/testgear/siglent-sds1000x-series-oscilloscopes/msg1273415/#msg1273415

I have the same issue, as per posted picture, and yes my generators are fine, it's the scope.
I would like to do the modification myself.

[borjam]

Thank you.

Now I wonder, will they honor warranty on this issue? It's going to be a problem, I am in Spain and I purchased it from Batronix (Germany).

[tautech]

Is there also a fix available for the SDS1102X ?

I'll do some research with an X I have here and ask the factory if a similar fix is needed.

Thank you.

Now I wonder, will they honor warranty on this issue?

I don't know how. It will take longer to do the paperwork, packaging and despatch than the rework.

One might ask the same of other manufacturers that have this same issue......you can see by the replies there are some ........... and how many other members are reluctant to share their 10x compensation mismatch experience.

It could be more common than we think. Good video topic for Dave.



The X-E is quite simple to get apart although you must engage brain so to not damage anything but the soldering is a little tricky as it's inside the shielded input box. The 4pF cap is paralleled alongside an existing cap, the value of which I should've measured when I had my demo unit apart. Grrrr. Will measure it later.
Then the scope must be partially reassembled and powered on to check the waveform and maybe adjust the trimmers. I think the job can be done in ~40 mins providing all tools are ready and at hand. (challenge for Defpom  )

All that's needed to disassemble X-E is #2 Phillips, medium flat blade screwdriver and 16mm or 5/8" AF Ring spanner or deep reach socket or even an adjustable spanner will undo the BNC nuts. (they're not recessed  )
Four connectors need to be unplugged, display flat cable (locking connector and fabric adhesive tape), front panel ribbon cable (locking connector) fan and PSU.
Undo the 3 BNC nuts and then you have the mainboard free to remove the input shielding cover and solder the caps in. The mainboard is covered in fragile components and need be handled with care, some antistatic sheet foam will be good to avoid any damage.

[borjam]

For now it's not a terrible issue for me, so I will probably pass on it.

[MrW0lf]

Nice to see dealing with issue in fast and decisive manner. If look at the other end of spectrum - some other brand popular scope change signal amplitude by 43% if switch Sinc ON|OFF (100MHz sine in). Also lie about rise times etc. Issues never acknowledged/fixed.

[nctnico]

Nice to see dealing with issue in fast and decisive manner.

I don't think so. Any A-brand would do a recall and fix the problem for free. Having people deal with soldering SMD and handling small connectors themselves is a recipe for dissaster. How is this going to work out when it comes to warranty? The second best option would be to have the local dealers handle it by sending the scopes to a local company which specialises in repairing electronics at a component level (for example a laptop repair service).

[StillTrying]

Is there also a fix available for the SDS1102X ?

This issue is even worse as on the E model. See picture below (picture is taken from other user in this forum in a previous posting).
https://www.eevblog.com/forum/testgear/siglent-sds1000x-series-oscilloscopes/msg1273415/#msg1273415

I have the same issue, as per posted picture, and yes my generators are fine, it's the scope.
I would like to do the modification myself.

AFAIK the VLF 100Hz square wave thing, and this HF X10 compensation when relay clicks thing are completely different. Of course you could be unlucky and have both!

[Hydron]

Nice to see dealing with issue in fast and decisive manner.

I don't think so. Any A-brand would do a recall and fix the problem for free. Having people deal with soldering SMD and handling small connectors themselves is a recipe for dissaster. How is this going to work out when it comes to warranty? The second best option would be to have the local dealers handle it by sending the scopes to a local company which specialises in repairing electronics at a component level (for example a laptop repair service).

I'd agree if this is the only option for the fix. Tautech - Is DIY SMD soldering the only way people will get the issue sorted? Or is this just an option for those who'd prefer to do it themselves, with return for service for anyone who doesn't want to do the soldering? If it's the latter then if I owned one I'd be happy to have both options (and would likely do the soldering in preference to mucking about posting it off).

[TK]

Nice to see dealing with issue in fast and decisive manner.

I don't think so. Any A-brand would do a recall and fix the problem for free. Having people deal with soldering SMD and handling small connectors themselves is a recipe for dissaster. How is this going to work out when it comes to warranty? The second best option would be to have the local dealers handle it by sending the scopes to a local company which specialises in repairing electronics at a component level (for example a laptop repair service).

I'd agree if this is the only option for the fix. Tautech - Is DIY SMD soldering the only way people will get the issue sorted? Or is this just an option for those who'd prefer to do it themselves, with return for service for anyone who doesn't want to do the soldering? If it's the latter then if I owned one I'd be happy to have both options (and would likely do the soldering in preference to mucking about posting it off).

SMD soldering + trimmer adjustment... it will void any factory calibration.  Will Siglent recalibrate for free all scopes repaired by customers?  And not all 4pf SMD capacitors are the same, so even if you do the fix, you might end up with a different result tested by Siglent.

[nctnico]

Another problem I see is that if the board needs further adjustment you have to run it with the power supply outside the case as well. And some connectors are fixated with glue. Look at Dave's teardown on how to take it apart.

[MrW0lf]

I don't think so. Any A-brand would do a recall and fix the problem for free. Having people deal with soldering SMD and handling small connectors themselves is a recipe for dissaster. How is this going to work out when it comes to warranty?

Well gotta put things in perspective. I have two quite ok 120MHz signal gens on the desk here, total expense ~1000€. What will A-brand offer for that money? To get similar (for my needs) ~10k€ easy.
Ok, I had to replace mainboard in one of mine - so what, works just fine afterwards.
Now if look scopes - what you get for similar price from A-brands? Mostly nothing at all, at best dumbed down toys with only +-*/ math capability*
But of course, warranty must remain and go in effect, even if user will break scope in repair process. I personally would actually prefer DIY repair because very reluctant to give out my usually very clean, well handled and scratch free equipment.
If all this too extreme may pay bit more and go for GWI or some other "middle ground" brand.

*beware, it may be considered racist practice if you know and use math
http://dailycaller.com/2017/10/23/professor-claims-math-algebra-and-geometry-promote-white-privilege/

[StillTrying]

I would have expected Step 3: Tweak the front end to start with compensate the probes on 0.5V/div or 1V/div using the scope's 1kHz before starting.

Edit. The 2 variable cap. check/tweak, is nothing to do with X10 probes.

[rf-loop]

I would have expected Step 3: Tweak the front end to start with compensate the probes on 0.5V/div or 1V/div using the scope's 1kHz before starting.

Afaik, this front end adjustment procedure with signal generator do not need probes.

[alreadystarted]

I would have expected Step 3: Tweak the front end to start with compensate the probes on 0.5V/div or 1V/div using the scope's 1kHz before starting.

Afaik, this front end adjustment procedure with signal generator do not need probes.

Yes some elaboration on the front end calibration would be nice.  I too assumed this step should be performed without probes (coax only), but reading it again it would make more sense to use middle-adjusted x10 probes.

Teardown and soldering is not a big hurdle, but getting the calibration right is a little scary.  Especially for casuals like me with my mostly bottom shelf aliexpress test equipement=)

Also any chance of getting a set of correct caps mailed out on request?  Or at least a list of supplier product page links for each region?

[StillTrying]

Yes some elaboration on the front end calibration would be nice.  I too assumed this step should be performed without probes (coax only), but reading it again it would make more sense to use middle-adjusted x10 probes.

That makes sense to me, I don't understand how it can be done without using X10 probes, but if it's not clear to at least 3 of us Step 3: Tweak the front end is not clear. I don't why it's done at 10kHz either, there's not much LF response left to compare the HF response against at that frequency.

Shall I ask about a similar fix for my CML+?   
https://www.eevblog.com/forum/testgear/siglent-sds1202x-e-compensation-issue/msg1326084/#msg1326084

[pigrew]

That makes sense to me, I don't understand how it can be done without using X10 probes, but if it's not clear to at least 3 of us Step 3: Tweak the front end is not clear. I don't why it's done at 10kHz either, there's not much LF response left to compare the HF response against at that frequency.

The datasheet says 18pF input, so I think you'd want to make a little thru board that has a 50 ohm load, plus a series 1Mohm || 18 pF.

You may need a slightly different capacitor value due to board parasitics....

Thoughts?

[nctnico]

I would have expected Step 3: Tweak the front end to start with compensate the probes on 0.5V/div or 1V/div using the scope's 1kHz before starting.

Afaik, this front end adjustment procedure with signal generator do not need probes.

Yes some elaboration on the front end calibration would be nice.  I too assumed this step should be performed without probes (coax only), but reading it again it would make more sense to use middle-adjusted x10 probes.

No. Connect a function generator directly with a coax cable! The adjustment is for the input circuit's capacitive divider and not the probe.