Siglent SDS1104X-E and SDS1204X-E Mixed Signal Oscilloscopes

[Performa01]

An up-to-date oscilloscope would display 7.6.1.26

Btw, the correct version string for a fully up-to-date SDS1004X-e would be 7.1.6.1.26

[4x1md]

I installed the new firmware. The cursors label is still hard to read. Its transparency can be adjusted but it can't be set to 0% and the background color is too bright. It should be darker. I would also move it to a corner instead of the right middle point.

7. Fixed the bug: The setting of Educational mode can’t be saved after power off.

What is Educational mode? Where can I find it?

[tautech]

7. Fixed the bug: The setting of Educational mode can’t be saved after power off.

What is Educational mode? Where can I find it?

Don't know....I'll ask. 

[kcbrown]

I installed the new firmware. The cursors label is still hard to read. Its transparency can be adjusted but it can't be set to 0% and the background color is too bright. It should be darker. I would also move it to a corner instead of the right middle point.

The problem isn't the transparency.  Even if it were possible to make it completely transparent, some of the text would be unreadable in your example because it's the same color as the trace that's underneath.

I think the best answer here is probably for the transparency to remain present (though I wouldn't object to making it possible to make it completely transparent), and for the text to have an offset black shadow, so that the text is readable no matter the background (the main text color takes makes the lettering readable on dark backgrounds, and the black shadow would make the lettering readable on a light background).

Here's what your high-transparency example would look like with the effect I propose (diagonal shadow):



And here's an alternate version of that, with the shadow going straight to the left:



Now, the above isn't ideal, since "noise" with the right characteristics will make the text difficult to read even with the shadowing effect, but it's still an improvement over the original.  But look at the improvement you get with a more opaque background (here I'm using the diagonal shadow, since I think it's more effective overall):

[bugi]

What would it look like with full 1 pixel halo/outline?

[4x1md]

The problem isn't the transparency.  Even if it were possible to make it completely transparent, some of the text would be unreadable in your example because it's the same color as the trace that's underneath.

Reduced transparency with a dark background can solve this issue.

[4x1md]

[Suggestion]

And moving the cursors label to a corner will improve the usability much more.

[kcbrown]

What would it look like with full 1 pixel halo/outline?

On the transparent background, this:



I'm outlining by hand.  I could do the same for the relatively opaque background, but I think it's clear that the better solution for a relatively opaque background is for the background color for the cursor data to be substantially darker.  That won't work well for dark-colored text, but I don't know if there's any such text that the scope uses in conjunction with a transparent background in the first place.  A darker background would work well even if there were some transparency of it.

So it may be that the best solution here (certainly the simplest) is to move the cursor data area to the corner and to change the background for it (and, I think, for most other things for which the transparent background is used) to a dark gray:



Or a very dark gray (12.5%):



That said, I think it's valuable to have a black outline around the text for those users who prefer a more transparent background.

[bluejedi]

The problem isn't the transparency.  Even if it were possible to make it completely transparent, some of the text would be unreadable in your example because it's the same color as the trace that's underneath.

Reduced transparency with a dark background can solve this issue.

I have already made several [Suggestion]'s for improving overlays (including adding black transparent background) in my post "Improve the readability of overlays" from July 31, 2018.

Please check this post: https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/msg1713224/#msg1713224

If now Siglent would only implement them..

[bluejedi]

@tautech
It would be useful if we could actually get some feedback from Siglent regarding our suggestions.
Whether Siglent has any plans to implement them and which (not).

[bluejedi]

[suggestion] for the layout of suggestions on the forum (this thread at least):

Suggestions for the layout of suggestion posts to make suggestions easier to find and recognize

Easier to find and recognize while manually scrolling through the posts and also when using the search option.

• Add a [suggestion] tag at the beginning of the post like above.
  Searching for the tag '[suggestion]' will give more relevant, more targeted results than searching for the words 'suggestion' or 'suggestions'.
  
  
• Use bold font and a specific (easy recognizable) color for the tag.
  I suggest to use green like above.
  
  
• Add a brief title that clearly (functionally) describes the goal of the suggestion(s)
  Make the title stand out (e.g. like above, in larger font and colored).
  
  
• Use a separate bullet per suggestion and describe each suggestion briefly.
  
  
• If there are multiple unrelated suggestions then create separate suggestion posts (group related suggestions in a single post).
  (e.g. multiple suggestions related to 'improving visibility of overlays' can be combined in a single post.)
  
  

After the list with suggestions additional information can be given (e.g. more details, background information, references etc.), just like this.

[tautech]

@tautech
It would be useful if we could actually get some feedback from Siglent for our suggestions.
Whether Siglent has any plans to implement them and which (not).

In any detail, that's not available to me but as most are aware Siglent get numerous pointers to ideas/bugs and improvements suggested in this and their Siglent Support thread.
I will thank all contributions thus far and as you might see member Siglent does spend some time on this forum following reports and suggestions you have all offered, please keep it up ! 

As you might understand their primary focus is correcting any bugs reported then adding enhancements/improvements suggested from the database they keep on each product. If we like to look back through this thread for suggestions and correlate them to subsequent firmware changelogs you'll see a good few have been implemented already but of course they are just those that have risen to 'important' status.
Each bug/suggestion is assessed and graded before adding to the database then depending on the resources allocated addressed when a FW revision is due/needed. Yes, this takes time to fully asses each issue/suggestion and find a solution that they are happy with. Sometimes beta firmware is offered to beta testers before public release.....more time ! Other times new firmware is just added into production and is available in new units before we have a public firmware release.

It would seem to me that most very important issues have now been addressed and those we can 'work around' have a lower priority will be looked at in forthcoming FW.

It's not like the SDS1*04X-E's are their only new product so resources need be allocated to areas of priority ........SDS5kX, SVA1015X, SSG3kX plus addressing the existing product ranges.

[rf-loop]

Somewhere time ago someone talk about need of shut off input channels trace display when using example math functions or if user want use one channel as Ext trig input. In these cases sometimes it is nice if trace display can turn off.

If someone have not noted this change in last FW versions...

If you use example math and you want only see math result you can turn channel(s) trace(s) display off. Also if you use one input for external trig. You can turn this trace off if you do not want see it.
Every channel trace display can turn off and it do not affect functions. Selection is in channel setup menu.

[Chuki]

Thank you for helping. Sorry for the long response - as I had Imaginary 3 (of 5) in university for electrotechnical I had to read a lot even to understand what you explained

As for 8bit and 25LSB - looks like that is right for 10 div systems. Our 1204 is 8 div. So 32 LSB per DIV is in our case?

So I do the math. At 500mkV/div my systems LSB should be 4mV / 256 = 0.016 mV = 15 mkV
At 1V per div - LSB should be 8 / 256 = 31 mV

At 500uV / DIV I have mean pk-to-pk 530uV on all 4 channels. Looks close to promised 0.5 div Std-dev.

Thank you. Looks like my "problem" and your explanation helped me to understand things I couldn't in Univ.

Like in old anecdote - if the problem could be solved with money that is not a problem this is an additional charge.


To be true. Before reading f.manuals I tried as always to kick the wheel - I made self-cal few times. No help with random results.
Then I upgraded to new software + cal. I got better results with 2 channels (1+2).


Also, could I make an assumption from my first screens that second part of my system (Ch3+Ch4) are less noise of better calibrated than the first one?
As these 2 channels always are close to 0 instead of Ch1+Ch2.

Also for me is not understandable why in GND coupling mode any voltage at all appears on CH3. I thought that It shouldn't be there theoretically.

Could you please also help what "Quck Cal. ON" function is. Tried to check with user manual - couldn't find.

There is nothing wrong with your SDS1204X-E (Regardless what firmware version you’re running).

It is an 8-bit DSO and one vertical division on the screen equals 25 LSB of the ADC. Now do the math…

At 1V/div, 1 LSB of the ADC equals 1/25V = 40mV.

It comes as no surprise that automatic measurements will always pick up at least 1LSB (peak to peak) since this is about the minimum noise generated by the ADC itself.

Change the vertical gain to 100mV/div and you’ll most likely get 4mV.

Try 1mv/div and you will get 40µV ADC noise, but actual reading will be more because at high sensitivities like this, noise from the frontend will contribute to the result.

[Performa01]

Chuki, you're opening a can of worms with all your questions

To explain it all thoroughly, it would certainly take much more than just a short posting.

First, trust me that one division on your scope is equivalent to 25LSB. The total screen height is 200LSB and there are 27LSB headroom above and below.

There is no significant difference between channels regarding noise, but the calibration has only finite resolution and cannot cancel out very small offset errors. This is why offset calibration can never be perfect for the very high sensitivities like 500µV/div and this is also why there will always be a slight offset difference between channels.
Yes even the noise can appear different with certain channel gains, just because it depends on the precise vertical position (= internal offset) setting how often the LSB of the ADC (analog to digital converter) is toggled. As I said before, one LSB of noise is always to be expected, and if it's less than that, it's just a special situation that can easily change anytime or needs not be the same on some other channel.

GND coupling is a very special case and it does not really shorten the scope input like it did on old analog scopes. It just shuts off the output of the VGA (variable gain amplifier) of that channel, so some tiny offset error of the ADC as well as some numerical offset manipulation in the signal processing path might still cause a small visible offset.

If you are really interested in a more thorough explanation how these things work, feel free to read my posts in this discussion (starting at reply #129):

https://www.eevblog.com/forum/testgear/siglent_s-new-product-sds2000x-series/msg1889885/#msg1889885

Yes, it is a different DSO (SDS2000X), but the principles are exactly the same.

As to Quick Cal, this is just fast calibration "on the fly" whenever the DSO detects a major temperature difference and/or a change in the channel gain setting. It will cause a a short "hickup" of your measurement and then results are more accurate again. This should not happen anymore once the scope is properly warmed up and ambient temperature remains fairly constant.

Some people don't like this feature, so it can be turned off or on, just as desired.

[Coldblackice]

Considering buying this scope, but wondering: is there any reason to hold off or wait on buying this scope? I.e., are there any hardware revisions that should be waited on, any downsides to current inventory, will the next hardware revision have some sort of improvement/benefit?

I don't usually jump on to new tech, instead waiting for a couple revisions to iron out any kings or bugs.

[tautech]

Considering buying this scope, but wondering: is there any reason to hold off or wait on buying this scope? I.e., are there any hardware revisions that should be waited on, any downsides to current inventory, will the next hardware revision have some sort of improvement/benefit?

We can already see revisions with different codes in the SN# compared to first units.
I think maybe a little quieter fan than in early units.

[Bakerboy]

Has anyone managed to get WIFI working on this scope with a  third party wifi adapter?
 ive got a TP link TL-WN725N wifi adapter if i put it in the scope it detects it and i can scan for and detect available networks, but when i try to connect it fails after a few seconds.

[tubularnut]

Has anyone managed to get WIFI working on this scope with a  third party wifi adapter?
 ive got a TP link TL-WN725N wifi adapter if i put it in the scope it detects it and i can scan for and detect available networks, but when i try to connect it fails after a few seconds.

Yes it works, but beware if you have special characters in your password, it is very choosy about what it likes and doesn't like.

[Bakerboy]

This happens to me every time, ive even tried hot spotting my phone

[tautech]

Has anyone managed to get WIFI working on this scope with a  third party wifi adapter?
I've got a TP link TL-WN725N wifi adapter if i put it in the scope it detects it and i can scan for and detect available networks, but when i try to connect it fails after a few seconds.

Confusingly I've had similar problems of late when previously none.
I've got a very early 1104X-E SN#0012 and TP Link adapter and previously it worked at good range but not so much now. I bought 2 units TP Link WiFi adapters and lately they both exhibit reduced range compared to previous performance........ 

Optimizing your WiFi antenna orientation can help just as moving closer.
I'll revisit this issue in attempting to find the root cause over the next few days.

[4x1md]

Has anyone managed to get WIFI working on this scope with a  third party wifi adapter?
 ive got a TP link TL-WN725N wifi adapter if i put it in the scope it detects it and i can scan for and detect available networks, but when i try to connect it fails after a few seconds.

Does the SSID of your network have spaces? If yes, they are the problem.

[tautech]

A hands up please of those using the Siglent WiFi adapter TL‐WN725N.
Some reports on your experiences and range with this USB adapter please.

Also please, reports from those using the genuine TP Link TL-WN725N dongle.


This is what I have and experience infrequent drop outs of connection to our WLAN.
I've even tried another TP Link TL-WN725N I have in stock.
Range is not good for reliable connections, maybe 10 M throughout a wooden dwelling while all our other WiFi devices work at much further distances.....phones, tablets etc.

In some attempt to find if there has been firmware changes that could impact on WiFi performance I've rolled the versions back one by one from V6.1.26 back to V6.1.20R1 with no noticeable changes to WiFi connectivity stability.

I seem to remember better WiFi performance but in the intervening time since receiving my beta unit we've changed WiFi hardware.............but all our other devices still work OK. 

Are these TP Link based dongles under-powered rubbish ?

Edit
While researching I found this:

On the WIFI front - I found I needed to turn DHCP on to grab an IP (which I then set in my router) and then turn it off to keep those settings - otherwise the 1104XE was constantly disconnecting and reconnecting for some reason. Rock solid once DHCP turned off and the IP saved.

More checks being undertaken.......

[rf-loop]

And why DHCP need be on in Oscilloscope? For what?

[4x1md]

And why DHCP need be on in Oscilloscope? For what?

DHCP client for obtaining an IP automatically.