Siglent SDS2000X Plus

[tautech]

Calling @Frex to post screenshots of the Sys Info page of the unit we helped to recover.
This info can be helpful to Siglent to find where the update issue is.
Previous FW version too if remembered.
TIA

[Frex]

Hello,
A screenshot of sysinfo of my "recovered" SDS2104X-Plus.
(Thanks to Tautech  )
Regards.
Frex

[Anthocyanina]

In XY mode the timebase doesn't matter, because the scope is not plotting voltage*time, but voltage*voltage

i've found that timebase does matter, at least on all my oscilloscopes, because the XY plotter will plot everything currently in memory, too fast a timebase compared to the input signals and you only get choppy bits of the whole picture, too slow a timebase, and if the picture changes too quickly, you get the full picture, plus remnants from the previous "frame", plus all the "connecting lines"

[KungFuJosh]

In XY mode the timebase doesn't matter, because the scope is not plotting voltage*time, but voltage*voltage

i've found that timebase does matter, at least on all my oscilloscopes, because the XY plotter will plot everything currently in memory, too fast a timebase compared to the input signals and you only get choppy bits of the whole picture, too slow a timebase, and if the picture changes too quickly, you get the full picture, plus remnants from the previous "frame", plus all the "connecting lines"

Interesting. However, it sounds like you're describing my experience with adjusting the memory depth. It's likely the timebase's effect on the memory depth that you're seeing affect XY mode.

[Anthocyanina]

In XY mode the timebase doesn't matter, because the scope is not plotting voltage*time, but voltage*voltage

i've found that timebase does matter, at least on all my oscilloscopes, because the XY plotter will plot everything currently in memory, too fast a timebase compared to the input signals and you only get choppy bits of the whole picture, too slow a timebase, and if the picture changes too quickly, you get the full picture, plus remnants from the previous "frame", plus all the "connecting lines"

Interesting. However, it sounds like you're describing my experience with adjusting the memory depth. It's likely the timebase's effect on the memory depth that you're seeing affect XY mode.

i do this mostly with the analog discovery 2, with 8k points, and have found timebase matters even if memory stays the same. when i did this on my keysight which has no way to adjust memory depth, i also found timebase mattered, especially since to get a good image i had to use the high resolution mode, which gets weird at the faster timebases, sometimes seeing just a bunch of thick blurry points if too fast, or chunks of the image if too slow, the keysight is dead now, so i can't see this new song on that one, sadly.

[KungFuJosh]

i do this mostly with the analog discovery 2, with 8k points, and have found timebase matters even if memory stays the same. when i did this on my keysight which has no way to adjust memory depth, i also found timebase mattered, especially since to get a good image i had to use the high resolution mode, which gets weird at the faster timebases, sometimes seeing just a bunch of thick blurry points if too fast, or chunks of the image if too slow, the keysight is dead now, so i can't see this new song on that one, sadly.

I'll try it out and see what happens with my Siglent.

[bdunham7]

Interesting. However, it sounds like you're describing my experience with adjusting the memory depth. It's likely the timebase's effect on the memory depth that you're seeing affect XY mode.

The main issue with DSOs and XY plots is actually "blind time".  For an analog CRO without cursors or indicators there should be no blind time, but on a DSO blind time can exceed 99%.  Blind time can be calculated if you know the timebase and refresh rate (waveforms per second).  Reducing the memory depth decreases the processing time and increases your waveform rate.  For rapidly changing XY patterns like your music test, small memory and short timebases with a high waveform rate may look the best even though the signal is actually getting chopped up into small segments.  There's an ideal setting for this that gives you a very high waveform rate (as in actually almost meets the advertised spec) but I don't recall that setting at the moment.

[KungFuJosh]

Interesting. However, it sounds like you're describing my experience with adjusting the memory depth. It's likely the timebase's effect on the memory depth that you're seeing affect XY mode.

The main issue with DSOs and XY plots is actually "blind time".  For an analog CRO without cursors or indicators there should be no blind time, but on a DSO blind time can exceed 99%.  Blind time can be calculated if you know the timebase and refresh rate (waveforms per second).  Reducing the memory depth decreases the processing time and increases your waveform rate.  For rapidly changing XY patterns like your music test, small memory and short timebases with a high waveform rate may look the best even though the signal is actually getting chopped up into small segments.  There's an ideal setting for this that gives you a very high waveform rate (as in actually almost meets the advertised spec) but I don't recall that setting at the moment.

That was my assumption regarding the timebase adjustments. I just tested it out, and the waveform rate around 2 - 5MSa/s seems to be the best performance for that music. Going too high on the timebase made it both slow down and reduce the waveform proportions on screen. That's with 10kpts set as max mem depth.

[KungFuJosh]

Here's a slightly extreme comparison for refresh/mem depth. At 1Gs/s, the image is squished. I wonder what the intended rate is from the music.

[TopQuark]

Here's a slightly extreme comparison for refresh/mem depth. At 1Gs/s, the image is squished. I wonder what the intended rate is from the music.

Shameless plug to my attempt with the HD
https://www.eevblog.com/forum/testgear/siglent-sds2000x-hd-12bit-(published-for-chinese-domestic-market-only)/msg5446511/#msg5446511

Scope settings:
Input settings (CH1, CH2) : 50-ohm termination, DC coupled, 20M BW limit
Timebase : 1 ms/div, 1 MSa/s, 10 kpts memory
Trigger : Normal, alternating edge, CH1, AC coupled, 0V offset

Using the high quality .WAV file, playing it back at high bit-rate and resolution makes a huge difference. Interested to see if that increases quality on the Plus as well. 

[KungFuJosh]

I bought a copy of OsciStudio to test this out, and eventually make a very stupid video.

In the meantime, it was useful to test both refresh rates and audio frequency effect on the XY results:




The first section is more obvious, that's scrolling through refresh rates (first 50 seconds). The last 30 seconds is screwing around with the frequency of the shape. Chris's tutorial shows him using A1 at 55Hz, which is (one of the reasons) why my results aren't as great with his music. A4 at 440Hz (and really most notes over ~100Hz) show much greater stability on my scope.

[KungFuJosh]

Scope settings:
Input settings (CH1, CH2) : 50-ohm termination, DC coupled, 20M BW limit
Timebase : 1 ms/div, 1 MSa/s, 10 kpts memory
Trigger : Normal, alternating edge, CH1, AC coupled, 0V offset

Using the high quality .WAV file, playing it back at high bit-rate and resolution makes a huge difference. Interested to see if that increases quality on the Plus as well. 

Your settings are pretty close to where I landed after my test video. However, the default edge trigger and your trigger settings didn't appear to make any perceivable (to me) difference, neither did changing coupling.

I will make another video to see if I can get better results than my previous attempt.

Thanks,
Josh

[KungFuJosh]

Interested to see if that increases quality on the Plus as well. 

I did some tests to see if anything made any significant impact in quality. The short answer: nope. The HD model has better hardware, and the results show it. At least that's my uneducated guess. 😉

I tried all sorts of things including running my audio card at 44.1kHz to 192kHz (with the wav file loaded in Studio One to match the output settings).

I also tried using some mp4 files that came with the software purchase (not sure if they're available free), and that let me tweak some settings. Unfortunately, the result was I couldn't get a happy medium between memory depth and refresh rate to reduce flickering, or maintain enough of the image. It still looked pretty good, but is just another thing (that doesn't matter) that makes me want an HD scope.

[famalex]

Hi all.
I need help. Lately, to turn the oscilloscope on or off, I had to press the power button long and hard. I assumed that this was due to contacts contamination (maybe wear?) and finally decided to disassemble the device to get to the keyboard (too much fuss for such a simple task) and try to fix this problem. After removing the power board, I accidentally discovered a burnt component.
This is capacitor C18. The device, of course, worked without it, but the breakdown needs to be fixed. If anyone knows the parameters of this capacitor, please share, it will help me a lot.

[tautech]

It's across rectified mains 
Therefore should be 4-500V rated.

One data point from warranty replaced SMPS IIRC from SDS1104X-E of which the area in question looks identical.

Required 2 irons to remove and measures 228nF @100Hz. ESR 45 Ohms with 0.5V stimuli.
Use 220-247nF 400-500V rated.

What does yours measure ?
Do you think the joint has failed or is the cap now shorted ?

[famalex]

Thanks for your prompt response. 
Yes. This capacitor is installed on the high voltage side immediately after the rectifier bridge.
As soon as I fix this problem, I will report the results.

[famalex]

It's across rectified mains 
Therefore should be 4-500V rated.

One data point from warranty replaced SMPS IIRC from SDS1104X-E of which the area in question looks identical.

Required 2 irons to remove and measures 228nF @100Hz.
Use 220-247nF 400-500V rated.

What does yours measure ?
Do you think the joint has failed or is the cap now shorted ?

The cap is absolutely dead. It turned into a 58.32 ohm resistor 

[Martin72]

Required 2 irons to remove and measures 228nF @100Hz.
Use 220-247nF 400-500V rated.

247nF does not exist, the standard values in the measured range are 220nF, 270nF, with a corresponding tolerance of 10 or 20%.
Looks like an MLCC guy, bad stuff.
Individual layers can be short-circuited or broken by mechanical stress (not perfectly flat, then soldered).

[famalex]

The exact capacity doesn't really matter. This is a high frequency noise filter capacitor.

[tautech]

It's across rectified mains 
Therefore should be 4-500V rated.

One data point from warranty replaced SMPS IIRC from SDS1104X-E of which the area in question looks identical.

Required 2 irons to remove and measures 228nF @100Hz.
Use 220-247nF 400-500V rated.

What does yours measure ?
Do you think the joint has failed or is the cap now shorted ?

The cap is absolutely dead. It turned into a 58.32 ohm resistor

See ESR measurement above.

It is very pleasing to know the only dead Siglent SMPS I have and have ever seen, is of use to someone. 

[famalex]

In general, I was not very pleased with the quality of the components in the power board. Maybe I'm wrong.

[Martin72]

The exact capacity doesn't really matter.

That's not what I meant either.

[famalex]

The exact capacity doesn't really matter.

That's not what I meant either.

I understand. Just complemented your thought.

[famalex]

It is very pleasing to know the only dead Siglent SMPS I have and have ever seen, is of use to someone. 
[/quote]
Apparently I was very lucky.

[Martin72]

In general, I was not very pleased with the quality of the components in the power board.

Well...
A "nice" power supply unit costs money.
You can install it, but then the scope can no longer cost the money it costs.
We manufacture power supplies for submarine and aviation applications, with corresponding quality criteria down to component level.
For me, anything less is just garbage.
As always, everything is relative from a certain point of view.