Siglent SDS2000X Plus

[tautech]

... and a question:

What is ths 'Para Switch' for in the Save menu?

CSV output seems to be identical on and off.

From the User manual P309:
CSV Data
Saves the waveform data to the external memory in ".csv" format. After selecting this type, you can touch the ParaSwitch to determine to include the scope configuration parameter information (horizontal timebase, vertical scale, etc.) or not. Touch the Save All Channel to save all displayed waveforms on the screen or select the Source to save. The available sources include C1 ~ C4 and F1 ~ F2.

[Peter_O]

Thx rob!
I just found that. See my edit.     We wrote in parallel.

So it's a newer feature that came with one of the firmware upgrades. And Silent did an update of the manual too. Fine!

But what about a FFT peak list export?

[tautech]

But what about a FFT peak list export?

I would try engaging FFT Peak Hold.
It will lift the noise floor and clean up the signal some therefore giving better data in the download to work with.

Sorry it's late here otherwise I'd offer some screenshots..... 

[balnazzar]

Folks, having played with the scope for some weeks, I can report it's indeed a good instrument. And I suspect that as I (hopefully) improve my knowledge, I'll find it even better.

I'd like, however, to make a few observations..:

1. Using the web interface, one could theoretically use it as it was an usb scope, with all the advantages that come with that. Two things, however, do prevent a truly satisfying experience. A) The web interface is lagged. Like a lot lagged. One could observe that that's due to the vnc tunnel and/or the ethernet link, but when I control other PCs in my home network using vnc (or a variant thereof, like rdp) I experience almost no lag. There is much to be improved. Possibly, compression protocols, or whatever. B) The scope's 1024x600 interface looks horribly blurred and coarse when scaled up brutally to 28" 4K (3840x2160). That would be easily addressable by making the client perform proper rescaling (as most vnc PC clients do).

2. I'm still trying to figure out the best way to make the scope impervious to dust. Originally, I wanted to invert airflow direction, that is, making the big opening on the back an air intake, and use the lateral vents as exhaust. The scope would have had positive pressure inside it when turned on, and the back opening would have been easily filterable.
I'm not sure, however, that inverting the airflow is actually a good idea. One alternative would be leaving the airflow direction as it is, and filter the lateral intakes. It'll be ugly, though. The only way to do this is by applying bi-adhesive, and then cut two strips of filter and attaching them to the bi-adhesive. Advice would be welcome.

[2N3055]

Folks, having played with the scope for some weeks, I can report it's indeed a good instrument. And I suspect that as I (hopefully) improve my knowledge, I'll find it even better.

I'd like, however, to make a few observations..:

1. Using the web interface, one could theoretically use it as it was an usb scope, with all the advantages that come with that. Two things, however, do prevent a truly satisfying experience. A) The web interface is lagged. Like a lot lagged. One could observe that that's due to the vnc tunnel and/or the ethernet link, but when I control other PCs in my home network using vnc (or a variant thereof, like rdp) I experience almost no lag. There is much to be improved. Possibly, compression protocols, or whatever. B) The scope's 1024x600 interface looks horribly blurred and coarse when scaled up brutally to 28" 4K (3840x2160). That would be easily addressable by making the client perform proper rescaling (as most vnc PC clients do).

2. I'm still trying to figure out the best way to make the scope impervious to dust. Originally, I wanted to invert airflow direction, that is, making the big opening on the back an air intake, and use the lateral vents as exhaust. The scope would have had positive pressure inside it when turned on, and the back opening would have been easily filterable.
I'm not sure, however, that inverting the airflow is actually a good idea. One alternative would be leaving the airflow direction as it is, and filter the lateral intakes. It'll be ugly, though. The only way to do this is by applying bi-adhesive, and then cut two strips of filter and attaching them to the bi-adhesive. Advice would be welcome.

It is using VNC. Scope is small embedded processor based. Comparing VNC performance with powerful PC  host is not really propper comparison.
I do see a bit faster response (a really small one) when using dedicated VNC client software on PC side.
Compression would not help, it is not network speed.
Also, you cannot resize non integer ratios and expect perfect rendering..

[balnazzar]

Folks, having played with the scope for some weeks, I can report it's indeed a good instrument. And I suspect that as I (hopefully) improve my knowledge, I'll find it even better.

I'd like, however, to make a few observations..:

1. Using the web interface, one could theoretically use it as it was an usb scope, with all the advantages that come with that. Two things, however, do prevent a truly satisfying experience. A) The web interface is lagged. Like a lot lagged. One could observe that that's due to the vnc tunnel and/or the ethernet link, but when I control other PCs in my home network using vnc (or a variant thereof, like rdp) I experience almost no lag. There is much to be improved. Possibly, compression protocols, or whatever. B) The scope's 1024x600 interface looks horribly blurred and coarse when scaled up brutally to 28" 4K (3840x2160). That would be easily addressable by making the client perform proper rescaling (as most vnc PC clients do).

2. I'm still trying to figure out the best way to make the scope impervious to dust. Originally, I wanted to invert airflow direction, that is, making the big opening on the back an air intake, and use the lateral vents as exhaust. The scope would have had positive pressure inside it when turned on, and the back opening would have been easily filterable.
I'm not sure, however, that inverting the airflow is actually a good idea. One alternative would be leaving the airflow direction as it is, and filter the lateral intakes. It'll be ugly, though. The only way to do this is by applying bi-adhesive, and then cut two strips of filter and attaching them to the bi-adhesive. Advice would be welcome.

It is using VNC. Scope is small embedded processor based. Comparing VNC performance with powerful PC  host is not really propper comparison.
I do see a bit faster response (a really small one) when using dedicated VNC client software on PC side.
Compression would not help, it is not network speed.
Also, you cannot resize non integer ratios and expect perfect rendering..

I know it uses vnc, that's why I was asking for the same behaviour we get on a PC-to-PC vnc connection. About the impossibility of having zero (or near zero) lag and the lack of computational power, I understand. But as for the resolution, note how vnc clients perform resizing & rescaling on PCs (I mean both proper vnc clients and rdp on windows): regardless of the resolution/size of the monitor attached to the host you are remotely operating (or even if it's a headless machine) the resolution is always properly rescaled for the monitor you are using on the guest PC. The OS's UI is rescaled accordingly.
That's because the non-integer ratios between pixels don't enter into the picture at all. The host sends the information to the guest about what's supposed to be rendered on the screen. The guest's vnc client performs the actual rendering at the native resolution.
I wonder why we cannot have the same for the scope interface.
If you want to bring ratios in the picture because, say, the PC-style rescaling is too demanding for some reason (but the heavy lifting is made on the client..), then it would suffice to rescale up to a window sized as the maximum integer-multiple resolution that's closest to the native one. E.g., I got a 4K monitor, then it could rescale at 1024*3 x 600*3.

[2N3055]

Folks, having played with the scope for some weeks, I can report it's indeed a good instrument. And I suspect that as I (hopefully) improve my knowledge, I'll find it even better.

I'd like, however, to make a few observations..:

1. Using the web interface, one could theoretically use it as it was an usb scope, with all the advantages that come with that. Two things, however, do prevent a truly satisfying experience. A) The web interface is lagged. Like a lot lagged. One could observe that that's due to the vnc tunnel and/or the ethernet link, but when I control other PCs in my home network using vnc (or a variant thereof, like rdp) I experience almost no lag. There is much to be improved. Possibly, compression protocols, or whatever. B) The scope's 1024x600 interface looks horribly blurred and coarse when scaled up brutally to 28" 4K (3840x2160). That would be easily addressable by making the client perform proper rescaling (as most vnc PC clients do).

2. I'm still trying to figure out the best way to make the scope impervious to dust. Originally, I wanted to invert airflow direction, that is, making the big opening on the back an air intake, and use the lateral vents as exhaust. The scope would have had positive pressure inside it when turned on, and the back opening would have been easily filterable.
I'm not sure, however, that inverting the airflow is actually a good idea. One alternative would be leaving the airflow direction as it is, and filter the lateral intakes. It'll be ugly, though. The only way to do this is by applying bi-adhesive, and then cut two strips of filter and attaching them to the bi-adhesive. Advice would be welcome.

It is using VNC. Scope is small embedded processor based. Comparing VNC performance with powerful PC  host is not really propper comparison.
I do see a bit faster response (a really small one) when using dedicated VNC client software on PC side.
Compression would not help, it is not network speed.
Also, you cannot resize non integer ratios and expect perfect rendering..

I know it uses vnc, that's why I was asking for the same behaviour we get on a PC-to-PC vnc connection. About the impossibility of having zero (or near zero) lag and the lack of computational power, I understand. But as for the resolution, note how vnc clients perform resizing & rescaling on PCs (I mean both proper vnc clients and rdp on windows): regardless of the resolution/size of the monitor attached to the host you are remotely operating (or even if it's a headless machine) the resolution is always properly rescaled for the monitor you are using on the guest PC. The OS's UI is rescaled accordingly.
That's because the non-integer ratios between pixels don't enter into the picture at all. The host sends the information to the guest about what's supposed to be rendered on the screen. The guest's vnc client performs the actual rendering at the native resolution.
I wonder why we cannot have the same for the scope interface.
If you want to bring ratios in the picture because, say, the PC-style rescaling is too demanding for some reason (but the heavy lifting is made on the client..), then it would suffice to rescale up to a window sized as the maximum integer-multiple resolution that's closest to the native one. E.g., I got a 4K monitor, then it could rescale at 1024*3 x 600*3.

Ahh.. now I get it what are you talking... VNC viewer built in web control doesn't have rescaling options.....
Yeah, this is  HTML5 VNC client in browser (in Java), and not fully featured desktop application...

I use web control non resized 100% screen most of the time (it's large enough) and if I need more control of the VNC parameters I use VNC Viewer on a PC..

I don't like or use 4K monitors because they make pixels so small you need to resize everything to be able to see... So I avoid problem in the first place..
110 DPI monitor I have is highest pixel density that makes sense on desktop PC to me.. 3440x1440 for a monitor with 800x335 mm display size. (34" SJ55W Ultra WQHD Monitor). And that is marginally on a smallish side already for my eyes..

[balnazzar]

Folks, having played with the scope for some weeks, I can report it's indeed a good instrument. And I suspect that as I (hopefully) improve my knowledge, I'll find it even better.

I'd like, however, to make a few observations..:

1. Using the web interface, one could theoretically use it as it was an usb scope, with all the advantages that come with that. Two things, however, do prevent a truly satisfying experience. A) The web interface is lagged. Like a lot lagged. One could observe that that's due to the vnc tunnel and/or the ethernet link, but when I control other PCs in my home network using vnc (or a variant thereof, like rdp) I experience almost no lag. There is much to be improved. Possibly, compression protocols, or whatever. B) The scope's 1024x600 interface looks horribly blurred and coarse when scaled up brutally to 28" 4K (3840x2160). That would be easily addressable by making the client perform proper rescaling (as most vnc PC clients do).

2. I'm still trying to figure out the best way to make the scope impervious to dust. Originally, I wanted to invert airflow direction, that is, making the big opening on the back an air intake, and use the lateral vents as exhaust. The scope would have had positive pressure inside it when turned on, and the back opening would have been easily filterable.
I'm not sure, however, that inverting the airflow is actually a good idea. One alternative would be leaving the airflow direction as it is, and filter the lateral intakes. It'll be ugly, though. The only way to do this is by applying bi-adhesive, and then cut two strips of filter and attaching them to the bi-adhesive. Advice would be welcome.

It is using VNC. Scope is small embedded processor based. Comparing VNC performance with powerful PC  host is not really propper comparison.
I do see a bit faster response (a really small one) when using dedicated VNC client software on PC side.
Compression would not help, it is not network speed.
Also, you cannot resize non integer ratios and expect perfect rendering..

I know it uses vnc, that's why I was asking for the same behaviour we get on a PC-to-PC vnc connection. About the impossibility of having zero (or near zero) lag and the lack of computational power, I understand. But as for the resolution, note how vnc clients perform resizing & rescaling on PCs (I mean both proper vnc clients and rdp on windows): regardless of the resolution/size of the monitor attached to the host you are remotely operating (or even if it's a headless machine) the resolution is always properly rescaled for the monitor you are using on the guest PC. The OS's UI is rescaled accordingly.
That's because the non-integer ratios between pixels don't enter into the picture at all. The host sends the information to the guest about what's supposed to be rendered on the screen. The guest's vnc client performs the actual rendering at the native resolution.
I wonder why we cannot have the same for the scope interface.
If you want to bring ratios in the picture because, say, the PC-style rescaling is too demanding for some reason (but the heavy lifting is made on the client..), then it would suffice to rescale up to a window sized as the maximum integer-multiple resolution that's closest to the native one. E.g., I got a 4K monitor, then it could rescale at 1024*3 x 600*3.

Ahh.. now I get it what are you talking... VNC viewer built in web control doesn't have rescaling options.....
Yeah, this is  HTML5 VNC client in browser (in Java), and not fully featured desktop application...

I use web control non resized 100% screen most of the time (it's large enough) and if I need more control of the VNC parameters I use VNC Viewer on a PC..

I don't like or use 4K monitors because they make pixels so small you need to resize everything to be able to see... So I avoid problem in the first place..
110 DPI monitor I have is highest pixel density that makes sense on desktop PC to me.. 3440x1440 for a monitor with 800x335 mm display size. (34" SJ55W Ultra WQHD Monitor). And that is marginally on a smallish side already for my eyes..

I didn't think about using a proper vnc viewer to access the scope interface.. Thanks, I'll try ASAP. Any recommendation on specific clients that work best with the scope 'server' would be welcome.

Anyway, 4K monitors (Or any 'high dpi' display) are not meant to be used at their native resolution without (once more) scaling. People usually set the scaling factor between 150% and 200%. For example, at 200% you'll see things exactly as you would with a 1080p monitor of the same size, just much sharper (and that's important if you work a lot with text, particularly pdf/ps/dvi docs that won't cope well with a low-dpi screen).

Note, however, that the problem I described is not directly related with my monitor being 4K. It's just that 1024x600 is a very, very low resolution with respect to a 28" screen size (I can clearly discern the single pixels even on the 10" screen of the scope at 40-50cm viewing distance). It would look blurred all the same on a 28" even if that display's resolution would be much less than 4K.

[2N3055]

I didn't think about using a proper vnc viewer to access the scope interface.. Thanks, I'll try ASAP. Any recommendation on specific clients that work best with the scope 'server' would be welcome...

Tried few of them. All of them worked well. Now I use RealVNC VNC Viewer for all my VNC-ing.

[BillyO]

I don't see any huge advantage of VNC over the web interface.  There may be a few ms difference.  That's it.

Not enough to make me install more spyware permanently on my system.

[Peter_O]

THD measurement

This would be a huge effort when e.g. the table could present at the end the THD.
Actually there´s no alternative from counting the shds together.
When one of our tested inverter shows a suspicious THD on the analyzer, we connect a scope at the output and watching the shds on the FFT screen.
Table function allows you to display the shds in numeric form, as mentioned above it would be cool if the scope would display the thd also.
EDIT:


http://www.sengpielaudio.com/calculator-thd.htm

THD table, calculator

I played some more:

It was easy to verify the calculation of the HP35665A THD value displayed; 
I did it this way:
- Define the number of peaks to take into calculation at the 'Harmonic' menu, e.g. 10.
- Grab the 11 peak values by marker and put them into Excel.
- Do the calculation, e.g. following the paper Rob found:

Just linked this in another thread for Peter:
https://www.researchgate.net/publication/362542224_Total_Harmonic_Distortion_THD_analysis_utilizing_the_FFT_capabilities_of_modern_digital_storage_oscilloscopes

Having read it now, it seems to be kind of a seminar paper on bachelor level, but it does give the basics you need.

Result: HP35665A shows 0,199% THD, calulation gives: 0,19931% So HP do it by rulebook of course.

On the Siglent:

1st Step: by hand

- I fiddled with the reference level to get the maximum count of harmonics into the table, which was 8 in my case.
- Typing the values into XLS
- calculate

The result is 0,2288% which is in the same ballpark.

EDIT:

2nd step: automated
After exporting the FFT as CSV and adding some  quick and dirty formulas, we get 0,2151%, which is feasible, because the formulas do take 10 peaks beside the base frequency:



The xls ist attached too. The export was some 500.000 records, I shortened it to 5000*, because all the action is in the first 5000 records.
- Open your csv export in XLS, convert csv into columns (make sure to address the decimal delimiter right) and paste the columns 1:1 into the left part of the attached XLS.
- Fill in the base frequency (yellow), and there you go. 

*) In case of regular need one should think about a better config of frequency range and number of points.   

And finally, to do the HP3310A justice: After some more adjusting it now shows a THD of 0,118% at 5kHz and 1V RMS.

[2N3055]

....
The result is 0,2288% which is in the same ballpark.
.....

Picoscopes have THD measurements:
4262 16bit combined has distortion floor of 0,002% (I don't know if this is source or scope limit).
So both generated and measured signal will be better than this.
Let's take it as a reference for this purpose.

3406D 8 Bit has distortion floor of 0,183% (with source with 0.002% distortion).
When measuring source with 0,036% it measures 0,194%.

4824A 12 bit has distortion floor of 0,019%
And measures that 0,036% source as 0,039%, really close.

FFT: 100kHz BW, 64k bins, 200kS/s
Sources: Pico 4262 (less than 0.002%) and SDG6000X 0,036%(5kHz signal, 600mv P-P.)

These would be realistic limits...

So in direct THD measurements 8bit scope would be good to measure some non critical distortion. 12 Bit already does quite well in this regard.

Of course if a classic technique of suppressing the fundamental tone with a notch filter is used, then you could measure very low distortion even with 8 bit scope.

[balnazzar]

I don't see any huge advantage of VNC over the web interface.  There may be a few ms difference.  That's it.

Not enough to make me install more spyware permanently on my system.

You have to install just the vnc client, NOT the server.

[BillyO]

You have to install just the vnc client, NOT the server.

Right.  Are you under the impression the client cannot be spyware?  It is.  It's confirmed.  When active, it communicates with an external site.  If your okay with that, that's fine.  Not that I'm assuming they are doing anything detrimental with whatever data they are collecting, but I just don't feel I get enough from the product to warrant that exchange.  As I said, the web interface works just fine and does not feel any more lagging than the VNC client, so why burden my system with more code that consumes resources I may need for other more useful things.

WE all run lots of spyware.  Chrome is spyware, MS Word is spyware.  Windows is spyware.  It goes on and on.  Almost everything you run on your system is spyware.  It's collecting data and sending it off your network to someone else.  Just be aware of that and balance that cost with the benefit of running the software.  In my case VNC offers me nothing for the space it takes up and the resources it uses.

[balnazzar]

You have to install just the vnc client, NOT the server.

Right.  Are you under the impression the client cannot be spyware?  It is.  It's confirmed.  When active, it communicates with an external site.  If your okay with that, that's fine.  Not that I'm assuming they are doing anything detrimental with whatever data they are collecting, but I just don't feel I get enough from the product to warrant that exchange.  As I said, the web interface works just fine and does not feel any more lagging than the VNC client, so why burden my system with more code that consumes resources I may need for other more useful things.

WE all run lots of spyware.  Chrome is spyware, MS Word is spyware.  Windows is spyware.  It goes on and on.  Almost everything you run on your system is spyware.  It's collecting data and sending it off your network to someone else.  Just be aware of that and balance that cost with the benefit of running the software.  In my case VNC offers me nothing for the space it takes up and the resources it uses.

Ok, if you enlist the operating system into the spyware troopers, of course almost every piece of software out there can be spyware/malware.
Of course windows or web browsers collect data, but the companies behind such software pieces are bound by law not to use them maliciously.
What I wanted to highlight was that with a vnc server installed one can potentially gain remote control of your machine (but you have to be very sloppy), whereas that's not possible with just the client.

As for latency, indeed a discrete vnc client doesn't seem to ameliorate the issue. But I still don't follow Sinisa's explanation: indeed, if the processor is powerful enough to send stuff to the instrument's own screen with an optimal update rate, I don't see why it cannot send the same stuff through an ethernet->vnc link. That's not computationally intense.

Vnc clients also don't do sh*t about the raw rescaling. I tested four of them (Real, Tight, Ultra, and Tiger).

[mwb1100]

Are you under the impression the client cannot be spyware?  It is.  It's confirmed.

There are several VNC clients, some open source.  Are they all spyware?

[BillyO]

There are several VNC clients, some open source.  Are they all spyware?

Haven't tested them all.  You can if you like.

My policy is:  Don't install software that does not offer benefit over and above the cost in terms of resources.

I've made a lot of money off people that just completely drown their systems by installing tons of junk "free" software they never use.  MOST of it IS spyware, poorly designed, installs persistent processes and makes the system unstable.  You don't get much for free these days (never did BTW) and most of the free crap out there comes with all the care and consideration for your system that you would expect for the price.

As for me, I will only install exactly what I need and nothing more and if I no longer use it, I take it off and clean up after it.

You do you.

[tautech]

Has anyone watched the short Siglent video on using VNC:

[BillyO]

Has anyone watched the short Siglent video on using VNC:

Thanks!

Just did and was aware of all that.  Still not enough added value for me to install it again.  Tried it and it does not offer enough advantage to warrant staying on my system.  My web browser is sufficient and I can use it for a lot of additional things I can't use VNC for.

[tautech]

Has anyone watched the short Siglent video on using VNC:

Thanks!

Just did and was aware of all that.  Still not enough added value for me to install it again.  Tried it and it does not offer enough advantage to warrant staying on my system.  My web browser is sufficient and I can use it for a lot of additional things I can't use VNC for.

Agree as the inbuilt webserver is already much more powerful but a customer used it to remotely monitor a new X Plus we sold them to solve an issue that had plagued them for a couple of years with a Single shot capture with carefully set trigger conditions.
Pleased they were too as they had other channels capturing data that led to the capture and all with a timestamp too. 

[balnazzar]

We are talking about nothing here, Billy. Using a vnc client or the web interface (which is just a frontend for the vnc server on the scope), apart from personal tastes is utterly irrelevant since none offer proper hi-dpi scaling or other benefits. For example, vnc clients can usually record the screen, but that turns out to be unavailable because "the remote host doesn't allow it".

[krasimir.k]

Hi,
I found other solution with small FAN PWM controller - ZF1X4, which can be found in AliExpress. It has 4 pin connector which is compatible with the Noctua NF-A9 PWM fan (in my case, it was the ChromaMax series of the NF-A9 PWM). The controller can be attached to the fan using double sided tape.
The controller has one button and 3 leds with which is possible to set the minimum temperature and maximum delta temperature. I set it to min 30C and delta 10C, which means that at 40C, the PWM will be 100%.
The setting is a not very intuitive, but the sequence is the following:
1. Press long the button - the LED 1 will be switch on, which is the default setting for min temp. of 30C. pressing once will increase the min. temp. pressing twice will decrease the temp. The new settings is encoding as binary code by LED 1,2,3 like:
1   2   3
on off off - 30C
off on off - 35C
on on off - 40C
off off on - 45C
on off on - 50C
off on on - 60C
on on on - 70C

2. Press long second time the button - you are entering in the delta temp menu. Here is the settings

1   2   3
on off off - 5C
off on off - 15C
on on off - 20C
off off on - 30C
on off on - 40C
off on on - 50C
on on on - 60C

[idolclub]

SDS2000X Plus Series Firmware Update
Version： V1.5.2R2
2022-12-22

Release notes:

1. Fixed a calibration bug only existing on Rev.G hardware (06-xx).


https://int.siglent.com/upload_file/zip/firmware/Oscilloscope/SDS2000X%20Plus_V1.5.2R2_EN.zip

[rowant]

Just received a SDS2104X Plus with V1.5.2R1, with disabled telnet. If you need telnet, the downgrade to 1.3.9R12 worked.

For some reason the telnet password no longer works, even after the downgrade (root/siglent_sds1000x_e), however to get around it just use

Code: [Select]

telnetd -l /bin/sh in the

Code: [Select]

siglent_device_startup.sh file, then it won't ask for any login details.

Updated then to 1.5.2R2, all good

[Amphion]

Am I doing something wrong?? Not sure its a bug but I can't figure it out so putting it in this thread.

SDS2104X+ with V1.5.2R2 firmware (this same behavior existed on V1.3.? firmware I had prior)

Probing a serial bus that sends periodic packets, it seemed obvious that the scope was not triggering on every packet. This problem can be reproduced, for example, by setting up the AWG for a 10ms pulse every 1s. Timebase set to 10ms/div. Trigger normal mode, on rising edge, 8ns holdoff,  holdoff start on acquisition start.

When I do this I can see Trig'd LED blinks only once every 2 seconds, or every other pulse. If I set holdoff start on last trigger, It does indeed trigger every second/on every pulse. Am I missing something?

I read the manual and can't really make sense of the difference between holdoff start on acquisition start vs. last trigger, but I wouldn't expect this behavior regardless.