Siglent Technical Support join in EEVblog

[tautech]

Does the SDS2000X series (I'm interested in the SDS2204X) support LXI on the LAN port?  If so, is the SCPI command set documented somewhere?

Yes.
The Programming guide is in the Documents page within the Product page on the Siglent websites:
http://siglentamerica.com/prodcut-wd.aspx?id=1488&tid=1&T=2

[nctnico]

Hello,
will there be the open socket LAN connection also for the SDS2000X series?

regards René

No, there won't be open socket on the SDS2000X but there will be on the new SDS1202X-E.

[boggis the cat]

What advantage does LXI give you over VXI-11?

I haven't used any of the LAN based communications options with test instrumentation.  Both protocols seem to offer SCPI / IEEE 488.2 support.  LXI appears targeted more toward very large systems, such as automating production processes.

[Siglent America]

The Siglent SDS2000X oscilloscopes are VXI-11 compliant but technically not LXI compliant.

If you want to program a SDS2000X oscilloscope using remote control over LAN you will need to use the NI-Visa layer and I would certainly want to have a copy of the Programming Guide handy. There are some example programs in the guide.

Please contact us at info@siglent.com if you have any questions.

[boggis the cat]

So the difference is ease of use, essentially?

Installing the bloated NI Visa software and getting it working is something that isn't a lot of fun.  If LXI let's you avoid that then I can see the advantage.

[nctnico]

So the difference is ease of use, essentially?

With LXI you can open a TCP/IP socket (telnet) and start sending commands. With VXI you'll need the bloated VISA to do the translation from VXI to text commands. LXI is way easier to incorporate in your own programs than VXI.

[kvd]

Hi, first time post on the EEVBlog forum.

I discovered a bug on the SDS1202X-E when using cursors in combination with a 10X probe. I don’t know if it has been reported before. Reading through 27 pages of posts is a bit too much.

Using a 10x probe and having correctly set the probe adjustment under the channel/probe menu doing a measurement with the Y-cursor gives a wrong readout. It divides by 10 one time too much. The same measurement done with the measurement button, and then for example ‘maximum’, ‘amplitude’ or ‘peak-to-peak’ gives the correct readout, properly corrected for the 10x probe. See attached picture for the bug.

Software Version: 5.1.3.8
FPGA version: 2017-03-21
Hardware Version: 00-01
Product Type: SDS1202X-E
Serial No:
Scope ID:

I discovered another bug, which I will post soon. It requires a bit more writing and images.

It is a great scope otherwise!

Koen

[kvd]

This is a description of the other bug which I found. It is about X-cursor readouts in the FFT spectrum, in combination with the zoom feature. On the attached two pictures the bug is clearly visible.

In the first picture (E3) you see a minimal zoom, and the ?X readout is 20.26 kHz, which compares fine with the 12.2 Hz/div for the FFT spectrum. In the second picture (E6) you can see that when using the zoom feature (zooming out) the same cursors give a ?X readout of 50.54 kHz, which cannot be right watching the 12.2 Hz/div of the FFT spectrum.

Hope this can be fixed in the next firmware version.

Koen

[colorado.rob]

So the difference is ease of use, essentially?

With LXI you can open a TCP/IP socket (telnet) and start sending commands. With VXI you'll need the bloated VISA to do the translation from VXI to text commands. LXI is way easier to incorporate in your own programs than VXI.

VXI is fine for me.  I use PyVISA and PyVISA-py for this.  Works like a charm with the Siglent SDG2042X.  A couple lines of Python code and you're off and running.  Not as nice as LXI, but usable.

[Siglent America]

So the difference is ease of use, essentially?

With LXI you can open a TCP/IP socket (telnet) and start sending commands. With VXI you'll need the bloated VISA to do the translation from VXI to text commands. LXI is way easier to incorporate in your own programs than VXI.

Yes, nctnico is correct about this.

Other Siglent instruments will have the open sockets - the SSA3000X spectrum analyzers do now -  but the SDS2000X scopes will not.
FWIW, the majority of customers we work with are okay with the Visa layer but we do understand that it is a large program and can slow things down some.

[nctnico]

So the difference is ease of use, essentially?

With LXI you can open a TCP/IP socket (telnet) and start sending commands. With VXI you'll need the bloated VISA to do the translation from VXI to text commands. LXI is way easier to incorporate in your own programs than VXI.

Yes, ntnico is correct about this.

Other Siglent instruments will have the open sockets - the SSA3000X spectrum analyzers do now -  but the SDS2000X scopes will not.
FWIW, the majority of customers we work with are okay with the Visa layer but we do understand that it is a large program and can slow things down some.

Just a short addition: Needing to install NI VISA makes it harder to distribute software to customers especially for small companies who can't spend a lot of time on support. Communicating with equipment over RS232 or LXI allows a piece of software to be installed as a single executable without needing third party libraries and reduces the number of ways an installation can be messed up, reduces OS dependancy and in general reduces the number of failure points. From a programmers perspective NI VISA may look easier but from a support perspective it is not especially if the customer doesn't have IT staff on-site.

[colorado.rob]

Other Siglent instruments will have the open sockets - the SSA3000X spectrum analyzers do now -  but the SDS2000X scopes will not.
FWIW, the majority of customers we work with are okay with the Visa layer but we do understand that it is a large program and can slow things down some.

 

I'd like to encourage you to continue to expand support for LXI.

It's not just a matter of slowing things down.  It's a matter of being able to easily integrate with modern tools such as scipy.

It is certainly a factor which impacts my purchasing decisions.

[Siglent America]

Other Siglent instruments will have the open sockets - the SSA3000X spectrum analyzers do now -  but the SDS2000X scopes will not.
FWIW, the majority of customers we work with are okay with the Visa layer but we do understand that it is a large program and can slow things down some.

o

 

I'd like to encourage you to continue to expand support for LXI.

It's not just a matter of slowing things down.  It's a matter of being able to easily integrate with modern tools such as scipy.

It is certainly a factor which impacts my purchasing decisions.

Thanks for your inputs, colorado.rob. We appreciate it and I can say that it is on the radar.

[boggis the cat]

With LXI you can open a TCP/IP socket (telnet) and start sending commands. With VXI you'll need the bloated VISA to do the translation from VXI to text commands. LXI is way easier to incorporate in your own programs than VXI.

...

Just a short addition: Needing to install NI VISA makes it harder to distribute software to customers especially for small companies who can't spend a lot of time on support. Communicating with equipment over RS232 or LXI allows a piece of software to be installed as a single executable without needing third party libraries and reduces the number of ways an installation can be messed up, reduces OS dependancy and in general reduces the number of failure points. From a programmers perspective NI VISA may look easier but from a support perspective it is not especially if the customer doesn't have IT staff on-site.

Yes, NI Visa is a PITA to install and get working.  Then it seems to be choosy about whether it will keep working...  LXI is obviously a cleaner solution.

Once you bite that bullet, though, does LXI deliver anything significant that is missing from the VXI-11 / NI Visa route?  I noticed that there are some additional functions supported in later implementations of LXI such as the timing synchronisation -- that particular function could free up triggering between instruments, I guess?  (Dependent upon what you're doing, obviously.  I am thinking of aligning manually captured data here, rather than triggering at a set time.)

I am only familiar with the old serial and GPIB instrument control implementations, so the VXI / LXI (and there is PXI, and other stuff) options are new to me.  Is there a good guide that you know of?

[Siglent America]

Hi boggis the cat.

Personally (and from my experience here and prior test equipment companies), installing and using NI Visa on our test equipment has always been straight forward and starts up the first time with no problems. Of course, I can't speak for everyone but that has been my (our) experience.

I found a couple of NI Visa guides you might check out:
http://www.ni.com/pdf/manuals/370423a.pdf
http://www.ni.com/pdf/manuals/370132c.pdf

 

[nctnico]

Try to use a Windows 2000 or Windows XP PC! Lots of these around in labs and people are hell bound on keeping them because they often run software which won't run on newer Windows versions.

[klaff]

I want to report a possible bug on the SDS1202X-E oscilloscope. Serial decoding doesn't appear to work correctly for History waveforms.  More specifically, if I capture a number of CAN waveforms, the CAN decode display will read the value of the last waveform, not the selected waveform.

[kvd]

Hi,

I found yet another bug on the SDS1202X-E.

1. Feed a sine wave into channel B, Vmin = 600mV, Vmax = 27V
2. Press Auto Setup
3. Turn the vertical adjustment knob twice to the right, so the scale is at 1 Volt per division (note that the purple zero indicator is below the bottom of the wave)
4. And now, turn the vertical position up. Now it is perfectly possible to move the zero indicator to far above bottom of the wave! This is clearly a bug.

Note that if the vertical scale is changed to 2 Volts per division instead of 1 Volt per division, the vertical adjustment works as expected.

Also note that a sine wave on channel A, Vmin = 800mV, Vmax = 10.2V does not show this bug.

Also note that after step 2 on each subsequent turn of the vertical scale knob Vmin gets another value.

Koen

Software Version 5.1.3.8
FPGA version: 2017-03-21
Hardware version: 00-01
Product Type: SDS1202X-E
Serial No:
Scope ID:

[kvd]

After having delved deeper into the last mentioned bug on the SDS1202X-E I have narrowed down the cause of the bug.

Taking a signal with a DC component and using DC coupling on the probe, and switching to different vertical scales the bottom of the wave seems to have rather arbitrary values.

Note: I have done self calibration twice on this scope.

With AC coupling on the probe I have not yet found these issues.
It happens on both channels, so it is not an issue of channel 2 only

Am I overseeing something? I am not an experienced oscilloscope user, but this behaviour of the scope baffles me.

[nctnico]

At first glance this looks like you are overdriving the input.

[tautech]

Am I overseeing something?

Not really.
There is a bug with measurements as you have identified and it is being addressed at the factory.

I am not an experienced oscilloscope user, but this behaviour of the scope baffles me.

Measurements are predominantly made from display information so to not have all the waveform on the display will result in corrupted values returned. There is a tradeoff between amplitude and frequency measurement accuracy depending on how you use the scope.  Same for most DSO's.

This thread investigates some of what you see:
https://www.eevblog.com/forum/testgear/testing-dso-auto-measurements-accuracy-across-timebases/

In this post member rf-loop gives examples for the SDS1202X-E:
https://www.eevblog.com/forum/testgear/testing-dso-auto-measurements-accuracy-across-timebases/msg1204964/#msg1204964

[boggis the cat]

Measurements are predominantly made from display information so to not have all the waveform on the display will result in corrupted values returned. There is a tradeoff between amplitude and frequency measurement accuracy depending on how you use the scope.  Same for most DSO's.

Most instruments will detect out-of-range signal conditions and not display erroneous calculated results.  Is there a reason for not taking this approach?

[ebastler]

Measurements are predominantly made from display information so to not have all the waveform on the display will result in corrupted values returned. There is a tradeoff between amplitude and frequency measurement accuracy depending on how you use the scope.  Same for most DSO's.

Most instruments will detect out-of-range signal conditions and not display erroneous calculated results.  Is there a reason for not taking this approach?

Let's not go there, shall we?
[deep voice and echo on] Summoning MrWolf!!! [/deep voice and echo] 

[kvd]

Regarding the Siglent SDS1202X-E:

I noticed the following bug/inconvenience: in the trigger settings, type 'Slope' the time setting for the "Limit Range <=" is not preserved on shutdown and powering up of the scope. As far as I have noticed thus far, all settings are preserved, although I am not sure of this of course. It would be nice if this would be so.

But: the SDS1202X-E is an awesome scope!

Koen

[boggis the cat]

The best approach is to save setups, then load them before starting work.  Unless you have an excellent memory there are bound to be things set differently than you recall, so starting from a 'clean slate' with a known state is the safest approach.

If you're doing any critical amplitude measurements, allow for a warm up time then run the self cal routines.  They take less than five minutes, and can improve the accuracy a great deal.