At last, Siglent's SDS5054X touchscreen DSO...

[bson]

Hmm.  1V 2ns pulse from my HP8133A pulse generator.  Looks like the -3dB point is at ~950MHz with a ~345ps rise time.  (The 8133 has a wonky connector and glitches if jostled, hence the large variance.  Not the scope.)

[bson]

Here's another question... is it possible to make the scope stop confirming that I really want to auto setup when I press the key?  Just do it!

[tautech]

Here's another question... is it possible to make the scope stop confirming that I really want to auto setup when I press the key?  Just do it!

Confirmation to use Autoset was added into the 5kX and 2kX+ new UI a few years back for those that pressed it accidently whereas other Siglent models provide a Undo Autoset. Pick your poison......I never use it as want to be master of instrument and NOT let it make choices for me. 

SDS1104X-E................

[bson]

Does anyone know what the external 10MHz clock levels should be?  Is a ±12V pk sinusoid fine?  I can't seem to find this anywhere in any of the datasheets or manuals...  And I hesitate to plug in a ±12V source if the instrument expects 0-5V or something.

BTW, if Siglent NA is watching: when downloading the SDS5000X User Manual from the documents page (https://siglentna.com/resources/documents/digital-oscilloscopes/#sds5000x-series) I actually get the Service Manual... I found the manual somewhere else through google though, but it would be nice to know that it's up to date.

[bdunham7]

I never use it as want to be master of instrument and NOT let it make choices for me. 

Like shifting your own gears, right?  I have two scopes with continuous Auto Range, and occasionally it is pretty handy.

[bdunham7]

Does anyone know what the external 10MHz clock levels should be?  Is a ±12V pk sinusoid fine? 

You could measure the clock out signal and see what that is--but 5V_rms doesn't seem out of line.  Of course you probably don't want to go guessing on your new $$$ scope.

[tautech]

Does anyone know what the external 10MHz clock levels should be?  Is a ±12V pk sinusoid fine?  I can't seem to find this anywhere in any of the datasheets or manuals...  And I hesitate to plug in a ±12V source if the instrument expects 0-5V or something.

BTW, if Siglent NA is watching: when downloading the SDS5000X User Manual from the documents page (https://siglentna.com/resources/documents/digital-oscilloscopes/#sds5000x-series) I actually get the Service Manual... I found the manual somewhere else through google though, but it would be nice to know that it's up to date.

Use the same spec as Siglent's AWG's:
Reference Clock
10MHz Input
Parameter Min. Typ. Max. Unit Condition & Note
Frequency 9.999M 10M 10.001M Hz
Amplitude 1.4 Vpp
Input impedance 5 kΩ AC coupling
10MHz Output
Parameter Min. Typ. Max. Unit Condition & Note
Frequency 10M Hz Synchronized to internal reference clock
Amplitude 2 3.3 Vpp HiZ load
Output impedance 50 Ω
https://int.siglent.com/u_file/document/SDG6000X_DataSheet_DS0206X-E02B.pdf P13

[2N3055]

Does anyone know what the external 10MHz clock levels should be?  Is a ±12V pk sinusoid fine?  I can't seem to find this anywhere in any of the datasheets or manuals...  And I hesitate to plug in a ±12V source if the instrument expects 0-5V or something.

BTW, if Siglent NA is watching: when downloading the SDS5000X User Manual from the documents page (https://siglentna.com/resources/documents/digital-oscilloscopes/#sds5000x-series) I actually get the Service Manual... I found the manual somewhere else through google though, but it would be nice to know that it's up to date.

Funny enough, on EU site it is showing right...
https://www.siglenteu.com/wp-content/uploads/dlm_uploads/2019/01/SDS5000X_UserManual_UM0105X-E01E.pdf

[luudee]

Guys,

I am trying to save some captured data. But due to it's size, I am forced to save it as binary.

But I can not convert/decode the capture.

Has anybody gotten the Bin2csv tools or the Python script working ?

Any pointers, appreciated !

Cheers,
rudi
 

[tautech]

Guys,

I am trying to save some captured data. But due to it's size, I am forced to save it as binary.

But I can not convert/decode the capture.

Has anybody gotten the Bin2csv tools or the Python script working ?

Any pointers, appreciated !

Cheers,
rudi

Bug !
Save of the BIN2CSV utility to USB stick is not working....reported.
However, grabbed from another model.
Unzip and install on PC.
A while since I last used it so can't exactly remember the details but it was quite simple and all you need to know is in the Readme.txt file.

[luudee]

Thanks Rob !!!

Yeah, it does seem to only work on my old lab computer with windows 7. On my notebook with windows 10 it does not work.

But I am good to go now, thank you so much !!!


rudi

[tautech]

Guys,

I am trying to save some captured data. But due to it's size, I am forced to save it as binary.

But I can not convert/decode the capture.

Has anybody gotten the Bin2csv tools or the Python script working ?

Any pointers, appreciated !

Cheers,
rudi

Bug !
Save of the BIN2CSV utility to USB stick is not working....reported.
However, grabbed from another model.
Unzip and install on PC.
A while since I last used it so can't exactly remember the details but it was quite simple and all you need to know is in the Readme.txt file.

The BIN2CSV utility can only be saved to USB via use of the File Manager accessed from within Save/Recall.
To save BIN2CSV the File Manager Save or Save As virtual buttons must be used.
This is not obvious until you take the time to study File Manager.

The Save and Save As virtual buttons could be improved with a bolder font and/or Save could be highlit to bring the eye to the Save virtual button.

Suggested feature request/improvement to Siglent for SDS5kX and SDS2kX+: make Save more visible in File Manager.

[luudee]

The BIN2CSV utility can only be saved to USB via use of the File Manager accessed from within Save/Recall.
To save BIN2CSV the File Manager Save or Save As virtual buttons must be used.
This is not obvious until you take the time to study File Manager.

The Save and Save As virtual buttons could be improved with a bolder font and/or Save could be highlit to bring the eye to the Save virtual button.

Suggested feature request/improvement to Siglent for SDS5kX and SDS2kX+: make Save more visible in File Manager.

Hello !

Thats exactly what I have done initially. I saved the binary capture, and when selecting
the safe file type, I saw the Bin2CVS utility, and saved it from the DSOs File Manger. I
unzipped the archive, and found an unreadable readme.txt file, as well as the Bin2Cvs
executable and the Python script.

The Bin2CVS from the scope did never work, the one I got from you did work. The
binaries are different.

rudi

[bson]

Somewhat tangentially, it would be extremely nice to be able to grab the entire acquisition record over LXI on ethernet, and not just the decimated display trace data.

[bson]

And on the topic of wishlists in general...

I'd like to be able to give it a "name" that's included with the DHCP request when it obtains a lease.  My NetGate firewall's DHCP server automatically puts these in an internal DNS view. 

Also, it would be nice if it ran an mDNSResponder for this name, so it shows up in the .local domain (zeroconf multicast DNS, on 224.0.0.53).  Even my Keysight 34465A does this... and also includes the name set in its DHCP requests, as mentioned above.  Since the scope runs Linux this is trivial - just run mDNSResponder and allow setting a name in the DHCP config dialog, and add it to the DHCP client config.  Even Windows has an mDNSResponder these days!

Add a "renew lease" button in the DHCP setup dialog.  Very handy for testing configuration changes.  Rebooting is kinda slow and disruptive.

[bson]

The trigger jitter on the AUX OUT is about 2.5ns.  The delay ~680ns.

1MHz in on C1, AUX OUT hooked up to C4 (shortest way around from the back).  Triggered on rising edge of C1.
Stddev of inter-channel delay (FRFF) measurement.

[luudee]

I've been trying to use the Decode Function on a SPI bus I have been debugging.
First I used the analog channels, then I switched to the digital logic analyzer.

The decoder seems to fail in several ways. Either it decodes incorrectly, or fails to
decode all together. I have verified that the analog levels and signal integrity is
pretty good. The SPI clock runs at about 22 Mhz.

Another issue was with exporting the decoded data. It does not work when using
the Logic Analyzer and decoding. I go to the Decoder->Results menu and save
from there. But it saves all zeros, no data at all.

These bugs seem to show up randomly. Sometimes it works, sometimes I get
garbage. Difficult to use an instrument like this when it unreliable.

I hope someone from Siglent pays attention here ...

Cheers,
rudi

PS: Signals are (top to bottom): CLK, CS_, MISO, MOSI

[bson]

Isn't that just a bug in the rendered box with 0x02 in it?  It's a little too short.  This is, assuming sampling on the rising CLK edge; if set to sample on the falling edge then it's not clear what the LSB should be since the trailing edge of MOSI coincides with CLK.

Oh wait, it's early.  It's should be 0x03, right?  Did you set the decoder CPOL+CPHA correctly?

[luudee]

Oh wait, it's early.  It's should be 0x03, right?  Did you set the decoder CPOL+CPHA correctly?

Hi !


yes, it should have been 0x03.  The clock edge is set to rising edge, data is MSB first.

As I said, it works about 70% of the time and I get expected data. But the 30% of the time when I really need it work, it doesn, lol. 


rudi

[nctnico]

Oh wait, it's early.  It's should be 0x03, right?  Did you set the decoder CPOL+CPHA correctly?

Hi !


yes, it should have been 0x03.  The clock edge is set to rising edge, data is MSB first.

As I said, it works about 70% of the time and I get expected data. But the 30% of the time when I really need it work, it doesn, lol. 

You seem to be running SPI at 20MHz. What is the maximum SPI clock rate the oscilloscope supports? It may be your SPI bus is simply too fast for the decoding to work. This can depend on the record length; unfortunately this isn't specified in the manual or datasheet.

[luudee]

You seem to be running SPI at 20MHz. What is the maximum SPI clock rate the oscilloscope supports? It may be your SPI bus is simply too fast for the decoding to work. This can depend on the record length; unfortunately this isn't specified in the manual or datasheet.

Hi,

SPI bus is rated up to 50 MHz.  If the scope says it supports SPI bus decoding. I would expect it to be able to operate withing valid parameters of the SPI specification.

rudi

[nctnico]

You seem to be running SPI at 20MHz. What is the maximum SPI clock rate the oscilloscope supports? It may be your SPI bus is simply too fast for the decoding to work. This can depend on the record length; unfortunately this isn't specified in the manual or datasheet.

Hi,

SPI bus is rated up to 50 MHz.  If the scope says it supports SPI bus decoding. I would expect it to be able to operate withing valid parameters of the SPI specification.

Unfortunately neither statement is correct. First: There is no limit to what speed is supported by SPI. You can run it at several GHz if you manage to get the timing right. Second: every DSO has a limit to the speed it can decode. For example: on my GW Instek it depends on the memory length. With a short memory length it has no problem decoding SPI at 125MHz but at longer memory lengths, the maximum speed drops. In the end a DSO has a limited amount of resources to do the decoding which in turn limits the number of samples it can process within a reasonable amount of time.

[luudee]

Unfortunately neither statement is correct. First: There is no limit to what speed is supported by SPI. You can run it at several GHz if you manage to get the timing right. Second: every DSO has a limit to the speed it can decode. For example: on my GW Instek it depends on the memory length. With a short memory length it has no problem decoding SPI at 125MHz but at longer memory lengths, the maximum speed drops. In the end a DSO has a limited amount of resources to do the decoding which in turn limits the number of samples it can process within a reasonable amount of time.

You are right, there are no hard limits for the SPI bus clock speed. But 20 MHz, is by today's standards low speed.

Also, I have switched to the Logic Analyzer mode, so it should record only one bit per sample, and not 8 bits
per signal. If there are any limitations due to memory or such, that the Data Sheet of the DSO should have
listed them. I have run my test using different time bases, and sampling rates. Short and large captures.
Yes, when I try to record 200mS worth of SPI traffic, it just chokes and doesn't do anything useful. On the
other hand, I can record 200mS worth of analog traces and save them to  flash drive and then decode it
myself.

The fact remains, that the decoder IS BROKEN. It works 70% of the time, and when you really need it does not.

Another thing I forgot to mention, the very first transaction in a capture is ALWAYS decoded wrong.


I would be happy to work with Siglent and provide sample traces if they are interested.

And don't forget, saving digital traces is also broken (saved as all zeros). And yes I have the threshold
set correctly ...


rudi

[bson]

Received the LPA10 LeCroy probe adapter.  Quickly checked it using the 1kHz calibration output on the front.

AP034 - differential probe - fully supported; allows setting the attenuation (the probe uses different hats to change it) and auto zero.  Works like a charm.

ZS1000 - not supported, but works fine with the scope set to 10:1 50Ω.  Plugging it in displays a message, "Got the probe error! Plug in the probe again".  I guess this means it's not recognized.

AP022 - doesn't work at all.
AP020 - doesn't work at all.

The adapter is very solid, and actually pretty sturdy while on the scope.  I had thought it might get a little rickety with the adapter, but it's really perfectly fine.  At some point I'll probably get an SAP1000 though, but it's nice to be able to use the probes I already have.

I suppose I should also try the LeCroy ADPPS, but it's kinda bulky and inconvenient.

[nctnico]

Unfortunately neither statement is correct. First: There is no limit to what speed is supported by SPI. You can run it at several GHz if you manage to get the timing right. Second: every DSO has a limit to the speed it can decode. For example: on my GW Instek it depends on the memory length. With a short memory length it has no problem decoding SPI at 125MHz but at longer memory lengths, the maximum speed drops. In the end a DSO has a limited amount of resources to do the decoding which in turn limits the number of samples it can process within a reasonable amount of time.

You are right, there are no hard limits for the SPI bus clock speed. But 20 MHz, is by today's standards low speed.

Also, I have switched to the Logic Analyzer mode, so it should record only one bit per sample, and not 8 bits
per signal. If there are any limitations due to memory or such, that the Data Sheet of the DSO should have
listed them. I have run my test using different time bases, and sampling rates. Short and large captures.
Yes, when I try to record 200mS worth of SPI traffic, it just chokes and doesn't do anything useful. On the
other hand, I can record 200mS worth of analog traces and save them to  flash drive and then decode it
myself.

The fact remains, that the decoder IS BROKEN. It works 70% of the time, and when you really need it does not.

Another thing I forgot to mention, the very first transaction in a capture is ALWAYS decoded wrong.


I would be happy to work with Siglent and provide sample traces if they are interested.

And don't forget, saving digital traces is also broken (saved as all zeros). And yes I have the threshold
set correctly ...

It would be interesting to find out whether the same behaviour can be observed on the SDS2000k+%$@*