AnsweredAssumed Answered

MSO9254A values mismatch between OSC and imported data with MATLAB

Question asked by guillermo_bcn on Feb 14, 2013
Latest reply on Sep 2, 2013 by algoss

Like • Show 0 Likes0
Comment • 7

HI there,

I'm interfacing a MSO9254 with matlab 2010b using SCPI commands and I read estrange values for OSC configuration parametres.
I receive DataRaw correct but when reading: +preambleBlock = query(visaObj,':WAVEFORM:PREAMBLE?');+
I gate strange values for Time/div, Volt/div, OrigenX,...
I see on my *OSC Volt/div = 50.0 mV* and in *matlab 0.053957427200000 Volt/div*
Also *OSC Time/div = 10ns* and in *matlab = 1.001000000000000e-08*

It could be that the real values are the ones read by matlab, and the ones shown at the scope are approximations?
If thats correct, is there some where on the scope where I could see real values?
I'm working on High energy Physics so 3mV/div error is important.

Thanks,
Guillermo

No one else has this question

Outcomes

7 Replies

algoss Feb 14, 2013 10:35 AM
Mark Correct
Correct Answer
First, make sure that you have the latest version of the scope SW, version 4.10.0010, available for free from the Agilent web site.

I have a MSO9404A and it works correctly. The value it is sending is the same as what's on screen. I don't have MATLAB immediately available, but I may be able to test it later today. If you can send a short section of standalone MATLAB script, I can give it a try. I have used this before and never seen any problems.

You need to verify exactly what the scope is returning because it may just be a problem with the translation going into MATLAB.

You should have the IO Libraries loaded on the same PC as MATLAB. If you don't have it, download it from www.agilent.com/find/iosuite. With that there's an interactive IO capability, so you can send a command like "chan1:scale?", and see the transmitted result.

You can also turn on the debug mode in the scope. Use Utilities -> Remote Setup... then "Debug". You can turn on a mode where all commands, and the results will be displayed on the screen. It's a little ugly, but it works. If the values the scope is sending are correct, but MATLAB is different, then it's a MATLAB problem. Change your script so you read the values into a text array.

Also, be sure that you're looking at the correct values in the preamble. For instance, if you are using 16 bit data (the best), the Volts/div = 2^16 * YIncrement / 8.

Here's some code that I wrote to get and use the preamble block:

preambleBlock = query(inst,':WAVEFORM:PREAMBLE?');
fprintf(inst, '*OPC?'); Junk = str2double(fscanf(inst));
fprintf(inst,':WAV:DATA?');
ScopeData.RawData = binblockread(inst,'int16');
Junk = fread(inst,1,'schar'); % Read the EOF character
maxVal = 2^16;
preambleBlock = regexp(preambleBlock,',','split');
ScopeData.Format = str2double(preambleBlock{1});     % This should be 1, since we're specifying INT16 output
ScopeData.Type = str2double(preambleBlock{2});
ScopeData.Points = str2double(preambleBlock{3});
ScopeData.Count = str2double(preambleBlock{4});      % This is always 1
ScopeData.XIncrement = str2double(preambleBlock{5}); % in Time
ScopeData.XOrigin = str2double(preambleBlock{6});    % in Time
ScopeData.XReference = str2double(preambleBlock{7});
ScopeData.YIncrement = str2double(preambleBlock{8}); % Voltage
ScopeData.YOrigin = str2double(preambleBlock{9});
ScopeData.YReference = str2double(preambleBlock{10});
ScopeData.VoltsPerDiv = (maxVal * ScopeData.YIncrement / 8);      % V
ScopeData.Offset = ((maxVal/2 - ScopeData.YReference) * ScopeData.YIncrement + ScopeData.YOrigin);         % V
ScopeData.SecPerDiv = ScopeData.Points * ScopeData.XIncrement/10 ; % seconds
ScopeData.Delay = ((ScopeData.Points/2 - ScopeData.XReference) * ScopeData.XIncrement + ScopeData.XOrigin); % seconds

Al
Like • Show 0 Likes0
Actions
- guillermo_bcn @ algoss on Feb 14, 2013 12:05 PM
  Mark Correct
  Correct Answer
  Hi algoss, thanks for the fast answer.
  
  Ok, this OSC has soft revision 02.02.0000, maybe I should upgrade a little, but what it needs is a fresh install because it has been used in so many parojects, thanks for the advice.
  
  I use similar code as you do, I take it from a mathworks page and change because there were some +E113 errors Undefined header+.
  
  Next I paste the important part of the code, the debug and the questions.
  
  CODE
  +
  fprintf(visaObj,':CHAN4 ON'); % Display on chan4
  fprintf(visaObj,':WAVEFORM:FORMAT WORD'); % Get the data back as a WORD
  fprintf(visaObj,':WAVEFORM:BYTEORDER LSBFirst');% Set the byte order on the instrument as well
  preambleBlock = (query(visaObj,':WAVEFORM:PREAMBLE?'));% Get the preamble block
  preambleBlock = regexp(preambleBlock,',','split');% split the preambleBlock into individual pieces of info
  fprintf(visaObj,':WAV:DATA?'); % commmand to read data
  waveform.RawData = binblockread(visaObj,'int16'); % read back the BINBLOCK with the data in specified format and store it in
  operationComplete = str2double(query(visaObj,'*OPC?'));
  while ~operationComplete
      operationComplete = str2double(query(visaObj,'*OPC?'));
  end
  
  %fread(visaObj,1); % the waveform structure. FREAD removes the extra terminator in the buffer
  for eindex=1:1:30
      ArrayError{:,eindex} = query(visaObj,':SYSTEM:ERR? STRING');
  end
  
  fclose(visaObj); % Close the VISA connection.
  fprintf('close MSO9254 \n')
  
  maxVal = 2^16;
  
  % split the preambleBlock into individual pieces of info
  preambleBlock = regexp(preambleBlock,',','split');
  
  % store all this information into a waveform structure for later use
  waveform.Format = str2double(preambleBlock{1});     % This should be 1, since we're specifying INT16 output
  waveform.Type = str2double(preambleBlock{2});
  waveform.Points = str2double(preambleBlock{3});
  waveform.Count = str2double(preambleBlock{4});      % This is always 1
  waveform.XIncrement = str2double(preambleBlock{5}); % in seconds
  waveform.XOrigin = str2double(preambleBlock{6});    % in seconds
  waveform.XReference = str2double(preambleBlock{7});
  waveform.YIncrement = str2double(preambleBlock{8}); % V
  waveform.YOrigin = str2double(preambleBlock{9});
  waveform.YReference = str2double(preambleBlock{10});
  waveform.VoltsPerDiv = (maxVal * waveform.YIncrement / 8);      % V
  waveform.Offset = ((maxVal/2 - waveform.YReference) * waveform.YIncrement + waveform.YOrigin);         % V
  waveform.SecPerDiv = waveform.Points * waveform.XIncrement/10 ; % seconds
  waveform.Delay = ((waveform.Points/2 - waveform.XReference) * waveform.XIncrement + waveform.XOrigin); % seconds
  +
  
  DEBUG
  +
  :CHAN4 ON
  > :WAVEFORM:FORMAT WORD
  > :WAVEFORM:BYTEORDER LSBFirst
  > :WAVEFORM:PREAMBLE?
  < 2,1,2002,1,5.0000000E-11,-2.1927192383E-08,0E+00,6.58660E-06,1.23823E-01,0E+00,
  > :WAV:DATA?
  < #44004 >ü;t; <²Š= =*
  >
  ? ^
  ? -410, Query INTERRUPTED
  > *OPC?
  < 1$
  > :SYSTEM:ERR? STRING
  < -410,"Query INTERRUPTED; "$
  > :SYSTEM:ERR? STRING
  < 0,"No error"$
  > :SYSTEM:ERR? STRING
  +
  
  QUESTIONS:
  
  Ok,now I understand some things, so the important values ar X and Y increment.
  but these increments are calculated with all the visible windows of the OSC or the signal, I think is the first option but I tried some things and I'v been unable to create a simple equation for calculating.
  
  If I use *fread(visaObj,1);* I get a *Warning: VISA: A timeout occurred *, so I deleted and all worked fine, but the thing is I've seen these fread on numerous examples in different program languages. Whats happening here?
  
  The last question, why I have a *E-410 Query INTERRUPTED* with the +fprintf(visaObj,':WAV:DATA?');+?
  I tried to modify buffer, and others but always have same error, witch doesn't interfer with the capture.
  Despite all these error I capture the signal dataraw correctly.
  
  Thanks,
  Guillermo
  Like • Show 0 Likes0
  Actions
  - algoss @ guillermo_bcn on Feb 14, 2013 12:42 PM
    Mark Correct
    Correct Answer
    Unfortunately I won't be able to provide much help until you upgrade the scope to the latest version of SW. That is version 4.10.0010, and it is available, for free, from the Agilent Web Site. There have been many changes, and there's just no way for me to be sure what's going on until you upgrade.
    
    The upgrade process will be relatively easy, just download the latest SW, and install it. It will probably require at least 2 reboots as various internal parts get updated
    
    If you have any applications on the scope, then those will need to be downloaded and updated, also. The new SW will tell you about anything else that needs to be updated.
    
    You should also make sure that you have all of the latest Windows updates and Service Packs. You should probably install those first.
    
    You will probably need to run a user calibration before you can use the scope, after the upgrades. If one is needed the scope will keep you from making any measurements until you are calibrated.
    
    Good Luck,
    
    Al
    Like • Show 0 Likes0
    Actions
    - guillermo_bcn @ algoss on Feb 14, 2013 5:28 PM
      Mark Correct
      Correct Answer
      Hi again,
      
      I have update windows and agilent software to the last version: 4.10.0010.
      And still have the same problem.
      I thing the YIncrement is relevant to the signal and no to the visible window.
      
      Another question, how can I read the current Average?
      I'm doing 65000 averages, so I need some way to read out the status of the averaging for pause the matlab program until average is complete, I've tried some querys of the programmer guide unsuccesfull, I'm only able to read the same AVG I write.
      +fprintf(visaObj,':ACQUIRE:AVERAGE:COUNT 65000');+
      
      Thanks again,
      Guillermo
      Like • Show 0 Likes0
      Actions
      - algoss @ guillermo_bcn on Feb 15, 2013 10:30 AM
        Mark Correct
        Correct Answer
        YIncrement is related to the A/D steps. I had the calculation to turn that into Volts/Div in my original code.
        Using WORD:
        maxVal = 2^16;
        ScopeData.VoltsPerDiv = (maxVal * ScopeData.YIncrement / 8);
        
        XIncrement works much the same way:
        ScopeData.SecPerDiv = ScopeData.Points * ScopeData.XIncrement/10;
        
        For getting the "average" information, you need to look at the :DIGitize command, and at "acquisition Synchronization" in the manual. :DIGitize is a macro, and will stop your program until the current acquisition is complete. If averaging is turned on, then it waits until the full number of acquisitions has been captured and averaged. You might want to use Polling instead, since that allows a timeout, so you rprogram won't hang forever is there's a problem with the signal.
        
        Al
        Like • Show 0 Likes0
        Actions
        guillermo_bcn @ algoss on Sep 2, 2013 6:29 AM
        Mark Correct
        Correct Answer
        Hi algoss,
        
        I'm back, I've been busy with other stuff.
        Back to the last post, I've solved everything except the -410 query interrupted warning/error.
        first of all, what makes the OSC generate this error(-410)?
        why is allways just after a fprintf(visaObj,':WAV:DATA?'); ?
        
        Thanks,
        Guillermo
        Like • Show 0 Likes0
        Actions
        algoss @ guillermo_bcn on Sep 2, 2013 7:41 AM
        Mark Correct
        Correct Answer
        I have seen that, too, on some code I've written. Basically it means that you issued a new query before the buffer was cleared from the previous query. If, for instance, you ask for 5 bytes of data and only provide a buffer for 4, then ask for more data, you will see that error. Some commands cause the scope to send one extra character, basically and "End of Data", and if you don't account for that, you'll see an error.
        
        That said, I have some code that always generates that error, but it otherwise works perfectly, so I just ignore it.
        
        Al
        
        Disclaimer: For more reliable response, you should consider calling your local Agilent Technical Call Center. The Agilent Forums are monitored on an "as available" basis, and aren't necessarily the fastest way to get technical questions answered.
        Like • Show 0 Likes0
        Actions

MSO9254A values mismatch between OSC and imported data with MATLAB

Outcomes

Related Content

Recommended Content