FRA GPIB data loss

[gireesh]

Hello

I communicate with FRA(Solartron 1250). When I am reading from FRA using VISA READ palette,at the starting time for example sweep from 0.1Hz to 1KHz, showing data loss like 0.1 then 0.0 next is 0.125 then again 0.0 next is 0.158 next is 0.0 after getting 0.2 or 0.3 it is giving continuous data upto 1KHz. And also some times the while loop is not stopping.

I programmed it as an automatic stop - it taking the 1KHz data and comparing with the given data(numeric DBL value) and it stops. It holds long time.

why the data loss happens, is it due to the timing interval? One time I tried with delay, wait palette with 1.5S it is not corrected. Please give me a solution for the issue.

Thank you 

[gireesh]

Hello

As I noticed that whenever ask questions about FRA solartron or pickering nobody will respond.

Why it  is happening that I don't know? 

[Dennis_Knutson]

I for one, do not understand your question. What exactly do you mean by 'data loss'?

Why don't you attach your code so someone can better understand what you are doing and please provide some details.

[gireesh]

hello denis

i am attaching my Vi. please just go through it. the main problem is in VISA READ palette. it reads data from FRA solartron 1250 as a string. For splitting the data i used scan from string palette. now i can see the frequency, magnitude and phase as DBL output. data loss happens in the DBL output only. At the same time i am saving the data as text file. in the text file it is correct, sometime seperator(comma) or + - signs are missing from the data. thats ok.

as per the solution got from the discussion i made a online plotter using XY graph. It also working fine. i satisfied. the real problem is when the data getting in as DBL output, sometimes it is showing all zero values
for example:-

freq,mag,phase,error
0.1,10.37,-179.2,0
0.0,0.0,0.0,0
0.125,10.36,-178.39,0
0.0,0.0,0.0,0
0.158,10.35,-178.10,0 and so on(please remember this will only in DBL output) in the text data the zero values are not entering.

Message Edited by gireesh on 07-04-2010 03:34 PM

[Dennis_Knutson]

It would seem that occasionally, you are getting a scan error and that is resulting in the all zeroes. You are clearing the error after the scan so this is obviously something you have seen. Instead of blindly clearing the error, you need to determine why the error is happening and fix it. You are doing the same thing after the VISA Read. Since you decided to not attach the text file with the raw data, I cannot determine the cause. I can say that doing a VISA Read in a loop is very, very unusual. The vase majority of GPIB instruments will not continuously send data. You actually have to write a query and then do a read. The instrument is non-standard and I'm not sure how much anyone can help without having the actual instrument.

If you want some more help, remove the clear error function and capture the input string when an error does occur. 

[gireesh]

hello Dennis

i will try this solution. actually i put the clear message palette on this is the VISA READ showing some timout message. this one already i disussed with you in earlier messages. And also scan from string also showing some error message(i forgot now) 

whatever i will try without this clear message and get back to you on this.

thank you for response and solutions.

[gireesh]

hello dennis

I tried without clear error message. But the timeout problem exists. due to that problem the run always in error. the above problems are exists. and one more thing, in the graph i used shift registers as per discussion from this forum the stored graph is not going blank when the next run starts. how to change this.

Please help me 

[Albert.Geven]

The FRA1250 is an old slow instrument. Not bad, just old.

Did you check the bus timing in MAX, you really should use a bus timing of 2 microseconds, not faster.

And maybe you have autopolling enabled, that sometimes also can lead to problems with older systems.

goodluck

[gireesh]

hai

1.   Timed out problem solved by changing the byte count from 1000 to 39. Actually data output of FRA was 39 bytes only.

       Now it is showing OK with some error code.  (I forgot to take the code from office).

2.    what about shift registers? the shifted data is not clearing the next run

thank you 

[Dennis_Knutson]

I'm not sure I would call that a fix. If you have to specify the exact byte count, the instrument is not using the EOI to denote the end of data and you have not properly set up your reads to correctly detect whatever termination character(s) that the instrument is using. An instrument that is compliant with IEEE-488.2 must send EOI and the VISA Read will automatically terminate when that is detected so specifying 1000 bytes would never be a problem. Look in the manual to see if the instrument can be configured to use EOI or read it to see what is the actual termination scheme.