Weird characters received from instrument using serial communication in labview

[altenbach]

We don't care about the program executable. How did you generate the pictured data? I assume using your "controller.vi" program from your first message.

[aj_dee]

This is what I got when I ran the code.

[aj_as]

This is what I had when I run the code. Earlier message was sent from my lab's PC.

[Bob_Schor]

Arrrggghhh!  NO PICTURES, please!  Save the bytes you are getting in a (binary) file, and attach the file!  We need to see the data, not some (unknown) program spewing out largely unreadable characters.

Bob Schor

 

 

[aj_as]

File attached.

[aj_as]

I was prevented from attaching the .txt file on here so I copied and pasted in word attached but they appear modified in the word version

[altenbach]

You can zip any file and attach it here.

[aj_as]

zipped file attached

[Bob_Schor]

Yes!  Real data!  Patterns begin to emerge!

What can you tell us about this stream of bytes?  Is this a "snapshot" of streaming data, by which I mean "data are coming in (at some rate) and this is an arbitrary sample of 337 continuous arrays of 8-byte records"?  I can see some periodicity in it, perhaps suggesting two "cycles" of about 160 x 8 byte data.  It does appear to be mainly numeric, mainly integer.  Haven't tried tricks like bit-reversal, but there are some intriguing periodicities in this sample.

Some questions:

• These data are coming from "something".  What do you think these data represent?
• Is this a "complete sample" of data, that is, can you say something like "This is a complete data set, obtained by running my program for 1 hour"?
• The opposite, of course, is "This is a snapshot of 10 milliseconds of data just chopped out of a much longer data stream".  Is this a closer description?
• What do you expect these data to represent?  How many channels of signals are represented here?   [It looks, to me, like 1].  What can you say about the sampling rate?  What can you say about the sampling time represented by these data?  [Note if I'm right about my guesses about these data, I would expect sampling rate * sampling time to be around 400].

A few years back, I was handed some 9-track "computer tapes" (back in the days of CDC and DEC computers saving data on "mag tape") and was given the puzzle of deciphering what the (scientific) data were on those tapes.  Fortunately, they were a mix of integer, reals (32-bit, or as we called it, "single precision", and (as we called it) "ASCII".  Arrays could be recognized by the (now I don't remember if it was a 2-byte or 4-byte) header that gave the size.  We managed to resurrect the data, as I recall ...

Bob Schor

[altenbach]

The data is highly repetitive with a period of 8 bytes.

The first two are mostly x0003, followed by two highly variable bytes followed by four bytes that are mostly x01000029. (very few exceptions, see below.

Two exceptions where it does not start with x0003:

and five exceptions where it does not end in x0029, but in x0028:

and two exceptions where the third group is not x0100

Obviously, we need to know much more about all this. Do you have the parsed table data for exactly the same dataset?