07-13-2010 09:54 AM
I have three identical (hw/sw) test systems, that were built by a third-party vendor. Recently, one of the systems randomly reads the incorrect number of bytes through the serial port (expect 17 bytes, read 18 bytes). It happens about once every 20 reads. I have a serial "sniffer" program that intercepts the serial data; and the message possesses a consistent 17 bytes. So the LabVIEW code must be stuffing a byte in somewhere. This displacement occurs within the first few bytes of the message. The weird thing is, it used to work fine. Is there something in the setup that could have changed (i.e. - what could I check)?
07-13-2010 10:17 AM
More information: A "FF" is getting stuffed into the first byte of the message.
07-14-2010 01:52 PM
Hi ,
What is the environment you are working with, is it LabVIEW ? Also, what is the version of the SERIAL driver you are working with?
Thanks and have a great day.
Esmail Hamdan | Applications Engineering | National Instruments
07-14-2010 02:28 PM
This sounds like a hardware problem to me - can you switch the serial cable with a known good test system and/or move the cable to a second serial port and see if the problem tracks the cable or if the problem goes away?
Thanks,
Jim
07-14-2010 03:06 PM
@TSJim wrote:
This sounds like a hardware problem to me - can you switch the serial cable with a known good test system and/or move the cable to a second serial port and see if the problem tracks the cable or if the problem goes away?
Thanks,
Jim
I agree.
If you have three identical systems then you should be able to swapp things around to figure what went bad.
Still hardware related but there is another possiblility. If something changed (new load on AC mains injects noise, new flourescent fixtures near wire,...) you may have noise where it wasn't previously.
I would break out an o-scope before I messed with any diagrams.
Ben