Moxa USB Serial Converter fails to communicate with LabWindows for the very first connect and after long breaks in communication

[mowgli]

Hi NI,

We are using a Moxa USB to Serial converter UPort 1130 for communicating with our ECU's over RS-422. A peculiar problem seems to show up while connecting for the very first time. We can open the port successfully using Docklight or Hyperterminal, however LabWindows refuses to communicate with the ECU for that very first packet exchange. We are using OpenComConfig for opening the Virtual Com Port and ComRd+ComWrt for the data exchange. We also close the port after our communication sequence is done. This phenomenon is repeated if we do not test for say 2 hours and try to reconnect using the Moxa. The odd part is that once the communication is established it never dies...however for the very first packet exchange and attempting to connect after long breaks in communication we find that anomaly. Has anyone noticed something like this and can someone help us with a solution ?

Regards,

-Ashish

[menchar]

I am seeing OpenComConfig problems with a virtual com port on a framegrabber.

I have had problems with USB-rs232/422 converters as well - we just gave up and used an NI RS-422 serial i/o card, that way we figured it would work (NI CVI serial library running on top of an NI serial driver on NI serial HW) and it does.  There are many USB - rs232/422 converters that are junk - there are three or four good ones available, NI is one of them.  You may have one of the poorly designed converters (the Moxa).  There's info on the web in re USB-serial converters, someone did a survey of the various brands.  You might try searching the KB for this too as we talked about it before. 

But you never know - of course NI makes sure their serial library / drivers work with their serial HW, but as soon as you step away from the all - NI scenario, if anything goes wrong you're in a finger-pointing contest.  NI has had problems with the serial library, they had a race condition at one point with queue flushes, plus whatever other problems they don't publicize 😉  I bet if you use an NI USB-rs422 converter it will work OK 😉

And I've had nothing but grief trying to run the CVI serial library on a DALSA framegrabber virtual com port - and it looks like it hangs up in OpenComConfig, though it's hard for me to to imagine how this could be.  We noticed that if you try to manipulate the modem control lines on the virtual com port it breaks it.  Seems sensitive to queue sizes too.

good luck.

[menchar]

As a follow up, rs-422 doesn't have the modem controls so kindof makes sense not to try it if it's supposed to be rs-422.  And the Moxa looks like it may be a well done design - though I've never used one of these - the ones I was disparaging are the $20 ones.

I see they offer a driver to go with it - have you tried using their Windows driver?

[mowgli]

Hi All,

We ahve tried the windows driver...it seems to work fine after the first attempt made through docklight.

Please find attached two log files. These have been generated usin the "PortMon" windows utility. Can someone at NI please interpret the data and help us understand why the LabWindows environment is not letting us connect at 460800 bps for the every first attempt ?

The Test.log file is generated in the following environment:

      PC running LabWindows GUI-----USB-----UPort-----RS-422----Device

The Test2.log file is generated in the following environment:

      PC running Docklight -----USB-----UPort-----RS-422----Device

The PC is running Windows 7 as operating system.

We have also tried the same on Windows XP operating system.

Please let us know if you need any more information from our end to process this query.

Regards,

-Ashish