Problem setting baud rate of 3rd party USB-RS232 cable w/ VISA 4.1 and LV 7.0

[rvest]

Hi,

I am trying to configure a USB to RS-232 converter cable to communicate with a serial device that has a fixed baud rate of 28800. I am running LabVIEW 7.0 with VISA 4.1 on MacOS 10.4. The cable has a Prolific 2303 chip set, and I have installed driver version 1.2.1r2 (the current release) from Prolific.

I can succesfully configure and use the cable at lower baud rates (for example, 19200), but when I try to configure the baud rate to 28800 I get a VISA error number -107380733, "The specified state of the attribute is not valid, or is not supported as defined by the resource."

The chip set in the cable claims to do arbitrary baud rates, and VISA claims to do arbitrary baud rates. Why is 28800 an invalid baud rate? Is there a work-around that will let me communicate with my 28800 baud serial device?

Thanks,
-Rob Vest

[mikeporter]

Could it be that you are bumping up against a limit in the chip set, in that it simply can't run that fast? For example, can it run at 38.4k? Even given the ability to produce "arbitrary" data rates, that frequency might not work right due to the way the clock is being generated - though I would be surprised if that were the problem as 28.8k is not that strange of a data rate. In fact, it used to be something of a standard years ago.

Have you tried an adapter from a different vendor? Have you talked the vendor that made this adapter? Does the adapter work at the desired rate on a Windows machine?

Finally, you didn't mention what the device was that you were communicating with, but does it have to run at that particular rate?

Mike...

[rvest]

Hi,

The claim from the chip set maker is arbitrary baud up to 6 million. We just tested it on a Windows box and it configured without error at 28.8kbps, although it did generate the same error I am getting when we tried 38.8kbps. So it seems the chip set is at least capable of 28.8kbps, although not necessarily the "arbitrary" in the manufacturer's claims.

I have tried a cable from another vendor, although I'm not sure what the chip set in that cable is. The result: identical behavior. One cable is from Zonet (Prolific chip set) and the other is from Coolgear (unknown chip set). I think the Coolgear is not a Prolfic chip, because the Prolific driver fails to recognize it.

The device I'm trying to communicate with is a residual gas analyzer from Stanford Research Systems. For reasons known only to SRS, it has a fixed and unalterable baud of 28.8kbps.

My instinct is to suspect that some piece of software in the chain of communications just isn't supporting the 28.8k baud rate, but every component (VISA 4.1 and the chipset) I can identify except the driver claims to do arbitrary baud. Two cables with two different drivers generate the same error, though, so that makes me suspect something that is common to both configurations, like VISA, but that doesn't seem right either. So I'm flummoxed....

Thanks
-Rob Vest

[mikeporter]

Well, if the thing works well under Windows at 28.8k that would tend to indicate an incompatability somewhere in the Mac/Linux OS.

Mike...

[_Bryan]

Hey Rob,

Did you get the RGA working?  We are looking into getting a RGA and I wondered if you liked the SRS unit.

[rvest]

I did get the SRS RGA working. Basically, I can configure the USB to RS-232 converter cable with ZTerm, a terminal emulation program. If you let VISA do anything after that other than read and write to the ASRL resource, it will reset the baud rate to 9600, but it will read and write without resetting the baud rate. The LabVIEW driver from SRS does not work, because it configures the resource. You'll need to modify the driver to take out all the serial configuration calls.

Another work around is to get a new Mac with the Intel processors and install a virtual machine running Windows. Then you can use the Windows cable driver and VISA, which works correctly. It makes it a little tough to integrate the RGA with other software, but that may or may not be a problem, depending on what your application is. In my case, that would work, since I'll just qualify the beamline vacuum and then open the valves and run in the MacOS environment without needing to talk to the RGA again.

The "use a virtual Windows machine" solution is necessary for several other NI parts that do not have MacOS drivers. We have another system that runs of a Mac laptop and uses the USB to GPIB converter. This can not be done under MacOS because NI will not port the driver over to MacOS. I have successfully used the cable under a virtual machine running Windows though. It seems stupid to have to do that. Can't NI port the driver? They already have a Linux version! If anyone from NI is reading this please take note. I've sent this complain up through my local NI representative several times, but I never hear anything back and nothing ever gets done.

-Rob

[_Bryan]

Did you just get a RGA?  Or something like the QMS100?  Are you happy with it?

We are looking into the QMS100 because our sample is at atmospheric pressure and they are half the price of other units on the market.  We are only trying to measure hydrogen so the RGA's are a bit of overkill, but they do the jobe better than leak detectors.

[rvest]

---

@_Bryan wrote:

Did you just get a RGA?  Or something like the QMS100?  Are you happy with it?

We are looking into the QMS100 because our sample is at atmospheric pressure and they are half the price of other units on the market.  We are only trying to measure hydrogen so the RGA's are a bit of overkill, but they do the jobe better than leak detectors.

---

We got the RGA only. We have several of them in use at our synchrotron radiation facility. They work very well for basic residual gas analysis. I don't have a good feel for how quantitative they are out of the box or what it would require to do really accurate residual gas metrology with them. But they are great for checking for vacuum contamination and helping identify the chemical species of the contaminant.

[CoreyNE]

Hi Rob,

Hopefully you are still being alerted to posts on this tag.  I am trying to set up a LabView program to communicate with a QMS 100 and seem to be running into similar problems as you were having using the SRS drivers.  I am using Windows XP and LV 8.5 and the data acquisition program that comes with the QMS works just fine.  However I couldn't even get the basic ID query VI to work.  Do you have any other insights?

Thanks,
Corey

[Coq_Rouge]

I Have used USB-RS232 converters in labview since 7.0. No problem at all. But swaping USB port will give the RS232 port a new adress. This can be very annoying, but is not a problem related to labview. I recomend using MAX for simple port diagnostics.

PS in old threads like this it is better to start a new one.