USB- 232/4 and USB- 485/4 hangs

[ranjini]

WE have purchased USB-232/4 and USB-485/4 two numbers each. We use RF and high voltage systems where these modules are also installed. if somwhere some H.V spark occurs , these modules hang and many times uninstalling the ports and reinstalling helps. and sometimes nothing helps even if the PC is restarted. the port does not respond at all. after few days they become usable again.!!! can anyone suggest why this occurs??

thanks 

ranjini

[Wayne.C]

ranjini,

Electrical noise will be generated anytime you have a spark.  The greater the potential behind the spark, the larger and more intense the electrical noise will be.  In your case I would venture that you are creating very large disturbances on either the ac power in your lab or the serial com wiring. 

I'd start by looking at how your cabling is run and make sure you keep the serial wiring seperate from any ac power or HV wiring. 

[ranjini]

Dear Wayne,

It is true that there is a lot of disturbances due to radiated RF in between and some sparks in some other High voltage lines. BUt there are many many instruments that work perfeclty . we used com port of PC without any problem. only these modules are very much prone to these noise. serial cable is not getting the noise since we are using shielded cables. AC power to PC may be getting disturbed. surprisingly only these modules hang??

we were assuimng that these can  withsatnd these intermittent noise and disturbance.

what to do now? it is not easy to avoid these noises....

[Wayne.C]

ranjini,

Sounds like the the USB converters are locking up.  Disturbance can only come from two places, serial cabling or ac power to PC. 

1. You can't completely rule out the serial cabling because it worked with PC com port.  I'm guessing you only had one or two serial cables connected to PC's com port at a time?  Now it sounds like you have more.  Does severity of problem change when you change how many of the ports on the USB adapters are used?  Your serial cables are shielded.  Is shield terminated on both ends, one or none?  This will make a difference and you may have to experiment a bit.
2.  Install a good filter on the ac to the PC.  Make sure the filter is optimized for the kinds of frequencies present.  

This also depends on how well the manufacturer tested the product.  What make and model USB-serial converters are  you using?

[ranjini]

dear wayne,

thank you so much fro the replies. 

shield of the serial cables are ground at one side. we use three serial ports . one with the PC com port which works perfectly. the other are used via NI make USB to serial converters. we have not noticed any relation with number of ports used with the functioning ofthe converters. 

but the problem happens very frequently and happened in two different labs with different set ups. 

we shall try to put a filter to Ac power of PC

what else could be?

[Wayne.C]

ranjini,

If the shield of the serial cables is connected at the USB converter end, then any noise coupled into the shield is being passed through the USB converter to the PC's earth ground.   I would consider making a change.  I would try to terminate the shield directly to a good earth ground and not through the USB converter.

[ranjini]

shield is connected to the field instrument end and not at the PC end.

[TiTou]

I know this thread is a bit old but we have the same problem with a NI USB-232/4 converter.

Roughly once a week, when the application starts, it freezes, no RS 232 communication can be achieved, rebooting everything 3 times doesn't improve anything. Come back the next day... it works again.

I strongly suggest to not use the NI USB-232/4 converter.

[ranjini]

Titou,

i had used an optical isolation between the RS 232 port of  NI USB 232/4 and the instrument. This solved the problem permanently. i am extensivley using the equipment now in the same RF environment.may be u can also try this.

But ofcourse, NI did not respond to any of my complaints. they told that it is the problem with my USB port!!!!!

regards 

Ranjini

[TiTou]

ranjini,

Thanks for the info, but in my case there is no electrical noise 😮

And in fact what happens is that even the RS232 port on the computer can't be used anymore, i wonder if it's not a different issue.

We're waiting for the issue to happen again to do some more tests to find out exactly what it is, just in case we purchase a different USB RS232 converter that we've used on other project and with which we never had trouble. If that new converter doesn't solve the problem then shame on me for ranting... wait & see