Electric interference on USB?

Technaton

New member
Dear forum members,

the same CFA-635 I had mentioned in my previous post now sits happily in a wooden enclosure. The same enclosure houses a computer - actually, the whole case. There are some fans and air filters to keep the interior cool and free of dust, but otherwise, there is nothing. I did a test run with a laptop, where the setup ran fine with lcdproc for about a day. Now, with the computer inside and the TFT's USB connected to it, it won't display a thing. Instead, I get the following messages in my kernel log:

Code:
usb usb3-port2: disabled by hub (EMI?), re-enabling...
usb 3-2: USB disconnect, device number 5
usb 3-2: new full-speed USB device number 6 using ohci-pci
usb 3-2: New USB device found, idVendor=223b, idProduct=000b, bcdDevice= 2.00
usb 3-2: New USB device strings: Mfr=1, Product=2, SerialNumber=3
usb 3-2: Product: Crystalfontz CFA635-USB
usb 3-2: Manufacturer: Crystalfontz America, Inc.
usb 3-2: SerialNumber: 0030001820353233
cdc_acm 3-2:1.0: ttyACM0: USB ACM device
I already tried changing the ports, but to no avail. So I'm wondering whether it could really be EMI. The CFA-635 has everything attached that would be needed to place it in a 5.25" drive bay, but the back is "exposed" to the interior of the case. (But the computer standing there is still in its original casing.) Does the TFT need some sort of Faraday's cage? Or are any other USB issues known that could be it? Google didn't help me find anything.

Thanks a lot in advance for any hints.
Looking for additional LCD resources? Check out our LCD blog for the latest developments in LCD technology.
 

CF Tech

Administrator
Thanks for your message.

Wood is not an ideal case material from an EMI-RFI-ESD standpoint, but I have had many of these displays running unprotected on the bench and have not seen any recurrent issues as you describe. It should not be particularly sensitive. I mean you do not want to shuffle across the carpet and give it a doorknob zap, but it should run OK on the bench without a case.

Was it working well before you installed it in the case?

Could the USB connector on the LCD be mechanically damaged when installing it -- maybe from pulling or bending on the cable?

But if the software reconnects -- as your log indicates -- that would point away from the connector.

Maybe try a different cable? You already tried a different port, right?
 

Technaton

New member
Thanks for your quick reply!

I did the 24h test with everything fully assembled. The only difference between the test and now is that the test was run with a laptop, while now a desktop PC tower sits in the case. Yes, really, its a wooden box with another PC case in it - the PC is not even disassembled (I wanted to do this, but I haven't gotten to it yet). So I'm wondering where the EMI might actually come from.

The TFT's USB cable seems undamaged from the outside. Pin connector is ok, everything fits well.

I have tried a different USB port, but not a different cable as I have none. I was suspecting that the host's hub might be the issue, but the same port the TFT was connected previously to now has an 1090 MHz receiver connected to it, which streams airplane transponder data happily since two days now, without a reset. And that one has a quite long cable (7m).

Are there any known quirks wrt. different USB host controllers? Anything I could look up and try, like a kernel parameter?
 
USB LCD Displays - Graphic and Character LCDs with a Keypad

CF Tech

Administrator
The USB on those modules has always been rock solid. I can't recall any USB trouble that was not related to some software (driver) issue (all of which were all resolved years ago).

All I can think of is doing "swaptronics":
  • Does the LCD still work fine with the laptop ?
  • Does the LCD work on the desktop system when it is the only thing plugged in ?
  • Does the LCD work on the desktop system if everything is connected except the receiver ?
  • Does using a powered hub between the LCD and the PC change anything ?
Just need to keep moving things around until you find that the problem always moves with X or only happens when A and B are used together :-(
 
Top