Instrument disconnect

[Potticary]

It's fairly inconsistent, I've been waiting for it to disconnect again which this time took 47 hours. The 4 machines are all connected through USB (virtual com ports). The two on the opposite side of the lab are connected through a 12 metre USB repeater that plugs into the mother board while the other two (one of which is the issue) have their own ports on the mother board.

Something interesting did happen earlier this week that may be significant: When restarting the VI and magnet, I actually forgot to turn the problem magnet back on. After a couple of days both magnets that are connected through the repeater had disconnected, which I had never seen before. The fourth was still running.

I'm about to start them all again with some of the suggestions (the ones I know how to do) by minions added. I've put some feelers out about the GPIB adapter to the physics department. If not, I'll buy a USB card and see what happens.

[cstorey]

Inconsistent problems are the worst to debug - I usually blame Windows for those. 😉  

I'd try running a "torture test" program with just the 2 magnets on the repeater, where you continuously read from both magnets every 1s and see if you can repeat the errors more often. 

Now that you mention the really long cable lengths, that might be an important issue.  12m (40feet) of RS232/USB cable might be causing you some issues if you use high baud rates.  There's info on cable length max and capacitance near the end of this article - https://www.lammertbies.nl/comm/info/RS-232_specs.html

Other tips would be to use lower baud rate (2400?), and look for higher quality (lower capacitance) cables for the longer runs.

If you were to go to GPIB, you might well exceed the cable length specs of a max 20m depending on how you run the cables.  See GPIB cable length specs - https://knowledge.ni.com/KnowledgeArticleDetails?id=kA00Z000000P8slSAC

At the very least you probably need an isoloator/expander for that long run - http://sine.ni.com/nips/cds/view/p/lang/en/nid/203221 but you should talk to NI about solutions for that.

Good luck.  Post code if you need help.

[Potticary]

Ok, still no joy with the neatened code. I am noticing however that the disconnects appear to be happening during the day, likely when the computer is trying to something else. I'm wondering if it's bugging out in windows in some way? I've ordered the PCI USB board and have managed to get a GPIB - USB as a loaner for a few weeks. Hopeully that will help?

Regarding the cables, ironically the 'bugged' power supply is the closest to the computer connected via a standard 2 m cable. I'l lpost updates when I have some.

Thanks all for the help thus far.

[Minions]

---

Regarding the cables, ironically the 'bugged' power supply is the closest to the computer connected via a standard 2 m cable. I'l lpost updates when I have some.

---

Ok, for your GPIB connections, are you using a daisy chain or a star configuration from that 'bugged' supply?  I would recommend using the star pattern where possible.

[Potticary · Accepted solution]

Just in case it helps anyone in the future, I changed the computer to a not-as-old one and it stopped happening. I assume the university did something to encumber the computer in a way I couldn't fix.

[ikaiser]

Hey Potticary, thank you so much for reporting back! This really helps. Please go forward and mark your own answer as Accepted Solution.