NI3000 I/Os jammed on

Almost sounds as if the IOs are configured as logic high. Setting the input as high logic disables the ability to use the channel as an output (so you won't be able to change status by turning the IO on or off) and the status when open circuit is 'on'. See the documentation on any integrated controller for the send commands for 'GET INPUT' and 'SET INPUT'.

Also, does the status of any input go from 'on' to 'off' if you short the IO to ground (like a contact closure)?

rynando wrote: »

I saw this happen once to an NI-700. The installers said they inadvertently connected the IO ports to a high-voltage output of some sort. All the IO indicators were forever stuck on after that.

Best I've been able to tell, they are direct collector outputs on TTLs. It doesn't take a lot to blow them up. I've done similar with individual IOs, so I wouldn't be shocked it someone accidentally blew them out. However, check the voltage too. If the board voltage is off, it would cause the TTLs to read wrong as well. I've blown up that onboard supply on a controller as well, and it sometimes causes unexpected results.