Why aren't DAQ ports being controlled by labview?

[Daniggy]

I am a new user.  The engineer who developed the labview program for 2 of our automated testers is no longer employed here.  Both testers use 6501 USB DAQs.  Both testers are controlled by the same PC (not simultaneously).  The DAQs, in both testers, control a bank of relays whose coils are powered with a separate source.  Both testers have the same symptom of turning on all of the relays simultaneously when the USB cables are plugged into the PC.  This produces an unrealistic test condition on the DUT.

The relays can be turned on and off via the Test Panels tab in MAX.  We don't know what changed to produce this new symptom.  The testers were operational and stable for many months prior to this problem.

Any suggestions?

[John_P1]

The 6501 powers up with the DIO lines set to tri-state (not driven high or low).  It has a 4.7 kOhm pull-up resistor, so when the board first is plugged in you may be driving the relays through the 4.7 kOhm pull-up (depending on the resistance of the relay).

Do you have a specs page for the relays and how you are connecting them to the 6501?  I'm hesitant to offer a specific recommendation without this, but you might consider connecting the relays between DO.x and +5V rather than DO.x and GND.  Inverting your logic should give you the desired behavior.  Driving the line low would sink current and drop 5V across your relay.  Letting the line go to tri-state (same as outputting high in open collector mode) will pull the line to 5V through the 4.7kOhm pull-up and you would have 0V across your relay.

Best Regards,

[Daniggy]

The DAQ channels that are driving relays are each connected to the gate of a 2N7002 N-FET, a separate FET for each relay.  The drain of this FET is connected to the relay coil, the source is to GND.  The other end of the coil is to a 24VDC power supply.

I understand that the DAQ output may be in the "wrong" state before the labview program is running.  The thing that puzzles me is that the labview program should correct the relay state when it is running.  This isn't happening.  Either the relays are all on, or all off, and the test fails because of this.  I can send a relay drive schematic if the words above aren't doing it for you.

Being a new user I'm not in a position to be inverting the logic on the fly.  Again, this worked for many months prior to several weeks ago and if something changed we are not aware of it.

Thanks for your help.