Compressor Relay doesn't fire???

[Invictus3593]

After an accidental +12v-to-frame short, our Compressor Relay suddenly will not fire. We put the robot into Test mode and tried manually starting the relay, but we don't even get the green relay light on the sidecar (keep in mind none of the robot code changed). We have changed the sidecar (confirmed fried), crio (smells fried), 9403 card, relay, anything that is inbetween the computer and the sidecar that could interrupt the code from reaching the SideCar but still no relay. The pressure switch isn't stopping it, that works fine.

The odd thing about the whole situation is that all of our motors, solenoids and everything EXCEPT that relay works. We don't have any other relays, we tried different relay channels and even bypassing the router and going straight tether-to-crio. Still nothing.

Any ideas on what our problem could be?

[Mark McLeod]

Did you try it without the Spike connected (just looking for the green Relay light)?
In case the Spike PWM cable is shorting the Relay.

I assume all DSC power lights are green?

Test the seating of your 37-pin cable, since the relay commands come in on separate lines in that.

[Invictus3593]

We didn't try it without the PWM cable, but we did replace the relay. We'll try that.

Yes all the power lights are green.

We have seated and re-seated the DB37 cable a couple times and even switched out cables, still nothing:/

[Al Skierkiewicz]

Invictus,
Before we discuss the original problem, the frame short concerns me. Often a single short doesn't produce the results you are describing. It sounds like you could have another frame short on the robot. Considering the amount of damage you have experienced, there is a good chance the Spike took a hit as well. The LED Mark is asking about is the Relay Tally lights on the DSC. When the DSC is issuing a command to a relay, there is an indicator on the DSC that lights for the relay(s). If one of these is not coming on when you think it should, the relay will never respond.

[Invictus3593]

Invictus,
Before we discuss the original problem, the frame short concerns me. Often a single short doesn't produce the results you are describing. It sounds like you could have another frame short on the robot. Considering the amount of damage you have experienced, there is a good chance the Spike took a hit as well.

The short circuit went from the +12v-frame to the GND in a PWM wire (there were sparks) so it totally fried the DSC we had on there (we can see and smell a burnt-out IC). We did check the GND-to-frame and the +12v-to-frame after we found the short and fixed it, and there's no continuity (this made us cheer). Is there another short that won't present itself in those tests? We also thought the spike probably took damage so we switched out the possibly-burnt-out relay with a working one.

The LED Mark is asking about is the Relay Tally lights on the DSC. When the DSC is issuing a command to a relay, there is an indicator on the DSC that lights for the relay(s). If one of these is not coming on when you think it should, the relay will never respond.

Right, this is my biggest concern. I mentioned this up in the original post:
...tried manually starting the relay, but we don't even get the green relay light on the sidecar

And after changing practically everything in between the DSC and the control station including the DSC itself, it still doesn't give that light, but Test mode says it should be running and reads the pressure switch just fine. :confused:

[Al Skierkiewicz]

If there is a short downstream from a controller, it won't present itself during a resistance check from either of the power input terminals on the PD to frame. This can occur with a wiring short to frame or with a defective motor. the only check for those issues are a continuity check from the output of each controller to frame. In this case though, the fault doesn't actually cause an issue until the motor is commanded except for Jaguars.

[Invictus3593]

We just checked both + and GND out from the relay OUT and no beeping from our fluke meter. Should we check all of the controllers? Even if they work fine?

EDIT: We checked all of them, nothing.

[Al Skierkiewicz]

I am of the "smoking gun" philosophy. If you don't find the problem, it is still out there and is waiting for a chance to occur again. If you know the cause then you can fix it. Have someone else look over the robot, what is obvious to them you might not be seeing. I have looked at hundreds of robots and pulled on a wire that was laying against the chassis that the team had been ignoring for weeks.

[Invictus3593]

Figured it out! Turns out the pressure switch was giving reversed signals to the computer even though we hadn't changed the wiring or code...:confused:

waste of about 6 hours

http://kinialohaguy.files.wordpress.com/2010/06/banghead.gif