Stuttering compressor

[sp33chy]

Hello. I'm wondering if it is normal that when the compressor gets to 115-120psi it kind of does a small stutter instead of turning off completely (the pressure does not continue going up however).

The pressure relay switch seems to be on correctly.

Thanks.

[hyperdude]

I'm no robot programmer, but I am pretty sure that you need to program the compressor to shut off when the pressure gets up to 120psi (when the relief valve opens). There's no way for the pneumatics system to "know" when it's at a certain pressure.

Now if you've already got this programmed, then you might want to post the code so someone can help you.

[sp33chy]

I am coding in Java. Other than when instantiating the compressor, there is seemingly nothing else to do with the pressure relay switch ... or according to the API

[hyperdude]

I think that the pressure release valve (with the manual valve on the end) can be wired up so that once it reaches 120 and starts releasing pressure, it sends a signal to stop the compressor.
Again, I am not sure about this, but it's what I remember from what's been going on with our team the last two years.

[Mark McLeod]

I'm wondering if it is normal that when the compressor gets to 115-120psi it kind of does a small stutter instead of turning off completely (the pressure does not continue going up however).

No, that's not normal, unless you have a fast air leak.

Normally, the pressure switch goes on when the pressure drops below 95psi and turns off when it gets above 115psi. Then doesn't turn on again until it again falls below 95psi.

What does the gauge show?

[JBotAlan]

If the compressor is staying on after 120psi, you will not pass inspection; this is not normal behavior.

Measure the resistance across the pressure switch. Does it toggle when the pressure reaches 120psi? (I can't remember if the switch is normally closed or open offhand) If not, you'll need to replace the switch.

As far as code goes, you should only need to instantiate the Compressor, I believe. It deals with itself after instantiation.

[MattD]

As far as code goes, you should only need to instantiate the Compressor, I believe. It deals with itself after instantiation.

The compressor will not operate until you first call the Start() method on it. Instantiation alone does not enable it.

[DonRotolo]

I think that the pressure release valve (with the manual valve on the end) can be wired up so that once it reaches 120 and starts releasing pressure, it sends a signal to stop the compressor.
Again, I am not sure about this, but it's what I remember from what's been going on with our team the last two years.No, you must use the pressure switch from the KoP to provide a signal to the cRio (via Digital Sidecar). That is the only electrical component that has the capability of signaling the system pressure for control.

[hyperdude]

No, you must use the pressure switch from the KoP to provide a signal to the cRio (via Digital Sidecar). That is the only electrical component that has the capability of signaling the system pressure for control.

Yup, that's it...

[Isaac501]

I'm no robot programmer, but I am pretty sure that you need to program the compressor to shut off when the pressure gets up to 120psi (when the relief valve opens). There's no way for the pneumatics system to "know" when it's at a certain pressure.

Now if you've already got this programmed, then you might want to post the code so someone can help you.

This is incorrect. There is a pressure switch that (if properly calibrated) will signal when the pressure is at 120 PSI. wpilib has code which turns off the compressor when this signal is received. The compressor requires no intervention past properly wiring it and instantiating and start()ing.

[hyperdude]

Yes, I know. I got confused over this, because I have never had to set up the pneumatics or programming.

[sp33chy]

Hm. I am at a standstill then regarding it. I can manually stop and start the compressor with my code. Here is how I am instantiating it:

Compressor compress = new Compressor(1,1);

Here is my wiring: http://i49.tinypic.com/ixvxxl.jpg
http://i48.tinypic.com/23wltx.jpg

Anyone see anything wrong? Please help:confused:

[JBotAlan]

Your wiring looks fine.

Have you measured the resistance across that pressure switch like I mentioned before? It is possible that your switch is defective. That switch should open (resistance jumping to infinite) when the air pressure reaches 115psi, and it should close once again (resistance jumping to 0) when the pressure drops below 95psi.

[Al Skierkiewicz]

Anyone see anything wrong? Please help:confused:

First thing, the red wire in the PWM cable connected to your switch needs to be insulated.

A little stutter is normal as the motor winds down against pressure. if it takes place for several seconds, it could be a noisey contact (rapidly turning on and off) within the switch.

Mark, might this also be a result of the same software issue with stuttering in other threads?

[sp33chy]

The stuttering stopped by itself seemingly. I fed the watchdog a few more times in the code but I doubt that has anything to do with it, although it could and is worth mentioning. Maybe it was just because it was a new compressor?

Also thanks for catching the part about the uninsulated wire.

[Al Skierkiewicz]

You're welcome.

[Jim E]

We had the same issue. I think the watchdog has more to do than you think.

Software control requires refreshing every so often, or it reverts to a shutdown state.

[sp33chy]

Interesting, the watchdog wasn't complaining on the dashboard however. Maybe I just fed the watchdog a step too late which wasn't enough to show up on the dashboard but was enough to make a difference.

So to any teams having this issue: try feeding your watchdogs more often.

[SteamChief]

We had the same issue the other day. For some reason, the problem seems to have corrected itself. No changes in programming were made. While we were having the problem, we changed batteries, thinking it may be a voltage/dying battery issue. A freshly charged battery made no difference.