Ontopic A Thread About Butt Mustard, For Those Who Drive Automobiles

Status
Not open for further replies.
Got the car back, and it's been touch and go.

We tuned it, it made 390HP, but it felt off the whole time. It wouldn't idle without misfiring, and it wouldn't idle below 1200RPM. I'd basically stuck it in the garage and let it collect dust. I went out today after I finished my training for work, and took a look at the base timing, and saw that it was way, way off (like 30° off). Took it back to the tuner, and we got it set back up correctly, and now it feels a billion times better.

Now time to turn the boost up.
 
  • Gravy
Reactions: HipHugHer
ECU in its home
IMG_20170831_163226.jpg


ECU, about to be mounted to the board (just with zipties for now).
IMG_20170831_163741.jpg


I ran the programming cable (a USB cable) up into the glove box. I'm also getting an AEM multifunction gauge that'll connect to the serial port on the box.
 
@gee, I may be overthinking this, but it's bugging me.

For whatever reason, the auto climate control is flaking on keeping the AC request line (for high idle and compressor clutch control) to the ECU grounded out when the AC is supposed to be on. The net result is that the compressor isn't held on for any appreciable amount of time, and it doesn't work really great. I checked all the shit that I can, and everything lines up right.

I'm now pretty certain that it's actually the auto climate control computer, as I swapped the request pins on the ECU from one switch input to another, and it still flakes out and short-cycles the compressor.

Is there a relay/timer setup that would hold that pin to ground for the 3 seconds or so that the computer flakes out before it cycles back on?
 
Last edited:
Just so I have this right:

When the AC is switched on, it's supposed to pull a line low that goes to the ECU, which tells the ECU "AC is on"... and the ECU in turn drives the AC clutch?
 
My first thought is.. is the climate control computer sensing something, thinking "hmm, the AC shouldn't be on for <reason>" and then turning the AC off. Is it sensing AC high side pressure and killing the AC 'cause it isn't high enough, or sensing the cabin temperature and thinking "it's cold enough", killing it?
 
My first thought is.. is the climate control computer sensing something, thinking "hmm, the AC shouldn't be on for <reason>" and then turning the AC off. Is it sensing AC high side pressure and killing the AC 'cause it isn't high enough, or sensing the cabin temperature and thinking "it's cold enough", killing it?
The climate control computer doesn't do that job. The window switch for the clutch to safeguard the compressor is on the other side, it kills power to the clutch, not to the AC request line on the ECU.

The cabin temp and vent temp sensors all read correctly.

If I put the old ECU back in, everything works fine.
 
The climate control computer doesn't do that job. The window switch for the clutch to safeguard the compressor is on the other side, it kills power to the clutch, not to the AC request line on the ECU.

The cabin temp and vent temp sensors all read correctly.

If I put the old ECU back in, everything works fine.
Sounds like the new ECU is the issue. Any way to reprogram its logic to correctly handle the AC signal from the HVAC computer, or is it limited to programming spark/fuel/etc tables?
 
Sounds like the new ECU is the issue. Any way to reprogram its logic to correctly handle the AC signal from the HVAC computer, or is it limited to programming spark/fuel/etc tables?
I have an iron in that fire, but I figured I would run down any alternatives in the mean time.

Right now, I have a switch in the glovebox that pulls the ac request line to ground manually.

I did confirm that the ECU didn't have a flaky input, by swapping the request line to a different input and seeing that it behaved the same way.

I'm down to blind troubleshooting, which isn't somewhere I enjoy being. I think the only thing that is left is some undocumented interface to the ECU that I won't be able to duplicate.
 
I have an iron in that fire, but I figured I would run down any alternatives in the mean time.

Right now, I have a switch in the glovebox that pulls the ac request line to ground manually.

I did confirm that the ECU didn't have a flaky input, by swapping the request line to a different input and seeing that it behaved the same way.

I'm down to blind troubleshooting, which isn't somewhere I enjoy being. I think the only thing that is left is some undocumented interface to the ECU that I won't be able to duplicate.
More thoughts:

Got a DMM? Measure the voltage on the HVAC -> ECU line when the HVAC controller hauls it low. It's possible the HVAC controller only hauls it down to 1.x volts because it's got a darlington output or something, and the ECU still treats it as a high signal.
 
Status
Not open for further replies.