ISC open too much during idle

[mjdarg]

Hey guys,

I have been having trouble getting a decent idle recently. I have been running a few months with the ISC unplugged while I try to figure out the problem. Stock 91 VR-4, using TMO to datalog. The ISC stays around 76 or 77 steps during normal idle, while the engine is around 1500 rpm. If I ground the connector on the firewall, the ISC steps drop down to 9 or 10, and the idle drops down to 800 where I set it with the BISS. As soon as I un-ground the connector, I can see the ISC steps increase in TMO first, and then the rpm and advance increase as an effect. I have also noticed that if the ISC is unplugged, grounded the connector makes no discernible difference to how the engine runs.

Is there something simple that I'm missing?

The O2 sensor cycles, but when it is idling around 800, the frequency seems to be about 0.5 Hz (once every 2 seconds). Around 1500, it cycles much faster, around 1.5-2 Hz. Does this seem normal or would it affect my problem in any way?

One thing I've wondered...how does the ECU determine the required number of ISC steps to maintain correct idle? Does it simply compare the current engine rpm and compare it to the setpoint? Maybe there is something wrong with my ECU... /ubbthreads/images/graemlins/dunno.gif About six months ago, my idle would increase while turning, which was the ISC correcting for the power steering, but I guess too much. I sent my ECU to ECMLink and they couldn't find any issues.

Lots of rambling...what do you guys think?

 

[tektic]

More likely something wrong with the isc. The coils are known fail points. Try to measure the resistance between each. If one are two are not in spec there is your problem, however I and others have had them check out ok and still not function. The best way to test is with a 0-5v power supply so you can watch each coil move.

I also notice if I have my looger pluged in and ground out the timing plug it doesn't work. I have to unplug the logger.

 

[prove_it]

Check for vacuum leaks, the throttle body shaft seals are horrible and a leak will mess things up.

If your ISC is white/yellowish in color, get a new black one asap. The white ones are old and will fry your ECU if they go bad, and chances are it will happen.

 

[mjdarg]

Thanks for the replies.

I've got a new ISC from Advance (carquest brand). I disconnected the battery last night to try and reset anything the computer might have "learned" while the ISC was disconnected. The car will now idle around 8-900 rpm with the ISC steps around 37-38. The BISS is almost all the way in. I am now back to my problem I had about a year ago when I sent my ECU to ECMLink. I was under the impression that the ISC would correct idle for any electrical load placed on the motor. However, I can only get the ISC to correct idle when I turn the steering wheel; power steering. If I turn on things like the rear defroster or headlights, the idle simply drops a bit and the ISC steps remain the same on the logger.

So, I'm wondering if it's an ECU problem...but when I sent it to ECM, they said it checked out ok. How does the ECU know to increase ISC steps when specific loads are placed on it? Are there sensors at each system (power steering, defroster,AC,etc.)? It's interesting because I can see that the ECU isn't even trying to regulate idle when I turn on an accessory.

 

[iceman69510]

The ps has a feedback device. The alternator does not on these vehicles. Only the cam sensor determining the engine rpm and the ecu desiring to keep it in the set range would be the determinants.

 

[mjdarg]

Ah, that's good to know! So, it is normal for engine speed to slightly drop while running things like headlights and defroster?

 

[prove_it]

Yes, very. The ECM does not have a electrical load sensor of any sort. Up to 200 rpm drop is normal. It should still recover after a bit once the ecm sees less rpms.

Remember, our computers are not super fast at processing and there are delays.