Idle momentarily jumps to 2300 rpm just after starting up

[sweetp01569]

I've seen some threads regarding jumps in idle rpms at startup. My idle speed motor seems to buzz and hum properly - I cleaned and lubricated that, vacuum likes appear intact and connected, the ECU module had power going to it, connector is intact with clean contacts and wiring, the idle microswitch is engaged. What happens for the last several startups when cold (ambient temperature about 70 to 75 degrees F), is that the car will start up normally, with idle at the proper 775-ish mark. Then, after about 30 seconds or a minute, the rpm will jump up to 2300 rpms, perhaps for another 20 seconds or so, then drop back down to normal, then it might pop back up to about 2300 rpm again for another 20 seconds, perhaps, then settle out at normal 775, then stay that way, as the car continues to warm up, and never an issue while driving. It just does this intermittent jump up to 2300 rpms once or twice just after start-up and then normal all other times. Any guidance on what I could check for this, or if anyone else has had an issue?

Thanks!

[Bitsyncmaster]

My guess would be an intermittent connection of the temperature sensor. You can test by shorting the two sensor wires at the idle ECU.

[sweetp01569]

My guess would be an intermittent connection of the temperature sensor. You can test by shorting the two sensor wires at the idle ECU.

Thanks. I'll locate those pins (I believe I measured the resistance yesterday between those). Are you saying to short them on the backside of the connector while connected to the ECU (or leave it unplugged while shorting it) and start the car from cold to see if that prevents the intermittent surge? If there is an intermittent connection, I hear that is hard to get to, correct?

[sweetp01569]

Thanks. I'll locate those pins (I believe I measured the resistance yesterday between those). Are you saying to short them on the backside of the connector while connected to the ECU (or leave it unplugged while shorting it) and start the car from cold to see if that prevents the intermittent surge? If there is an intermittent connection, I hear that is hard to get to, correct?

Stupid me - lol! I need to short the pins in the ECU, not in the connector! That would do nothing. Not thinking it through Thanks again. I will try that.

[Bitsyncmaster]

Stupid me - lol! I need to short the pins in the ECU, not in the connector! That would do nothing. Not thinking it through Thanks again. I will try that.

Yes the short has to be going to the ECU but you can leave the sensor connected and still short the wires without harming the sensor.

[sweetp01569]

Thanks, as always, for the great advice, Dave!

I spent some time this weekend to isolate the issue of a possible intermittent connection at the or near the temperature sensor (thermistor?) Using an endoscope, I was able to view the connector at the sensor. The sensor is new (installed just 2 years ago), and visually, the connector and wiring appeared fully engaged. While keeping an ohmmeter connected to measure the resistance (while cold), I reached my hand as far as I could to the rear of the engine, feeling for the wire leading to the sensor. Sure enough, as I wiggled the wiring, I could see the connection "open" and "close" on the meter! I struggled to reach further in, and expected I was only a few inches from the sensor itself. I wiggled the wiring a little more, even giving it a little "tug" to jostle the connection. Fortunately, I was able to stabilize the connection with these actions, enough so that continued wiggling of the wiring did not disrupt the connection and it is holding tight. I hope it remains that way, to prevent disassembly to provide a proper cleaning of the connector.

Curious, Dave, did you state somewhere that you could actually reach the sensor yourself? I wonder if I propped myself higher on the engine block, if I could have pushed my hand further into the valley to wiggle the connector itself to seat it better? Or can it be reached better from underneath?

Anyway, thanks again, Dave. You initially suspected it was this connection that was failing and causing the rpm to jumps to 2000. It was definitely the cause!

Paul

[Bitsyncmaster]

I have never tried to reach that sensor without removing the intake. Maybe if your up high you can do it.