Just a follow up. I replaced the spider assembly with another one I had and the problem seems to have been resolved.

 

Follow up to the follow up, problem is still there. I have spare ECT and AIT sensors, so will swap them out and see what happens.

 

What's to say the replacement ones are good? Best to put a scan tool on it and check the temperatures being reported.

 

I actually did and there was about an 8 degree c difference between them. In the garage, the ait read 12c and the ect read 20c. I did briefly start it a little while before, so some of the differential may be due to that. The replacement ect sensor is new, but the ait was an old one. I read the temp after I put it in and it showed 18c which considering the car was recently run seemed ok. The ect read 69c.


We'll see how it goes tomorrow.


Tim

 

Cool. Before you start it in the morning, connect the scan tool and compare ECT, IAT and ambient temperature with the ignition in the RUN position, engine off. All 3 should be within ~2C of each other.

 

New ECT sensor and starts first time every time. Old one never did throw a code. Why the heck didn't I replace it first!


Tim

 

That explains the lean symptoms when cold. Sometimes you can't go by the book, you have to think outside the box. If the ECT or IAT sensors report warmer than actual temperature, (especially the ECT), the fuel mixture commanded by the PCM will be lean on cold start up, (open loop). When the PCM is operating in open loop, it ignores oxygen sensor data and uses data from ECT, IAT, MAP, MAF & TPS to determine what fuel mixture will be commanded. Once in closed loop, the PCM relies almost entirely on oxygen sensor data for fuel mixture.

 

Another follow up. -26c last night and the problem is still there. On fuel prime, engine will just catch then immediately die. This happens much quicker than a Passlock problem. On continued cranking, the engine will occasionally and barely, try to catch. Seems to be too lean to start.


I'm beginning to think the car has a demon. Still no codes, new ECT, replaced AIT, different spider, new regulator, new fuel pump. In warmer weather, it runs perfectly, smooth idle.


Is there a flood clear mode? I.e, wide open throttle and cranking? No codes, but could the TP sensor be showing it is open? I have not read the TP voltage, but will a little later. Is the VTD mode the same as the Passlock shut-off?

 

There is a flood clear mode, unloader, dechoke, whatever you want to call it: Wide open throttle, then crank. It shuts off the injectors. Best way to see what's going on with the TPS is connect the scan tool and watch it on start up, should stay low and steady. With the ignition on and the engine off, watch TPS data while slowly pushing the throttle to the floor and release it. Should be slow, smooth increase and decrease, no spikes, no drop outs. Cranking without touching the throttle, TPS voltage should be close to zero volts. Wide open throttle it should be just under 5 volts.

99 times out of 100, if there are no DTC's, there are no problems with anything that is monitored by the PCM. Fuel delivery is not monitored by the PCM, and therefore can not set a DTC or turn the SES light on if there's a problem. I have seen pumps that will run on initial prime, but the pressure gauge needle didn't budge for a second or so. On your 99, you can check this very easily. In the underhood fuse panel, next to the fuel pump relay, there is a vacant terminal referred to as the "pump prime terminal". The circuit runs directly to the fuel pump motor.


Connect the fuel pressure tester to the service port. Connect one end of a fused jumper wire to battery positive. While watching the fuel pressure tester, touch the other end of the jumper to the pump prime terminal in the fuse panel. The needle should jump up immediately. The ones I encountered, acted up maybe one in twenty times.

Another possibility is a funky spot in the ignition switch, (also not monitored by the PCM). We can check that too if the fuel pump checks out OK.

 

I was chasing the wrong thing. Problem turned out to be a bad BCM. Only happened when cold, worked perfectly when warm. No other obvious symptoms. Only suspected it when I tried to do a re-learn and the BCM would not enter re-learn mode until it warmed up.


I feel pretty stupid for chasing a lean start when it was this!