Slightly annoying idle speed 1000rpm
 

When I clutch in at speed the rpms drop slowly to about 1100-1200 then after stopping they normally drop to 1000rpm rather than 800rpm, so i would assume there must be a signal change recieved from the speedo to the ecu. I measured 13vdc at idle on ecu pin 34, I dont know what it should read at idle. I have on two occasions started a warm engine and it initially has idled at 800rpm, once driven the idle increased to 1000rpm. Also interestingly my afr at idle with a warm engine drops to low 13s at times. Mid 14s is what I see most of the time at idle.


I've grounded pins 14, 15 and 30 at the ecu. I'm fairly sure the car is increasing idle speed intentionally. I've set the base idle to 500rpm, tested the throttle switch resistance and it works perfectly, set throttle switch position ok, swapped to a spare iacv, swapped a spare maf, sprayed easy start all around the mani for air leaks (although I did notice that when I pull the vac hose from the fpr and introduce an air leak there is no noticable change in rpm) I assume this is the ecu adjusting the iacv to compensate.

I hadn't mentioned already but the car is not standard. Here are the mods that could effect idle but I dont believe they are. AC removed, Vxr injectors, chipped ecu ezk, breather is VTA, V cam, ttv flywheel.

Any more insight as to where to look next would be appreciated.

Also have fitted a new bosch coolant temp sensor.

Have you read the codes?

My initial thought is temperature sensor, it thinks its a bit cold. Note there are 2 sensors, one for the gauge and one for the fuel ECU.

If you are seeing a decrease when stopping then sounds like the speed sensing is working. Normally this comes from the rear ABS/diff sensor and is a pulsed signal so DC measurement will not indicate much. You need as scope, but the speedo is a perfecly good detector for most issues.

If you use full power without speed pulses appearing the fuel ECU will create a limp mode and raise the idle to about 1800rpm.

I have seen linkage adjustment and dirty throttle body causing problems, so make sure its clean and maybe pop off the linkage to eliminate adjustment, but sounds like you have ensured it is properly adjusted.

Assuming your AFR is to be believed, it's running a bit rich at idle. What's the current mileage and when was the Lambda sensor last changed, if at all?

I will need to buy a code reader, i think sombody on the forum sells them.

TB is clean as a whistle. There is ample slack in the linkage that its not possible to get hung up off the throttle stop. The temp sensor is new but i could try the old one to see if its different. The AFRs from cold start and driving are ok. Its only when the engines hot and im idling for a bit that it becomes richer.

I swapped a spare lambda sensor in earlier and went for a drive, idle speed is the same though. Need to monitor AFRs.

I need to replace my steering rack and pump as the rack leaks and the pump complains, still perfectly drivable but they are not quite right. Is it possible the ps pump could be loading the engine and causing the ecu to compensate? Or am I clutching at straws.

Where did the spare Lambda sensor come from? :thinking:

It's possible the PAS system is loading the engine unnecessarily, especially if your rack isn't properly centred in the straight ahead position before wheel alignment took place. That will cause the pump to boil the fluid and eject it from the system, out of the filler cap breather hole if you have an integral pump/reservoir like mine.
Also boiling the fluid will cause air pockets and hence cavitation within the pump, not good for the pump, rack or anything else.

"Where did the spare Lambda sensor come from? "

The middle drawer in my garage. 😝. I know they can be sensitive but its unlikely to have an identical fault? Where do you test the lambda voltage? Am I right in thinking if it appears to swing back and forth from 0-5vdc then its functioning ok?

You have a drawer that grows Lambda sensors? :err: Where did it come from before that, was it new or used?

If it was used, buy a new one, subject to test. You test on the output wire of the sensor, +ve probe of the multimeter on it and -ve probe to a good earth.

The voltage will be in the range of 0-1V, on a hot idle it will swing from an ideal 0.45V to 0.001V then up to near 0.9V then swing back down again continuously. If your AFR is correct then in theory should be reading more in the range of 0.1-0.999V but i suspect it will be lower and the sensor is old and lazy.

If it is old and lazy, it's probably likely to swing from 0.001V to ~0.8V, if it's steady then there is a fault in the rest of the fuel injection system or the sensor is dead/near-dead.

If both sensors were used there is every chance they will give the same fault.

I can't see the O2 sensor affecting the idle, you need more air for that especially if its rich. Adding more fuel will not increase the rpms on its own. There is a slow rpm compensation at idle, were you can let the clutch out very slowly without using the accelerator. I would have thought it would work both ways so it can close the IACV down if idle is too high.

I mentioned there are 2 temp sensors, did you definately change the ECU sensor? This is the only possibility I can think of where it is doing it deliberately, everything else you mentioned being ok.

Another possibility is an air leak post throttle body, inlet gasket is easy to mess up if you don't tighten up the bolts evenly. Then there is injector seals and brake booster pipe and idle air valve pipes. Some spray liquid like WD40 is good for testing and obviously inspect all the rubber.

What Engine/model/year?

A feature of the LH system is if it detects (or even suspects) rich running, it will raise the idle speed, it even does it on my LH-2.2 without a Lambda sensor if it's had a tricky start.

With the LH-2.4, it will always aim for the lean side of stoichiometric so again, if the Lambda has gone lazy and raising the idle speed for a short period doesn't reduce the reading, it is likely to keep the idle speed raised while reducing the injection time. If it can't get the results it's looking for it will keep the idle artificially high.

As for the air leaks etc, that's entirely possible and/or probable. Also if it doesn't detect as much of an increase from the Lambda as it expects from a longer T-inj then it also keeps the idle speed higher, not sure how/what it bases that on but it's something i've deduced from the multiple threads we've seen about similar.