[fordem]

The IAC and the TPS are two totally unrelated items - the only "commonality" is that on this vehicle, they both mount to the throttle body - and neither one is likely to cause a misfire under wet/damp conditions.

There are several different styles of IAC, this vehicle uses a stepper motor IAC - it rotates a screw that opens & closes a passage that allows air to bypass the closed throttle plate - if it fails, it's simply not going to turn and the ECU will be unable to adjust the idle speed - depending on the actual position at the time of failure that could mean the engine won't idle, has an unusually high idle or it may even run reasonably well when warmed up but won't idle properly when cold.

Most TPS are simple potentiometers, some have an idle position switch (this one doesn't), and this one is non-adjustable which implies that the actual throttle position is not critical - all it's doing is letting the ECU know how rapidly the throttle is being opened, fuelling is being done primarily based on airflow, with secondary information from the O2 sensor and MAP (manifold pressure) sensors - if fitted (my car has neither).

If I recall correctly, the TPS was replaced because of a flat spot when accelerating - the ECU couldn't tell how rapidly the throttle was being opened so there was no "additional enrichment" - ie. no "accelerator pump" functionality - the disappearance of the misfire is probably co-incidental and most likely relates to work that may have been done at the same time - disconnection/reconnection of a harness connection - removal/replacement of a screw which improved a ground connection - something along those lines.

I just re-read that thread - the MAF was also replaced after it died during cleaning - that could account for the change in fuel consumption, but I don't see it as the cause of the misfire. For what it's worth - I've cleaned many a MAF sensor, and you'll almost always find a MAF aerosol on my shelf.

[losfer]

Ok , so im the one who fixed the car and i will tell you what i found...

I was presented the car with this information.

engine rpm would dip on depressing the accel pedal then rise- verified on test drive
a miss/stumble on light cruise- verified on test drive
a miss when wet/moisture in the air- not verified

i was also informed that the air flow meter had been replaced to verify it wasnt the problem and
a new tps was fitted to verify it wasnt the problem.

So of course first of all i started by doing the usual thing . hook up scan tool and check codes , none were reported not even history codes.
from here i went around and did other electrical and mechanic checks . Check ecu voltages, power and grounds. Checked things like EGR (which turned out to do nothing because of his extractors/blanked EGR port) checked for vacuum leaks checked operation of pcv valve , went around with hose crimp tool to crimp on hoses to see if i could eliminate deadspot. checked injection pulses , NOTHING i seemed to do could get rid of it.

I knew the problem wasnt related to the coil packs/probes/leads/ spark plugs because although it had a miss on accel and miss/stumble on light cruise i could not replicate the problem under a heavy load scenario.

So this is the stage where i am clutching at straws and things start to get interesting.
put a timing light on and lock the timing in the scanner YEP 5 degrees right where it should be , unlock the timing in the scanner YEP 14 degrees right where it should be , rev up the car with timing light on and WTF. The timing before advancing would retard to 0 degrees!!!! (right where the flat spot was)
Check the live data in the scanner and sure enough timing was dropping from 14 to 0 before advancing , every other data in live data was acting exactly how it should (even tps)

Ill just point out that prior to noticing this problem the only standout thing in the scanner live data was the tps position would sometimes read 1-2 % on full closed throttle so i elongated the holes for the tps mounting and adjusted it so it would read 0% at closed throttle at all times. I know this tps doesnt have an idle switch but was at this stage clutching at straws and just wondering if that was causing some sort of weird fuelling issue causing the jerking on light cruise. i know if the GTi tps are not set properly with the idle switch they will cause a similar jerking.

so back to the retarded timing.

I started checking the things that would or could possible affect timing , especially RETARDING it. first off i started with the knock sensor , plugged another in with no change. hooked up an oscilloscope to the ecu to check cam position sensor signal to check waveform. So the square wave signal the cam sensor puts out should oscilate between 5v and 0v , i was getting .5 of a volt and 0 volt. Confirmed that i had a 5 volt reference signal from ecu and a good sensor ground. i tried another cam sensor i had here to no avail.

So at this stage its getting late in the afternoon i ring Max and inform him that i think he could possible have a faulty ecu. After a short discussion i agree that i would recheck all the wiring for the cam position sensor in the morning.
Now this is where it gets a bit fuzzy for me because the next morning when i checked continuity of the wiring from sensor to ecu and then rechecked the cam signal with engine running i had a perfect signal 0-5v , so whether i misreaad the ecu pinout , accidentally probed the wrong pin for ground , or the wiring diagram was simply wrong i dont know , and didnt look any further into it since i now had a good signal.

This is the stage where you would usually hand the car back and tell the customer you cant fix it , but i dont because 1. i usually wont charge them 2. i hate when something beats me 3. you spend the next week awake half the night wondering what it could possibly been.

Really really clutching at straws now so i just randomly go about unplugging sensors to see if things change , coolant temp , maf (even re -cleaned)and then tps and bang stumble gone with tps unplugged , scratching head now , previously confirmed with Max that nothing changed when he changed original tps with brand new tps , so i hook up the oscilloscope to tps. As fordem mentioned tps is a pontentiometer , it gets a reference voltage of 5v and the voltage rises from the base voltage and gets greater the further the throttle opens. so the base voltage was .75Mv as the throttle was ever so slightly opened the voltage dipped to .68Mv before rising. YES .07 Mv change was enough to cause the timing to retard 14 degrees !!! The live data on the scanner couldnt even pick this up !!

Went and got another tps and plugged in and checked voltages/test drove car FIXED.

Just to confirm to Max i went and got a $60 used tps from wreckers , last time i priced a new one from suzuki they were about $400.

Moral of the story dont even trust new parts (especially cheap aftermarket ebay ones)

[Gwagensteve]

Thanks for the update losfer, very interesting.

Steve.