Notice:
If your engine/vehicle is or was ever affected by the engine compartment fire hazard recall this tip is for you!
Background
I'm a 55 yr old accomplished EET who's worked for (Veeder)-KBR, Philips, IBM, Atari, Williams Electronics and a number of high end Hi Fidelity and AV stores, and as a pinball video distributer Service Mgr. and ace troubleshooter, for many years, now a semi-retired web programmer and webmaster/designer.
I studied this seemingly random stalling problem for over a year with my 2000 GTP 4DR 3.8L V6 Supercharged (L67) with 140 KM on it, it worked fine after I bought it 2 yrs ago all fall, winter and spring but started acting up on me last summer usually always on a hot day. The problem just progressively gets worse and worse in any sort of slow traffic and/or hot weather. It now has 196K on it, gets 24-26 MPG, burns no oil, (but still leaks around the heads and valve covers) and goes like madd. Otherwise I really love this car.
The former owner, a truck driver friend of mine (who also had a 2001 2DR for his daughter) told me he had only ever had this problem in the summer a few times and (on this one) the dealer had replaced two of the three HV modules, the MAF sensor, the Fuel Pressure regulator switch, and the Throttle Position potentiometer, but the problem was finally solved (he said) when they put a new CPS sensor and ICM (the driver board under the ignition modules).
It has also had the resistor mod for the fuel pump done. The idiot dealer tech put it over the hot back manifold on the rear firewall I moved it over to the passenger side fender wall which is a much cooler location.
This car also has a new catalytic, new EGR valve, new O2 sensor, and i cleaned out the IAC and tested the purge solenoid, but the odd intermittent behavior continued getting worse and worse.
Due to the incompetence and idiocy of a RapidLube monkey my fuel filter went unreplaced while the moron had twisted the fuel line like a pretzel obviously by not having gripped it properly, they said I should get a new fuel line, because it was "rusted"! This damage to the fuel line by them resulted in the failure of my fuel pump which I (my lawyer) forced them to change and repair the fuel line they had damaged.
I got my recall notice read the report and pulled off the flammable plastic ignition wire holder thingy scraped off all the leaked oil tar crud from the front valve cover dripping out over the front exhaust headers, and rerouted the ignition wires out over closer to the rad, not having time to risk some dealer screwing up something else, I'm still waiting to get my (only one?) new valve cover gasket replaced, just can't be without the car! (My lawyer is interested too)
Symptoms:
Engine runs perfect, great mileage, no engine lights save for odd loose gas cap occurrence. but often it just stalls in city driving and won't restart for a while or won't start after a short trip and park on a hot day.
Problem(s):
Stalls suddenly on hot day usually at a traffic light (brake idle) or immediately after a long traffic light, occasionally will also intermittently stall out in very heavy bumper to bumper traffic, usually under 25 MPH, or after exiting from a highway at/to a light.
Engine may also stall out if the car is parked for 15-20 minutes restarted and run a short time, or may not start 'till engine has totally cooled, never cuts out on the highway over 45MPH
In all cases the engine has no spark, good gas pressure and only a period of cooling time is necessary before it comes back on and runs like a charm again, problem is also worse or likely to happen whenever the car is left parked on a slight uphill incline.
Fire Hazard Problems
Due to TIAE (Totally Incompetent Amateur Engineering) these engines are ill equipped with any actual engine temperature sensing telemetry. The electronic temperature sensor only measures the circulating coolant temperature (what you see on the dash) which is normally around 84˚ to 86˚ C (185˚ F) it will go up appreciably with low octane fuel or on hot days or when the A/C is on. The ECM has only that belated information as a past clue to how hot the block might actually really be.
On Pontiac GTP there is barely no airflow due to the tiny grilles and the low hood design concentrates engine manifold heat under the cowl due to passenger air intake seals and compartment-top can-pressurization the only convective air escape is through the side slots of the hood. A small amount of air escapes under the firewall, wheel wells which is barely adequate when the car is moving over 35 MPH.
Even though the rad is equipped with two dual speed fans, they hardy ever run unless the ECM detects some really high temperature nearing 100˚C (212˚F) in the coolant (only) or if the A/C is on (auto-low). In fact I have never gotten this engine hot enough for the stupid ECM to ever turn on fans for engine heat! The standard program fan trip points are obviously FAR, FAR too high, totally inadequate and outrageously defective! (86˚C would be much better for low, 92˚C for high)
This GROSS TIAE factory defect (no block/head temperature sensor) doubtlessly results in oil damage, oil thinning and burning (IAC evidence), CAT and EGR oil-tar corruption and the serious oil seal valve and cylinder head gasket damage that also resulted in the fire recall !!
Meanwhile there is a horrific low RPM and Parked heat riser build-up in the cylinder heads that normally exhibit 95-98˚C (205˚F) atrociously high cylinder head temperatures even when you use premium gas! . (much worse with regular)
At any rate, I totally hated that heat so put a 90˚C (192˚F) old wall furnace blower thermostat I had hanging around on the rad wedged behind the fan and wired back to ground out the two required fan relays to make the fans come on auto-magically whenever the engine compartment/top of the rad reached that temp. Surprisingly the fans came on quite a bit even in the cooler early summer weather city traffic, though never did the temp indicate anywhere near such on the dash!
In the meantime my stalls became far fewer though still happened especially on hot days, heavy traffic or after parking uphill.
Most interesting was the heat build up after the engine was shut down. Even on a cool day when the fans never ran while the engine was running, after 15 minutes parked they would start up and run 5-8 minutes. But not just once! After ANOTHER 25-32 minutes they would come on again and run for 25 minutes! This indicates heat sinking from the headers back into the unventilated block and heads with no coolant flow !
TDE (Totally Defective Engineering)
Now that I understood what the temperature fatigue cooling defect situation was with this car I got out my remote recording thermometers for a few test ride/sessions. I was sorely tempted to change the CPS and ICM (also reading tips here) but got out the schematics and noticed the "fail-safe" daisy chain configuration. One electonic engineering friend of mine who'd had an old Grand Am told me he was sure it was just the ICM (driver board and ADC (analog timing signal to digital chip and some darlington transistors in a flat pack under the HV coils), and just couldn't be the Hall Effect sensors.
Hall Effect sensors are the most reliable semiconductor switches ever developed they are impervious to wear, not near any heat down low there. and other than physical damage are susceptible to only three types of problems, metal dust/ferrous rust build up, overvoltage, or overdrain-current loading. None of those things could ever happen to them.
Moreover I was looking for a better (stabler) engine block temperature sensing position to relocate my auxiliary fan thermostat to, and that was when I had a revelation! I measured the temperatures on the ICM mounting platform....
The huge cast iron tray that holds the ICM and the HV modules over the fanbelts is bolted to the damned cylinder head! It is baking your ICM! Moreover even though it produces piss-all heat, the ICM is heat sink-coupled DIRECTLY to the engine block's #1 fire hazard producer!
When I measured the fire hazard ICM platform's temperatures they ranged from 199-238˚F!! It is higher than the heads and totally uncooled so cylinder head heat convects UP into it with nowhere to ever go!!! This outrageous heat changes the analog input sensing characteristics of the integrated circuit that receives and converts the CPS engine timing signals from the Hall Effect sensors (also cam one) and in turn directly computes (based on ECM ratiometric and control instructions) and outputs the spark signals to it's high gain darlington output transistors which drive the coils plugged into it! Even without ANY HEAT SINK other than it's own, (hanging loose) the ICM would never EVER get hotter than 125˚F!!
Due to TDE the ICM is being used to cool your overheating cylinder heads! What happens is that when heat rises up into the ICM it's IC gets totally screwed up and it's input buffer is also naturally filtered for both low and high frequency "noise" rejection. When it heats up the resistance values change the filter characteristics especially for low ranges which causes it to lock up and be unable to reliably read the Hall Sensors signals below 1khz (800 hz idle and below especially) The CPS and Cam Sensor are connected directly to it. It's integrator sends the RPM to the Dash/ECM, while it also directly controls ignition sparks.
When the car is parked on a grade uphill the heat rises into the grossly defective ICM mounting "heater plate" even more, requiring a much longer time to cool off! Yhe ICM's mounting platform is an oven-baking component heater, stressing, disabling and and destroying the ICM!
The FIX (ignore the materials mentioned here use the newer improved method in my later post below!)
This will get the car working well enough to drive a while again, but it is not the best solution! Still lets too much heat up into the ICM
One must put a heat shield (or airspace) between the ICM and it's aluminum (so-called-"heat sink" ROTFLMAO) mounting hotplate which is bolted to it's cast iron cylinder head "heat-radiator" platform!
Obtain two approx 3" X 7" pieces of foil (silver or gold metal looking) Arborite phenolic (not truly effective enough, not doing it this way any more! see BELOW)
(Rest here is OK, just that six special washers must be used instead of the two Arborite sheets mentioned above - see below!)
Disconnect the battery positive and set safely aside.
Remove the 6 screws holding the coils and ICM to the hot block platform. You may have to temporarily take off one spark wire boot to get the socket in there on the HV side, replace it loosely immediately so as not to mix them up.
Unplug the HV coil modules and set aside in order (use tape and number them 1.2.3)
Lift the ICM from the heater, you don't need to unplug it but watch out for the passenger side wire near the sharp center bolt-head holding the connector in, bend the (blue) wire down so it doesn't get it's insulation scraped on the stupid connector bolt-head.
Take your six special washers (see below) and place them around the stud-bolts (6 threaded studs) of the heater plate
Lightly re-bolt the ICM to the heater plate with the heat shields in place using the long screws on the lower HV module. (just to hold temporarily)
Get some removable grade Lock-Tite (a dab of Krazy Glue is fine) and permanently remount the upper two HV Modules being careful not to over tighten the bolts (snug-fit and hand tight is good, the glue will hold them! DO NOT OVER TIGHTEN THEM you might crack the HV modules!
Repeat for the bottom HV Module, recheck all cables and HV plugs and reconnect the battery!
Your ICM will now operate at it's own temperature and never cause you a problem again!
Smile and wonder how they stay in business....
Your car will never stall or not start on you again!
PS I relocated my auxiliary fan thermostat to the little 3/4" X 3/4" exposed corner of exposed iron on the upper left corner of the ICM heater plate, the fans function perfectly now as well. The car has run daily over a month now with absolutely no further problems!
