Losing couple cylinders after hot soak condition for a short time

48K mile 85 GLHT Log manifold with Spearco intercooler kit. I have green top FFV injectors in it, running around 47 PSI static pressure. DC Stage II LM. One of my nice flowing cylinder heads and 60 trim t3 compressor wheel. Many of you have seen this car at SDAC shows throughout the years.

Anyway here is the issue. When I get into a hot soak condition during summer months and car sits for maybe ten minutes, it starts and only runs on two of it's four cylinders. 1 and 2 are dead. I have a wide band o2 and it is showing severe leanness (greater than o2 band limit). After about 1-2 minutes the engine will once again start firing on all four injectors, picking up the dead cylinders gradually one at a time, and wide band o2 will return to normal and switch back and forth trying to achieve stoich. The vehicle will run perfect until it is shut off again and allowed to heat soak. It has done this problem rather consistently over about a years time. The vehicle is not running hot at all, just normal operating temperature. If allowed to cool for an hour or so, the vehicle starts up and runs completely normal. It most definitely is heat soak related. It does not happen when temps cool down in the early fall.

At first I thought maybe the 2 pump internal/external fuel pump had finally decided to start to show it's age, so I changed that out to a 255 Deatschwerks pump, single internal pump assembly, along with new sock filter and fuel filter. That wasn't it. I then had the recently cleaned, flowbenched FFV injectors retested. Seeing how the car was acting like many TM running +40's on low static fuel pressure, my fuel injector guy (Bill Johnson @ Mr. Injector) thought maybe the injectors were dribbling fuel into a couple cylinders and were misfiring due to overrich condition. It mimicked that exactly, but I had my doubts because of the extremely lean wide band o2 readings when this occurs. Well that wasn't it either.

Something seems to be overheating possibly in PM and 1-2 injector driver shuts off temporarily, only to come back on after a couple minutes. Injector harness is perfect (very low mileage) and power module still has inlet air being pulled through it. So my question is has anyone else seen this happen before?

Todd

what are you using for a fuel rail?

sounds like there might be air in it causing vapor lock

Thanks for the reply Dr. Johny. Stock log fuel rail. It has the factory insulation blanket the turbo cars run on top of the turbo and under the injectors still on it. That is in very good shape. The stock rail also has 3/4-1" wide Thermo-tec header wrap applied to the entire rail as well. Then over the top of that is high temp electrical tape and over that is more heat reflective tape. The AFPR has the relocated mounting bracket. That was a recall back when these cars where near new. Car has been together like this since 1987. Never had this issue until the last year. Turbo also has external heat barrier applied to exhaust housing. I would surmise that should help hold heat down a bit over a stocker. At least that was the claim of the coating company. Using non-oxy premium fuel.
Yes I agree it does sound like vapor lock, but fuel flow is plentiful through lines, rail and regulator. One of my questions to Mr. Injector (Bill) was is there any test or history for possible injector failure under high heat conditions? He said not that he was aware of. Possible it could be Accufab regulator acting up. Hum...that is the one thing I changed a few years ago. I had an older Buick GN/Bosch AFPR regulator on it before. I changed it out strictly for cosmetic reasons.
One thing I could check to rule out vapor lock would be to temporarily manually run fuel pump by applying 12 volts to positive side of coil. Could run for a minute or two before attempting to restart in hot soak condition. I also could leave my pressure gauge attached to Schrader valve to see what happens to fuel pressure in hot soak conditions.
Todd

Check the wires in the fuel injection harness at the engine harness connector. Two light brown and two dark brown wires from the injectors join together (Light to light, dark to dark) at the harness connector. I've seen these connections fall apart before. Expansion from heat could cause the connection to be lost.

Could be the HEP too.

Could be a ground, maybe one of the many in the harness where several ground wires join together.

Does your car have the TSB update to relocate the fuel regulator?
This will be a longer angled bracket that lifts the regulator further from the intake.
Additionally, is the thermal shield installed and without any deterioration of the fiberglass?

A few additional thoughts...
Pre-1990 engines are 'banked' fuel delivery.
1 and 2 are paired as are 3 and 4.
This is worth mentioning due to the early Turbo-I fueling strategy.
The original turbo-I software has a feature that would bias the fuel delivery for each bank, with cylinder 1 -2 enriched due to intake air distribution issues.
This feature was removed when the larger/longer runner intakes were introduced.
Furthermore, when the DC intercooler kit was introduced the fuel biasing was switched to enrich cylinders 3 -4 because the DC intake has the air entering from the opposite end.

If your S-II LM is for the DC intercooler kit, it will inherently have 1 - 2 even leaner with the Spearco system because that system flows into the original driver side.
If your S-II LM is for a T-II, the fuel biasing is non-existent altogether.

Sorry, tried to reply with quote but this isn't working for me. Andy the 6 pin connector terminals and associated wiring are in perfect shape. Wires are still flexible with zero cracking or anything like that. I will remove all the pins, clean and reassemble just to make sure it isn't that. The injector harness maybe has 10k miles on it. The engine bay harness has 48K miles on it. I tried a different HEP a year ago when it first started acting up, that wasn't it either.

Thanks for the reply Ken. As noted in post #3, the factory fiberglass shield is in very good shape. It also has additional 2" Thermo-Tec header wrap sandwiched between the intake and exhaust runners to further keep heat away from intake manifold. I also have a NOS heat shield should this one ever deteriorate. Also the regulator has the TSB relocated bracket and is actually further moved away another 2" or so I could use an Accufab Buick type AFPR regulator on it. This has been relocated like this since 1987.

DC LM is for non intercooled 85 g-head applications. I had bought a 85 DC intercooled LM for Turbo-Auto intercooler and it was a pretty substantial downgrade. I remember back in the day to have been pretty disappointed in that LM. Car felt like it lost 50+ HP! It seemed like spark advance was altered a whole bunch too. Car was a pooch running that LM compared to running the non-intercooled stage II LM. As noted, #1 & 2 cylinder was very lean using the intercooled LM. Could easily tell that back in the day just by reading plugs. It wasn't in my car long before it was removed and thrown in a box. Interestingly though, I've only hurt ring lands in stock cast pistons on #3 & 4 way back in the day with using the non-intercooled DC stage II LM. Has Ross forged pistons in it for the last 20 years or so, with top ring gap opened up to around .020".
Todd

I would swap 1 2 injectors with 3 4 see if it follows.. if it doesnt. swap in another LM.

Quote:

---

Originally Posted by 4 l-bodies

Sorry, tried to reply with quote but this isn't working for me. Andy the 6 pin connector terminals and associated wiring are in perfect shape. Wires are still flexible with zero cracking or anything like that. I will remove all the pins, clean and reassemble just to make sure it isn't that. The injector harness maybe has 10k miles on it. The engine bay harness has 48K miles on it. I tried a different HEP a year ago when it first started acting up, that wasn't it either.

Thanks for the reply Ken. As noted in post #3, the factory fiberglass shield is in very good shape. It also has additional 2" Thermo-Tec header wrap sandwiched between the intake and exhaust runners to further keep heat away from intake manifold. I also have a NOS heat shield should this one ever deteriorate. Also the regulator has the TSB relocated bracket and is actually further moved away another 2" or so I could use an Accufab Buick type AFPR regulator on it. This has been relocated like this since 1987.

DC LM is for non intercooled 85 g-head applications. I had bought a 85 DC intercooled LM for Turbo-Auto intercooler and it was a pretty substantial downgrade. I remember back in the day to have been pretty disappointed in that LM. Car felt like it lost 50+ HP! It seemed like spark advance was altered a whole bunch too. Car was a pooch running that LM compared to running the non-intercooled stage II LM. As noted, #1 & 2 cylinder was very lean using the intercooled LM. Could easily tell that back in the day just by reading plugs. It wasn't in my car long before it was removed and thrown in a box. Interestingly though, I've only hurt ring lands in stock cast pistons on #3 & 4 way back in the day with using the non-intercooled DC stage II LM. Has Ross forged pistons in it for the last 20 years or so, with top ring gap opened up to around .020".
Todd

---

Thanks Todd
It sounds like most things are in order and if you take the time to swap the injectors as another member suggested, inspect the "pinch" connection in the injector harness.
Back in the day, the harness wires were pinched/struck together, without soldering, which was more like a friction weld.
Over time, they like to come apart or corrode due to poor insulation and isolation from the elements.

Additionally, the associated pins to the injector circuits, at the power module and at the injector, should be inspected for corrosion or pin spreading.
There are two opposed fingers inside each connector that should touch each other as you look in the power module/injector connectors.
Don't crush the pin sockets - use a strong mechanics pin or stick pin to bring them (only the connector fingers) back to center.
Btw - the pin sockets at the power module are the same pin sockets used on the black and gray distributor plate connectors.

Since this seems to be isolated to one injector bank circuit checks may be worth while.

My friends minivan did this exact thing. Unfortunately he didn't find a fix before selling it and he didn't give it to me to drive for a while to see if I could figure it out.

He'd park run into the auto parts store for 5 minutes, start it up and mis-fire on a couple cylinders. Like you said if it cooled for an hour+ it started fine. He swapped fuel rails from a billet back to stock, ECM's from a boost button flashable with my cal on it to a stock ECU. No changes. That's about all I know about it so I'm curious to see if you find a fix.

Todd, are you using your wideband output for your O2 signal? WB2NB?
I recall having some similar issues, however i never narrowed it down to specific cylinders. Mini woud be hard to start hot after just a few minutes off on a hot day. And when she did she was lean.
It was probably over a year ago now, and i haven't driven in much under the same circumstances due to having the electric water pump.
I believe it was resolved through adding a bit of key on prime, and the default value for warmup of the WBO2 (which shouldn't matter since starts should be in open loop) I think i may have also altered the warmup fuel curves.

Any type of misfire will cause an O2 sensor to read lean. O2 sensors read oxygen and a non firing cylinder is not using up any of the available oxygen so looking at the wideband for a diagnosis will not help. I would suggest picking up a set of noid lights at the parts store or HF and having them with you. Next time a misfire is imminent, throw the noid light on and verify you are getting an injector signal. If that checks ok, I would then let it hot soak again and pull 1 and 2 plugs and check for coolant seepage into the cylinders.

Quote:

---

Originally Posted by shackwrrr

Any type of misfire will cause an O2 sensor to read lean. O2 sensors read oxygen and a non firing cylinder is not using up any of the available oxygen so looking at the wideband for a diagnosis will not help. I would suggest picking up a set of noid lights at the parts store or HF and having them with you. Next time a misfire is imminent, throw the noid light on and verify you are getting an injector signal. If that checks ok, I would then let it hot soak again and pull 1 and 2 plugs and check for coolant seepage into the cylinders.

---

Good point and well said.
Raw fuel un-burnt fuel will produce a lean sensor read.
Although, coolant intrusion into the cylinder comes with the witness of white smoke upon start up.

Thanks for the replies and the ideas guys. I got the freshly flowed and tested injectors back in the car. Also took both sides of injector/body harness connector apart and cleaned and inspected the terminals. All seemed good there, but it has light coating of fresh grease on them now. Also inspected the injector terminals. They looked like new (10K miles on them). It is up and running again, but haven't got it into a hot soak condition yet. Had other things come up the last few days, but I will report back when I have some answers. I'm also going to look at 1-2 injector wire on LM connector. I do have one of FWD's stage II LM's I can throw in there to rule out a LM issue should I continue to have this issue. That is supposed to be close to a duplicate of what I currently have installed in vehicle now.
Wayne, This GLHT is pretty old tech. It is running on it's narrow band (in stock location) and has a AEM wide band around 25" downstream in 3" downpipe. About the ONLY thing newer tech on this car is a wideband, otherwise it is straight out of 1987-1988 (lol). I stopped trying to make this car faster after it kept busting 525's as fast as I could install them.
Todd

I was gonna say what about the fuel hoses?
but after reading ^ I'll modify my thought to , how OLD are the fuel hoses

I know the log motor fuel rail is a loop but I can't remember for sure without looking , which fuel injector would see fuel first
number 1 or number 4
if it's 4 , and the fuel hose is pinching shut then you might not have enough fuel flow to feed injectors number 2 and 1

Todd, do you have a spare ignition coil to swap in as a test?
When i had one going bad (and then eventually fail altogether) it gave similar lean readings due to the misfires. If your coil is what is getting heat soaked and flaking out, it could be on its way out.
And is the ignition coil in the factory location? Mine was also, with the long coil wire.

When the problem was intermittent, it was really wearing me out, trying to find just what was causing those conditions.

Once again sorry for not replying with quotes, still not working for me. Dr. Johnny, car is impeccably maintained. It just hasn't had much performance enhancements done to it in a long time besides wide band and normal maintenance. So to answer your questions, all fuel hoses, filter, and injector clamps are fresh all the way to the tank. Both feed and return were checked for flow at time of new fuel pump install. The log rail feeds #4 first and #1 last. As noted earlier, they're both heavily wrapped, so I sort of doubt it is vapor locking or pinched as it reacts the same way after new fuel pump, filter, and lines were installed.

Wayne, the car has a MSD 6AL (hidden in fender) for the last 25+ years. It has had both Blaster II coil, and new stock coil in stock L-body location. Recent black 8.5mm MSD wires on it too. It also has Rich Bryant's HEI brass terminals inside the 1.7 VW dist. cap. So in other words, it does away with the spade terminals (that wear) like in std. 2.2/2.5 later model caps/wires. I replaced Blaster II coil after I lost a HEP in car. There was some theory of different resistance in aftermarket coil windings was to blame for premature HEP failures. So it now has a stock Mopar coil that is about 5 years old. Has maybe a couple thousand miles on it. Yes, I can certainly throw a number of replacements coils in it to check that. I know a few people aren't impressed with MSD stuff, but this CD ignition was purchased and installed back when they were assembling in El Paso, TX. Long before MSD probably was outsourcing and assembling parts outside North America. FWIW, I've had MSD 6AL's on half dozen cars over the years and have had zero failures. Wire them correctly, and you can run jumper wires on coil bypassing the MSD box should something ever fail internally in the MSD box. Feel pretty confident in the ignition system, but since it is electronic, anything certainly can fail. It should be superior to a Mopar standard ignition system.

Cold and rainy day outside today. High of 57°, so no chance to test heat soak conditions today.
Todd

ok I hadda ask

have you tried swaping the 1&2 for the 3&4 injectors yet ?

the only times I've ever had an injector drive fault code it was for 1&2 and it was a bad injector both times .. and the issue came and went until I swaped the injectors out

lol
I always liked the stock cap and wires .. granted they're a cheep as hell set up but I always got the extra use out of them by flipping the terminal on the end of the wire around and bending it slightly towards the rotor

fresh new contact after only a couple turns of the dist when started again

Quote:

---

Originally Posted by 4 l-bodies

I replaced Blaster II coil after I lost a HEP in car. There was some theory of different resistance in aftermarket coil windings was to blame for premature HEP failures.

---

that sounds improbable. i can see if the HEP was the kind that has transistors in it that directly drive the coil but all the HEP does is send signals to the ecu.

unless the theory is that the hotter spark dancing around in the distributor is somehow arcing to the HEP but thats improbable as well.

my thoughts on HEP failures is from the small wires getting jiggled to death from unsupported wiring harness stuff.

Brian

Todd,
Here's another thought, since I'm determined to try and contribute something useful...
When you have this hot soak condition before you try for a start, can you log your coolant temp sensor reading the ecu sees?
If the sensor is reading faulty or off, it could be starting you off with the wrong injector prime, and in the wrong spot of your warmup curves.

Wayne,
That is a good idea. I appreciate your thoughtful responses. You other guys too. I'll have fuel pressure gauge and DRB hooked up so I can try to figure this thing out. Injectors all passed with flying colors for dribbling. In fact they were left on the machine overnight trying to get them to pee onto table. What they couldn't test was how they would respond when hot. Hopefully, tomorrow will be hot enough to try some more heat soaks. I have several stock and another stage II LM to also try should it still have the same issue. I'll have a different coil ready as well.
Todd