Cold start trouble - but no issues when warm...

[mikee30]

Hi all
Looking for some ideas on resolving a start die condition only occurring when the car is cold. The issue can be described as follows:
Startup attempt #1: Car almost starts – catches but will die immediately.
Startup attempt #2: Car will start and die within 1-2 seconds
Startup attempt #3: Car will start and will idle well for ~5-10 seconds – then starts to get erratic and will eventually drop to 400 rpm, then eventually die.
Startup Attempt 4 – Same as #3 or I give gas until it is warmed up and I can drive off.

I can sometimes keep the car running on Attempt 1 or 2 if I give a lot of gas. Sometimes giving gas at a steady idle of 1500-2000rpm will work, but sometimes while giving gas to get the car warmed up, the engine will try to shut off forcing me to give more gas to try keeping it alive.
This issue does not occur when the car is warmed up. I can start up without any issues if the car’s temp gauge is at ¼ to ½ way mark.
Here are some items I tested/replaced (also I should note the prior owner installed Authority chips):
1) New DME temp sensor
2) Found a short in my DME temp sensor wiring under the connector boot as indicated in the carpokes test procedure – spliced and repaired. Tested wiring all seems fine.
3) Tested TPS – checks out ok
4) Tried a replacement AFM from a working car – it did not help
5) New Fuel Pressure Regulator
6) New Fuel Pump Check Valve
7) New Fuel Filter and air filter
8) New spark plugs
9) New heater control valve (unrelated…)
10) Prior owner did new ignition wires and new fuel injectors
11) Checked for gas in the vacuum line of the fuel dampener – nothing…
12) Checked for any fuel pooling in the trunk under the cap that is under the carpet – nothing.

Some other items I noticed:
1) The car has a smell of raw fuel when sitting – seems to come from the rear…can not pinpoint. No leaks. Also very light smell near one injector, no leak. I only smell it when I am outside the car, especially when parked indoors.
2) I noticed my ICV connector has some exposed wiring near the connector boot – but does not seem to be shorting out. I will replace this…but doubt the ICV is the issue. When car is running, I can hear the ICV cycling.

I was told by the previous owner work was done to the fuel tank…he thinks it was replaced but didn’t know for sure.

I did a smoke test and found a few small leaks…repaired the ones that were largest, but there still remains a few (here's a post regarding my smoke leak test: viewtopic.php?t=3442)
1) Smoke leak from the bottom of the ICV valve (near the small openings of the casing)
2) Smoke leak from the top cap of the cycling valve (one hose was also leaking but that has been fixed).
3) Smoke leak from the throttle body spring on the right hand side.
4) The smallest port on the jboot had some smoke. It has the original squeeze clamp - it is either the clamp or the joint of the hose entering the port that is leaking slightly. I would like to replace the hose and clamp but can’t seem to find the hose part number. Can I use a regular screw type clamp versus the squeeze clamp here?

I don’t know if small smoke leaks are acceptable…but even if I repaired them all, it seems odd that the car does not have any issues when it is warm. It’s almost as though the car loses fuel pressure when it cools down. Looking for some ideas on how to proceed before I remove my intake manifold and see if I can find anything and tackle each air leak from the smoke test…. Any help would be greatly appreciated!

[Tom]

Sounds like it is not getting the cold start enrichment it needs. Be sure to test the temp sensor at the DME pins, along with the temp sensor in the AFM, and the idle contact in the TPS, all at the DME connector (to root out harness issues). Also check the O2 sensor. If all that checks out, a weak ignition can also cause this. Have you checked the spark when cold to see if it looks strong? Bad cap, rotor, plugs, wires and even speed and ref sensors can make the spark too weak (or non-existent) on a slow cold crank.

[mikee30]

Sounds like it is not getting the cold start enrichment it needs. Be sure to test the temp sensor at the DME pins, along with the temp sensor in the AFM, and the idle contact in the TPS, all at the DME connector (to root out harness issues). Also check the O2 sensor. If all that checks out, a weak ignition can also cause this. Have you checked the spark when cold to see if it looks strong? Bad cap, rotor, plugs, wires and even speed and ref sensors can make the spark too weak (or non-existent) on a slow cold crank.

Hey Tom! I did test the temp sensor at the DME pins and it was fine. Since I swapped out AFM's with a known working one, I eliminated the actual AFM as a problem spot. The TPS was also tested based on the TPS test procedure you created...so I eliminated that too.
I was under the impression the O2 sensor does not activate until the car is warmed up. I did try disconnecting the O2 sensor on cold start up, but same problem.
I haven't tested the spark when cold - but given the ignition wires and spark plugs are new, I eliminated that too.
My reference sensor holder is not in great shape - but still making a connection...

Should I rule out a vacuum leak and/or fuel pump?

[Tom]

Humor me and take a look at the spark - see if it looks strong. And how is your tach bounce? Decent bounce or little twitch?

[mikee30]

Humor me and take a look at the spark - see if it looks strong. And how is your tach bounce? Decent bounce or little twitch?

Checked the tach bounce...it did bounce up to the first rpm line (actually a hair below the first line).
I see sparking occurring - not sure if is considered strong or not, but seems normal?

[barnwerks]

"ruling things out" because particular parts are new is not good diagnostic procedure. What are the injectors? What is the fuel pressure? Try bypassing the fuel pump and DME at the DME relay socket to see if the symptom changes.

[mikee30]

"ruling things out" because particular parts are new is not good diagnostic procedure. What are the injectors? What is the fuel pressure? Try bypassing the fuel pump and DME at the DME relay socket to see if the symptom changes.

How would I bypass the fuel pump and DME and what should I expect when doing so? I don't see how the car would run when bypassing these items?

[barnwerks]

What I wrote is confusing, bypassing the relay by connecting terminals 87 and 87b to terminal 30. This will power the fuel pump and the DME directly.

[Zirconocene]

If/when you power the fuel pump directly from that relay socket, I think it's worthwhile to a) use a fused circuit and b) use a momentary switch to that it only happens when you want it to happen, and not the second that you connect the terminals.

Cheers

[mikee30]

What I wrote is confusing, bypassing the relay by connecting terminals 87 and 87b to terminal 30. This will power the fuel pump and the DME directly.

I'm not sure I follow what exactly this will troubleshoot given the issue does not occur when warm, but I will try!
Is the idea I short the connections as described above and start the car to see if it starts first time around? On a different note, I have a few new DME relays which I swapped out and it did not make a difference with the issue.