Cold Start: Runs for 20 seconds, then dies. Cause?

[RRoller123]

This is fantastic! I have long questioned why there is no focus on possible air leaks at the throttle plate? Every other joint and component in the FI system gets examined and discussed as a source of air leaks, but not the throttle plate shaft..... So it looks like you have uncovered something new! Well done, looking forward to receiving the o-rings, I will put them in mine and see if we can replicate the results. I also have a bad cold idle that I have been vexed by for literally years, after having tried almost everything, even swapping in/out a known good AFM.

[maxm50]

Thanks Pete! Let me know if it helps your idle issue also.

Here's the circlip pliers that I used - they are really great. You can adjust the tip spacing so that you actually fit the tiny circlip on the shaft. The other bigger pliers I had just wouldn't work for this one. Your mileage may vary. https://www.amazon.com/dp/B001B82G16/re ... UTF8&psc=1

And be sure to have a really good 1/8" tip flathead screwdriver for the throttle plate screws. If it has a small handle, use pliers to increase leverage.

[maxm50]

New Update: I figured it out!

My resistance of the coolant temp sensor was good at the sensor and good at the ECU. But, I noticed that the connector was quite wobbly/loose on the sensor - one half of it is basically broken off, so it's only holding on to one side. The metal "fingers" inside also weren't oriented well. I bent them back into place so that they'd have a good bite, cleaned with electronic contact cleaner, reinstalled the connector, and put a tight ziptie around it to keep it from wobbling around.

Fired her up, and for the first time in my short month or two of ownership, did not have to touch the gas pedal to get her to warm up! She settled in to a nice idle - maybe still a little rougher while cold, but she warmed up all on her own! Woohoo!

My theory: The coolant temp sensor was loose enough that while the resistances looked OK with engine off, as soon as I turned it on, the vibrations were probably enough to cause the connector to make rapid on-off contact with the sensor, rapidly alternating between infinity ohms and the actual reading. Thus, the ECU saw signals of "Cold as hell" to "correct reading", rapidly oscillating, and probably was freaking out trying to alternately pump in a ton of gas.

New spark plugs tomorrow (or later tonight), new O2 sensor, will go through the idle setting procedure again, and then hopefully all good to go!

Oh, and: Does anyone know where to buy new wire harness connectors to the coolant temp sensor (2 wire) and the cold start valve (2 wire)? Both of mine have one tab broken off. Maybe my zip tie is good enough, but eventually I'd like to replace the connector I think...

[tima01864]

I bought one connector from AR, Just be minful when cuttung and soldering the replacement. One at a time

[RRoller123]

I have an old harness that I can cut a couple of them off of if you would like, but they are of course unknown. Worked when the engine was stripped years ago.

[maxm50]

Thanks Pete! I won't take you up on that yet, but if my zip ties don't hold up (and I get more time on my hands) in the future I just might.

Related question - should I be applying dielectric grease to all of these wiring harness connectors (AFM, Coolant Temp Sensor, Cold Start Injector) as I reassemble them? Or do you all recommend leaving them "dry"?

[RRoller123]

That dielectric grease issue will cause crazy discussion here. I think it is snake oil, never use it.

[courtenay]

I use it on the elec. connectors on my boat - salt water Environment - no corrosion!

[RRoller123]

That is a reasonable application, in a marine environment. We always just "paint" the connections with spar varnish or poly after everything is wired up, but the Dielectric Grease is probably a better solution. Need something that is an O2 barrier, yet is easily removable for service. Not sure the long polymer molecules actually stop O2 very well. But it always seemed to work. At least keeps condensates from direct contact with the connections.

[maxm50]

Thought I'd post an update here for anyone who comes across this later. Here's everything I did on this issue:

First:
- replace all vacuum hoses
- replace main intake rubber hose
- replace throttle shaft seals (see previous posts)
- clean gunk out of AAV and check operation for smoothness and fully closing/opening
- replace cold start injector seal
- replace intake plenum gasket
- replace 2-way valve in emissions system and broken vapor lines near fuel tank
- secure connector on coolant temp sensor at front of engine so it makes good contact (big one!)
---- At this stage I had an engine that would start and idle, but after my post above I still noticed some odd behavior, occasional idle coughing, misfiring, etc.

Then:
- replace alternator voltage regulator and brushes (EASY to do without removing alternator from car) - I noticed my voltage shot up to 17V occasionally at idle and the idle would sputter and cough. This fixed that behavior.
- replace spark plugs
- replace O2 sensor
---- Now I had a smoother engine, most misfires/coughing gone, but not completely.

Finally:
- Send AFM to www.fuelinjectioncorp.com for rebuild. PO had definitely screwed with mine at some point.
- wait a week and then reinstall shiny rebuild AFM
---- Now it runs like a NEW CAR! Incredible difference in smoothness of idle and general operation. This is money VERY well spent.

Now I just need to fix an exhaust leak and get her smogged!