Fuel injected: good cold start ; poor hot start

[micbrody]

1981 Fi

I have a nagging problem; not major; but I can't seem to solve it: and I know others on the board have had problem; but never saw anyone who, after checking the common things, found the "uncommon" problem.

When car is cold, it starts immediately.
When warm (say after an hour or two since last ignition), the car will sometimes start. Other times it won't. When it starts, it barely starts (starter works, ignition takes place, and then car continues to run.....but always starts at extremely low rpm for a few seconds and then comes up to RPM of 900-1200.
Many times it will not "catch", and after the initial ignition, it conks out (may a half second). When immediately attempting to start again, it cranks but takes a good 5-10 seconds to start (sometimes I hit gas pedal....though not sure it does anything).

Things I have checked:
CSI and TTS works; and it is not stuck open. It have seen it turn on and then off (depending on how cold )

AAV - it is definitely functioning . I have cleaned it; I can see the shutter move depend on if in freezer vs heat; and when car runs , I am positive the electrical part works because the rpm goes down after a minute (when electrical connection closes valve).
I have measured resistance of coolant T thermo sender. Resistances appear normal at cold and hot; and I checked resistance of connector to ECU..... essentially no resistance.

I don't see any air leaks. I have inspected the big air tube for cracks; I have new brake booster. I cannot find sources for any unmetered air. I have tried spraying carb cleaner on all these potential airleak places ; no change in rpm.

I tested, inspected, and cleaned internals of double relay. All testing seems to show it very reliable.
I tested throttle position switch.

I followed procedure to set idle throttle: Ran car for two fan cycles; adjusted throttle linkage nuts to run at 850-900 with idle screw closed; and then loosened screw to raise RPM by an extra 500.

That leaves the AFM.

When I purchased car, the AFM had been opened by PO. I cleaned up all contacts; adjusted swing arm to glide over new part of resistive board; the fuel pump contact is very reliable. Door on AFM does not drag.
The only thing I can't be sure about is the spring force of the AFM door. I have not fiddled with it; there are no marks on it; so theoretically a PO could have adjusted spring tension.
One thought is about this tension:
I noticed when car is running and warm, if I take air filter off and just barely (as little movement as I can actuate) open the AFM door slightly more, the engine rpm goes down......
If I make assumption that PO did change the tension, how should I proceed? Does that tiny amount of extra door opening mean: I have made engine too lean? Or is the tiny movement electrically signaling ECU to send more fuel.....so too rich?
My hypothesis: the PO has loosed the spring so AFM makes mixture slightly more rich; but since I set idle by adjusting everything else, the actual baseline , non-load state of AFM is actually functionally lean: in other words: my baseline idle AFM door is too open; and I compensated by having throttle door and idle screw closed more than it should. When car initially turns over when cold, the AFM door opens up too much, but the CSI enriches mixture to eliminate the lean starting condition; when car is warm, csi does not fire, and now the mixture is too lean when starting....
Of course, this is all mental experimentation.

The big question: how should I proceed ?

[RRoller123]

If you are leaving the throttle plate at its current position and then manually advancing the AFM door, you would be creating a rich condition. If you think it is running too lean, it would probably seem to be too hot in the exhaust manifold.

You can check the voltage in the O2 sensor, it should average about 0.45 VDC or so, it will jump around but average right around a half volt. Check it at the sensor itself (T off), and then verify wiring to the ECU is ok. Given that this problem occurs when it is warm, I suppose that the O2 sensor could be the culprit, although I think I have read that it makes only minor adjustments.

Also possible that the injectors are leaking? Maybe causing a rich start condition after warmed up, left to sit, and the restarted. Any backfire or popping?

Pete

[micbrody]

I did check the oxygen sensor and it hovered around 0.45ish. However, that was at idle. What I am suggesting is adjusting AFM spring tension, and then recalibrate throttle plate and idle screw based on new idle flap opening of afm

[So Cal Mark]

have you checked to see if the system holds pressure when shut off?

[micbrody]

Not sure I understand: which system should hold pressure?

If you are referring to fuel stuff: I cleaned all injectors; replaced all high pressure fuel lines; fuel pump; and injector lines. They all held pressure

[micbrody]

Maybe another way to approach issue:
Let's say I know for a fact that po adjusted spring tension; but left no markings as to direction and how many teeth the spring wheel was moved: how would I proceed to reset spring tension to factory original spec?

[So Cal Mark]

your pump has a check valve that should maintain pressure in the system once it's shut off. That prevents vapor lock and long crank times. If the system doesn't hold rest pressure you either have a leak or a faulty check valve. Leaks could be internal or external.

[micbrody]

Last year I went through the whole fuel system; I did have a pump that's was noisy; and when checking pressure in fuel rail (off of csi line), was anywhere from 23-37. I replaced with new fuel pump; pressures better. Pump quiet.
I really don't think it's a bad check valve as pressure was good when installing new pump (pump has the internal check valve). The other reason is that if I let car completely cool overnight, the engine starts immediately. It's only when engine is warm within a few hours of running.

[RRoller123]

It seems you have gone through most everything, but I keep coming back to possible leaky injector/s or maybe CSI?

[spider2081]

If the temperature sensor in the coolant "T" or its connection are not in good shape that could be your problem. The lower the temperature the greater the resistance the temperature sensor resistance. So when its cold the resistance is high like the connector being unplugged. As the engine temperature increases the resistance of the sender decreases telling the computer to lean the mixture. I think when the engine is at 190 degrees the resistance between the pins on the sender should be about 200 ohms. When the engine is cold it should be 2000 ohms or greater.

[micbrody]

I did check resistance of coolant Temp sender when cold and hot. Maybe there is a "linearity" problem with sensor....meaning resistance is changing at wrong "rate" relative to temperature change. I do have an extra "used" one; Maybe I will switch out.

With that said:
PO did open up AFM. When I reopened up, it looked unremarkable in that the sweep arm was still carving the same groove on resistance surface. The fuel pump contact was working. So............what did the PO do when they opened it up? The ignition system was a little screwed up (fuel pump always running; and parallel, separate switch to energize starter). I cleaned up double relay; and some connectors ignition switch to return ignition to factory condition; so maybe PO checked AFM and did absolutely nothing, seeing everything looked fine.
The other hypothesis:
Previous owner was looking into making car perform better. The air filter housing is not factory . It is some mushroom looking thing with a dense nylon , washable filter. And he/she put a roll bar on car. Maybe he/she saw some of the posts on internet as to adjusting spring tension to have car run richer........

[satufiat]

I had the exact same problems as you described. Low idle...almost stalls during warm up and hard starting when hot. ('81 2000 FI)

I just replaced the Pickup Coil in the distributor(corroded wires), Ignition Control Module, Rotor and Cap and Plug wires. While I did that I thought I would check the Idle Mixture screw. Per the manual, I closed the mixture screw(was out about 4 turns). Adjusted the throttle stop for 800 rpm. Then adjusted the mixture screw to bring the idle up to 900 rpm. (about 3/4 turn out).

Perfect! took care of all of my problems. Starts just fine when hot and no low idle during warm-up.

Of course I also just changed out the rear trailing arms(Short and Long), as well as the Panhard Rod, but I don't think that contributed to the solution.

Edit: I also took out the AAV, clean and tested, but it seems it was working properly prior to cleaning.(heated with a heat gun and cooled in the freezer)

[satufiat]

I also wanted to add that I didn't check the throttle position switch. But after adjusting the Throttle stop /mixture screw, it may have had an impact on the throttle position switch adjustment. The previous owner had the mixture screw out way too much wihich means the throttle stop was also wrong.(could have impacted the Throttle position switch adjustment??)

Also, I tested the Coolant Temp Sensor in the 'T', 4K Ohms cold(60 deg F ambient), and 664 Ohms at operating Temp(180-190 deg). Maybe a little high at operating temp, so I may replace but-she is running sooo smooth right now, maybe best to leave it alone.

[aaalbedo]

I have a similar problem with my 84; good cold start and often no warm start.

I think it is my time temperature switch, if it can fail this way: I think it is running my cold start injector even when the engine is hot.

At any rate, a work-around until I can investigate further is to give it half throttle when starting when warm. This has never failed. And if I forget the half throttle and it fails to start, I can disconnect the cold start injector and give it full throttle to start. In this case there is a brief delay before it fires, consistent with flooding.

Hope this works for you, too. Someday soon I'll get around to testing that switch.

Larry

[Daveje]

I have a similar problem with my '85. Cold start is fine, hot start takes extra cranking, and when it does start the tach is at about 500 and works it way up to 900 with a little driving. When it is especially hot the car will start very rough and struggle to stay alive until it picks up.
Today I tested the fuel pressure, seems low at 26psi@idle, but also the pressure bleeds down to zero in less than a minute when the car is shut off. When I pinch off the fuel inlet hose it holds pressure so I assume the pump check valve is shot.
I've ordered a new pump and I'll see if it solves the problem.