82 FI Won't Start Hot

[5strings]

Hi,

The car starts like a champ cold, runs perfectly even when warm, but it will not restart. It will splutter on the first attempt, then there's nothing during subsequent tries. Not a kick or a cough, just endless cranking. When it cools off, it starts and runs like normal. There is spark. When I connected a test light to the injector harness at the end of one of the injector plugs as per the troubleshooting guide, there is no pulsing light. So the injectors aren't being told to fire. The fuel relay checks out according to the tests in the guide. I haven't had the wiring harness off to check for a pulse at the #1 pin, that will be the next step, though I wonder why it would would only be absent on hot starts? If anyone has had similar experiences and can point me in a direction, I'd be much obliged.

I did change the fuel injectors and alternator this spring.

Thanks very much in advance, Robin

*UPDATE*

I got home tonight and was able to do some trouble shooting. To make a long story short my airflow meter flap is sticking and giving the ECU fits. I pulled the cover off the AFM to observe things and it sticks open at WOT. When that happens, the car predictably dies. But, even once it's freed , the car won't start again. What DOES get it to start is unplugging the ECU and starting the car. I didn't think it would start, but it does. I was just trying to check and see if there was power getting to pin 1 in the wiring harness, and was surprised when the car actually started. I dies fairly quickly, but it will idle for a few seconds. Then when I plug the ECU back in, it's fine until the flap sticks again. Why did this come up during hot restarts? I don't know, maybe the ECU relies more on the AFM when the engine is up to temp, and it was giving erroneous readings and triggering the "death mode."

This has left me with a few questions:

1. How does the car idle with no ECU if even for only a few seconds at a time? Doesn't it have the final word on telling the injectors to fire or?
2. Why won't it run if I reset the AFM valve unless I unplug the ECU, and start the car? Unplugging it and plugging it back in even if I let it rest for a minute or two doesn't work. I was so curious I tried it a few times to be sure. The car needs to be started without the ECU to reset everything.
3. Where's the best place to get an airflow meter? Bayless seems the cheapest by far from what I've seen.

Thanks again in advance. I think I've got a handle on the problem, but I'm still really curious about the questions above.

[rlux4]

Where are you located? The best thing to do would be to swap out the ECU with a known good one before you spring for a replacement. If you're not close to someone with a good one to put on your car you may try in the "parts wanted" section, there's guys parting out FI cars pretty often.
Ron

[So Cal Mark]

sounds like it's flooded and unplugging the ecu keeps the injectors shut, allowing the cylinders to get purged of excess fuel. After cranking it in the no-start situation, pull a spark plug to see if it's wet

[rlux4]

Have you cleaned the chamber of the AFM? Look inside and see if you can see where it's binding, it may need a light sanding to get whatever is causing it to hang up.
Ron

[majicwrench]

Mark has an excellent point...pull a plug when no start and see is soaked or dry.
Coolant temp sensor can fail causing horrible rich, thus NS. On many cars, for whatever reason, this occurs most often on a hot soak. Pull plugs and look. ECU pretty dependable.

[5strings]

Thank you for your responses and suggestions.

I'm in Frelighsburg, Quebec Canada about an hour southeast of Montreal. I know of a couple of Fiats in the area, but they've got carbs.

I did pull a plug and it was dry.

Once the AFM screws up, there seems to be no more power going to the injectors as indicated by my test light.

I have sanded my AFM before, and had it working well but it's back to being sticky, so the problem is getting worse. The little trap door doesn't close squarely to the wall when it's manually opened wide, so the shaft must be misaligned.

At this point, I figure whatever else might be wrong, the AFM needs to be addressed. So, I think, having nothing to lose, I'm going to take it apart and see if I can make some new bushings for it. I'll take pictures, and post them.

Thank again for your help on this!

[rlux4]

Good luck with that! (Just kiddng). According to Brad Artigue's manual the flap can be bent by a back fire. That may be your problem. Have you downloaded Brad's FI section of his maintenance manual?
http://www.artigue.com/fiat/tuning-book
Ron

[So Cal Mark]

power to the injectors has nothing to do with the afm, the injectors are powered from the dual relay. I'd check power from the ignition switch to the relay if you have no power to the injectors

[rlux4]

Diagnosing the dual relay:
http://www.wcmotors.com/personal/Fiat%2 ... agnostics/
Ron

[5strings]

Guys, thank you all again for your posts.

I have the Artigue guide. It's great!

I was misleading saying the injectors weren't getting power. What I meant was, according to the ECU diagnostic guide that I got from a link in this forum, they weren't "being told" to fire. Using the same guide, I did all the tests on the the relay, and it checks out.

Yesterday I did take the AFM apart.

I have taken a bunch of pictures, showing what I did in case anyone else wants to try it, but my ftp site is down at the moment. I'll upload them later if you guys would like to see them.

First, I took the guts out of the top, being careful to mark the orientation of the spring adjustment teeth. I drilled out all the fasteners on the bottom of the unit, and popped the lower cover off. I took the "gate" or "flapper" out, and used a dremmel with a flat sanding disc to do an initial pass on the upper surface of the housing. I used the scrape marks from where the gate was rubbing in the housing as a guide. I also took a little off the upper edge of the gate itself. Then, I put the gate back in, ran it through its travel, and took off a bit more, using the fresh scrape marks as a guide. I did this several times until there was no more interference. There are 2 bearings in there SKF# 394911ACNP. I might still order a set and drop them in, but I mine seem to be alright.

Once I got the upper surfaces to stop interfering with each other, I did the same thing on the bottom. I took off a bit, put the bottom back on, swung the gate back and forth a few times, and repeated the process, using the scrape marks as a guide. I guess I could have taken all the material from the gate, but I really wanted to leave as much metal as possible, and only deal with the offending areas. My unit was only rubbing in certain spots, not across the whole range of motion.

To put it back together, I used a dot of black RTV sealant on every point I drilled in the bottom cover, and a bit more in each the corner. On the top side, I put the spring back in and set it 3 with notches less tension. I figure I took enough metal out that more air is going to bleed around the gate, so it should need a bit less tension in the spring to let it open. 3 Notches is a TOTAL guess. I'm going to drive the car a bit, check the plugs, fuel economy, performance, and quite probably make adjustments. I did read on another thread that someone had moved theirs 5 notches just to get more power, so I figure at least still in the ball park with 3. But, it was snowing here yesterday, so I thought I'd leave testing for another day. After reading yet another thread, I bent the arm on the potentiometer, and reset the height of the arm so it would take a new path along the surface, and extend its life a bit.

The car runs fine now. It starts reliably hot and cold. If the temperature gets above freezing today, I'll get some testing and tuning done. I have a suspicion that I can go a few more clicks on the spring.

Robin