This problem has been hard to narrow down but I think I've finally done it. There are three choke states on my 1989 351 fuel injected windsor. Cold, Warm, and normal. This problem only happens in the "warm" state. This is why it took so long to get it to consistently reproducible.

Only in the "warm" state there is a dead spot on initial acceleration. Idles fine, runs fast fine, but when trying to get off the line from a dead stop the engine faulters and dies. When in this state it happens every time, without fail. Outside of this state everything works perfectly.

Cold, starts, drives, runs perfect. Of course this state doesn't last long so now we're in warm state and constant fumbling. While driving this is enough to kill it. Then once it gets into run state everything is fine again.

I'll see if I can pull codes but last time I did it they were all over the map.

Any ideas?

 

Might not be quite the same issue, but my 93 5.0 did something similar. It would drive fine for the most part, but pulling away from stops, and pulling into parking spots, it would buck and stall, shift into neutral, it fired fine, and off it went. Had this issue for 4 years, no one could diagnose it, until one shop checked the numbers the computer was seeing.

What it turned out to be, was the intake air temperature sensor was faulty, so it would read much higher than it should have in those situations, causing the computer to attempt to compensate, and killing the engine.

If you have a scanner that can monitor the data for the sensors, it wouldn't hurt to have someone look at it during a test drive, see if things are in spec when the engine falters.

 

When it stops raining I'm going to try to get a few failures and engine codes.

I have an 1987 shop manual. Is there an area that explains the warm up modes this goes through? It very obviously has three distinct stages as the engine warms up.

 

Strategies:http://web.archive.org/web/201104140...tion.com/?p=64

 

The fuel mixture needs are sent to the PCM via the temp sensors and o2 sensors… they change zillions of times per minute (yes zillions IS a technical term)… there aren't any 'stages'… and no choke.

 

The PCM uses different strategies based on the feedback from those sensors. Click on the link in my previous post.