I have read many of the threads on this topic about the DME sensor, fuel pressure regulator, AFM, etc, but it seems mine is acting a bit different than many. Basically because it is so inconsistent. Sometimes it will start right up hot or cold. Other times if I turn it over a few seconds, stop and do that several times it will start and sometimes it will not start unless I give it a shot of starting fluid. Once started it runs like a champ with no problems, but if you stop somewhere you just don't know.

Any ideas?

 

How many miles?

You've got to do some troubleshooting .... I would start with a rail mounted fuel pressure gauge based on your indication that starting fluid helps it. I like the one from Pauer Tuning.

 

Yes, basic troubleshooting time. You need to know what's missing when it is not starting. Fuel or spark, probably. Figure out which and drill down from there.

You can try the DME relay temporary jumper, check the fuel pressure (as suggested) hook up a Noid light, Spark checker, etc.

 

My bet ia a failure in the injector wiring. My turbo exhibited the same symptoms.

 

If it helps, I experience a similar issue:

At first I thought it was completely random. I'm getting tach bounce, new DME relay etc... What I noticed though, is that the amount of time it sits usually dictates the starting condition. Further observation told me that the amount of time I allow the engine to cool dictates the starting condition.

For me: once the water temp needle is down below the first large tick, its almost guaranteed to be a no start. I isolated this to the DME temp sensor. On second thought, I had replaced it with a new one, so it likely isn't the issue. The second condition that I noticed having an effect was ambient temperature. This could be feasible, since coolant temps will likely have a direct correlation to under hood temps (which aren't measured by the DME temp sensor). Two variables that I noticed increasing the likelihood of a healthy start were temperatures below the 40º-50º range and/or the car sitting for more than 3-4 days.

My current guess without busting out any fancy tools such as pressure gauges, noid lights etc. is that there is an air temperature sensor (probably in the AFM) that is starting to get wonky.

This may not be your issue, but my point is that even just observing all of the small details and conditions that a no start goes by can help cut down significantly on the number of possible causes (or at least make the process of elimination easier [and cheaper] to work with).

 

I have thought about all the points you all have said, but kind of dismissed a few (probably prematurely) based on other factors. I didn't think it was a spark or fuel pressure issue because it cranks easily with a small shot of starting fluid and runs fine from there. I wouldn't think it would do that is it had fuel or spark issues, but it wouldn't be the first time I was wrong!

It does have the tach bounce but the amount of time it sits and the temperature don't seem to be a factor. I have had it sit weeks at a time and one time crank within 3 rotations and again not at all. Same with being up to temperature.

 

I am not sure if the check valve would fail intermittently, but if it always starts once it gets running, and usually has a hard time starting after a period of time, I would suspect a bad fuel check valve.