Intermittent hard-starting, no codes

About 25% of the time, my wife's 98 Lumina 3.1L presents us with a "hard starting" condition: it won't start on the first crank, and requires 1 or 2 more cranks before it will fire up. When this happens, we can usually smell gas resulting from the unsuccessful starting attempts. Once it does start, it runs rough for about 5-10 seconds (probably until it burns the excess fuel from the cylinders), and then it runs fine. Other than this, the car runs perfectly: it never runs rough, stalls, lacks power, or knocks. In other words, this seems to be strictly an intermittent starting issue.

The weird thing is that it doesn't trigger the check engine light or generate any OBD fault codes.

I've tried new plugs and wires; and I also did a primary/secondary resistance test on the coil packs - all 3 were well within spec.

My next guess would probably be either the ICM or the Crankshaft Position Sensor, but I'd love to have some other opinions before spending any more $. Or maybe I'm completely barking up the wrong tree by automatically assuming it's an ignition problem...

Any info or advice will be much appreciated!

First question does it only do it first thing in the morning or randomly anytime during the day? How about only after a hot soak?

Almost need to put a fuel pressure gauge on and watch decay, may have one or more injectors leaking. Should not drop more than 5 psi in ten minutes.

Sent from my Droid X. Typos probable.

Thanks, GM.

To answer your first question, it's pretty random. As best I can recall, it tends to NOT happen when starting cold first thing in the morning. It also doesn't seem to be related to ambient temp or wet/dry conditions.

As per your advice, I'm gonna try to get ahold of a fuel pressure gauge and see if I get a drop like you described.

A couple questions:
If it indicates a possible leaky injector, what's the best way to determine whch injector(s) might be the culprit - visual inspection?

Could a bad injector o-ring cause this, or is it most likely the injector itself is leaking?

Also, how hard is it to replace injectors on this engine?

As you can tell, I'm no expert mechanic - I probably know just enough to be dangerous. :smile:

thanks again,
JohnB

The way to isolate injectors is to first install a shut off valve on the fuel feed line,pressurize the system and then shut the valve if you did have pressure leakdown and assuming your pressure gauge is connected at the fuel rail test port if the psi doesn't drop with the valve closed then ur injectors are ok. If u were to still have a psi drop it would be injectors or leaking fuel psi regulator. It may be hard for u to get ahold of a shut off valve it is a special tool and u don't want to pinch off the nylon lines. But see if u do get a psi drop first and post ur results. As for replacing injectors not too bad. The upper intake comes off first to acces the injectors. If u did have a leaky injector best way to isolate them is to remove the injectors from the lower intake and with the fuel line still connected pressurize the fuel system by cycling the key a few times and place a paper towel under the injectors and see if the paper towel catches fuel drips.
Your second response leads me to believe u may have another problem though. U might try pulling the vacuum hose off of the fuel pressure regulator, which is near the throttle body and see if u have any fuel in the vacuum line or the regulator vacuum port.
Hope this helps. I do think fuel pressure drop should be verified first since it is an easy test to do if you have access to a fuel psi gauge.

You could also try tapping on the icm with the engine running and see if the engine misfires and also wiggle the connector.
Sent from my Droid X. Typos probable.

Thanks again. All good info.

I'll grab a gauge & check for a fuel pressure leakdown as you suggested. If yes, then I'll probably just take it to a mechanic to let him diagnose/replace injector(s) as needed. The process you describe for isolating the injectors makes perfect sense, but it's probably more than I feel comfortable attempting.

I'll definitely check the fuel pressure regulator & make sure there's no fuel residue on the "wrong side" of the diaphragm - good idea. It went bad a few years ago & I replaced it at that time. Not sure if that makes it more - or less- likely to be the culprit here...

Then again, the car has 230k+ miles, so one or more bad injectors wouldn't be surprising.

One thing that still mystifies me: wouldn't these injector/fuel-related issues trigger some kind of OBD fault code?

JohnB

Not necessarily. If it happened longer, probably but say if it misfires at start up the pcm needs to see so many misfires within a certain period of time(forget specs)to set a dtc.
In the case of a leaking injector the pcm has no way of monitoring that other than if the injector causes a misfire.

Sent from my Droid X. Typos probable.

Again, what you say makes perfect sense. Next time it happens, I'll try cranking it longer on the first try - hopefully long enough for it to throw a code. Although it'll probably just throw one or more "misfire on cylinder x" codes, which really won't tell me anything anyway.

In any case, I'm gonna try the fuel pressure test & check the fuel pressure regulator today. I assume the pressure test port should be readily visible somewhere on the fuel rail. Gonna consult a Chilton/Haynes book @ local library to get the pressure specs.

I'll keep you posted.

If you can have a fuel psi gauge and a spark tester connected while this occurs u would be able to see exactly what's going on when it happens if u can duplicate the problem.

Fuel psi spec 41-47psi with key on engine off.cycle the key a few times to get the highest pressure. The test port is at the front of the engine beside the power steering pump on the fuel rail. U may have to deposition the p/s hose to connect the gauge.

Sent from my Droid X. Typos probable.

Good idea. Of course, the tricky part is to get the thing to cooperate and "misbehave" while I have the fuel pressure gauge and/or spark tester attached. Call me crazy, but I may attach an inline spark tester, leave it installed as my wife goes about her daily business, and ask her to open the hood & watch it every time she tries to start the car. I could probably even use one of the old wires as an extension & rig it with wire ties so it would be more visible from the drivers seat. Know what I mean?

About the fuel pressure regulator: I pulled the vacuum line and there didn't appear to be any visible fuel residue in the line or vacuum port. There is, however, a very strong fuel odor in the vac line. Maybe it's normal for some fuel vapor to get thru the diaphragm, or could it just be the residual odor from several years ago when it failed and fuel poured into that line?

btw, I see the fuel psi test port and it looks to be easily accessible. Hopefully this engine doesn't require the TBI adapter or some other crazy thing.

JohnB

Yea there will be a fuel smell in the vacuum line. U also shouldn't need any special adapters to connect the fuel psi gauge.

Sent from my Droid X. Typos probable.

gmtech79,

Some diagnostic progress to report; hopefully you can help me interpret the results.

I left an inline spark tester installed for a few days while using the car, and I was able to determine that I'm getting spark during the the hard-start condition (or at least that one cylinder is getting spark :smooch:).

I used a fuel psi guage to run some tests; results as follows:

With key on / engine off, I get an initial pressure of 30-31 (spec is 41-47), and then it rapidly bleeds down as I watch. It falls to 0 psi within about 3 minutes. So besides failing the leak-down test miserably, the initial key-on pressure is way low to begin with.

The fuel pressure at idle is 30-31. Spec is 31-44, so it's at the very low end of spec.

When I pull the fuel pressure regulator vac line, it goes up by about 7 psi. So it seems to be doing its job.

Here's another important thing to mention: after all this testing, I noticed some fuel had accumulated in the fuel pressure regulator vacuum port and vac line. I had previously reported a negative on that.

Well, what do you think?

thanks again,
JohnB

Well with that info I would probably replace fuel pump and filter and also the regulator. That pressure is too low. Does it seem to have a lack of power also?

Sent from my Droid X. Typos probable.

Thanks, man. I already bought a new filter to install, I was waiting to access a friend's lift tomorrow.

I think I'll try replacing the regulator first before the fuel pump (cheaper & much easier); especially since it seems to be blowing fuel into the vac line.

Here's another thing worth mentioning: when I first turn the key to On, the gauge initially jumps to around 41 (low end of spec) for the first 2 seconds while the pump is energized, then it immediately falls down to 30-31 when the pump shuts off. So I'm thinking (hoping?) the pump is doing its job.

I'll keep you posted. Again, I really appreciate all your help & expertise.

JohnB

Good news: the fuel pressure regulator turned out to be the main culprit. The fuel pressures are now dead on spec, and there's absolutely no leakdown.

Also, prior to replacing the fuel pressure regulator, a new fuel filter brought the pressure up by 4-5 psi. So it was probably also a contributing factor.

Bonus: the car now has quite a bit more power too. This problem probably developed gradually over time, so I didn't realize how much power it had lost.

Thanks again, gmtech79. If it wasn't for your help & advice, I'd probably still be chasing a phantom spark problem.

best,
JohnB

Your welcome. Glad I could help and you solved your problem.