So help me diagnose my engine sputtering

[georryan]

Ok the symptom:

Ramdomly when driving my car will sputter. It usually will happen if I floor it, but recently it has been happening at different times. For instance, I pulled out of a parking lot and went to put it in first and the car idled fine, but the second I'd give it any gas it would sputter like it was going to die, and as soon as I let off the gas it was fine. Obviously I can't drive it like that, and miraculously it would be ok and I'd be able to drive normally again. That was the first time it did that and kept me from driving. Normally it was just a sputter when pushed but only at random times. Yesterday I got stuck about a block from my house for about 10 minutes while I tried to see if I could get it going. It was acting so bad it took forever to turn around the street because I couldn't really give it any gas, it would sputter every time I tried.

I changed the fuel filter today and checked the pump's filter. I couldn't diagnose the pump because the filter was fine and the car never acted bad while it was in my garage. I played with some wiring and nothing seemed to make the pump act any differently. Plus, I'm not sure it would be the pump, because if it was the car would die pretty quick, not idle ok.

What I'm kinda thinking at this point is it could be some debris in the injectors themselves or maybe a fuel pressure regulator problem? It is really hard to test because the car runs fine most of the time. It just randomly sputters like this but there is no codes, no check engine light, everything seems to be normal, it just sputters. It sometimes doesn't sputter all that long, and then it is back to normal operation. Yesterday's 10 min session was a record.

Is there anything else I could be missing?

[Legacy777]

I know you said no CEL, but have you pulled the codes or done the D-check mode?

I'd suspect an electrical issue rather then mechanical....especially since it's intermittant and goes away. Mechanical problems tend not to go away, but rather get worse.

My WAG would be the MAF sensor.

[free5ty1e]

I second the MAF sensor having issues.

[douglas vincent]

Look for air/vacumn leaks AFTER the MAF.

[georryan]

I will check up on the MAF and play with the DCheck mode. Thanks guys. I just wish it was consistant enough to test. Sometimes it is only for a second or two, other times it takes so long to correct itself that I am dead in the water.

[vrg3]

Everyone's advice is good, but I'll add some:

It could be bad traces on the throttle position sensor. If the signal seems to change when it shouldn't, the ECU might go into power enrichment/enleanment.

You say you are pretty much guaranteed to see the problem happen within 10 minutes of driving? Try this when you have some time: unplug the oxygen sensor, throttle position sensor, knock sensor, and pressure sensor. Leave only the camshaft position sensor, the crankshaft position sensor, and the MAF sensor in place. Then try driving gingerly for a while. You'll get a CEL, and it will cut fuel if you try to go on boost, and it will stumble if you move the throttle pedal too quickly, but other than that (yeah, I know) it should drive fine. If the problem remains, things point more to the MAF sensor or a mechanical problem. If the problem does not appear, connect things one at a time until it does and you'll find the cause.

Can you run my scantool? It would be potentially very valuable to know what some of the readings do when the problem appears. You'll have to wait for the next time the problem happens for more than a few seconds though.

[georryan]

I'm not guarenteed of anything. The times when it is the most problematic is when I'm going slow without much gas on, or just starting to get going from a stop or slow crawl. It will sputter to the point where I can't go but more like coast along.

So far it has never happened at any other time, at least not as bad as that.

[vrg3]

So it could be the TPS traces near the "fully-closed" position.

[georryan]

Could be. What I meant by it not happening any other time was that it hasn't happened as bad any other time. The initial signs I'd see happening, which I wasn't sure if they were related, but I think they might be, is if I'd hit the gas fast, rather than easing onto it a bit I'd sometimes get a slight sputter before it reacted (this would also happen on the freeway when I would try and hit the gas fast while in 5th so that I could pass a car). Also, as I said before, fully floored and boosting would sputter sometimes. I have been noticing that when I drive it just after some of those times when I'm dead in the water, and it is suddenly driving better, it will take off fine, and I'll get the increase in power from the turbo spooling up, but then feel a sudden loss of power as if timing has been retarded for some reason. I haven't been able to get the scantool on it yet. I'd have to have it just sitting there waiting for to act up again. I hope to catch it soon.

[georryan]

Ok well i drove it around today, got warmed up and put on the d-check. It came up with nothing. While drving it in that mode I had some minor sputtring at speed and some sputtering when starting off here and there. No codes from D-check.

I had the scan tool running as well and everything looked normal. I didn't actually look at the airflow signal, I should have. I did look to see if the pump was cutting out here and there and it wasn't. I noticed the throttle position sensor didn't do anthing abnormal.

What does fuel trim represent, and how should it be behaving. It was all over the board. At speed it seemed that when I'd get my sputtering I'd be seeing the trims at 0%, yet when I was starting off, I'd notice the sputtering when the trims were at 7-19%(7 was the lowest I saw, but I mostly saw 14 or 19). Should it be that high when you are just starting to move the car from a stop? Sitting at an idle the trims are at -6 to 3 or something like that.

side note: What are the three sensor connectors directly under the throttle body? One is the knock sensor, what are the other two? A couple of those could probably use some rewiring work.

[vrg3]

Fuel trims should normally stay around +/- 10%, I'd say. 19% is close to the maximum it can get. But it's not clear whether this is a cause or an effect -- misfires throw the fuel trim off. I say disconnect your oxygen sensor until this whole thing is resolved.

The other two connectors under the throttle body are for the cam and crank angle sensors. If those wires or the shielding around them are not in good shape, it could be causing the misfires you're experiencing.

[georryan]

Yeah a couple of wires are in bad shape down there. I think I may swipe part of a wiring harness from a local yard and splice it in. We'll see if it helps.

I'll update soon. Thanks.

[georryan]

Well I'm crossing my fingers, but I think it was the wiring for the cam and crank sensors. They were seriously shot. I had no idea. When I finally got the harness in a position where I could actually get a good look at it, I saw several of the wires were exposed and barely conneted. One of the ground wires snapped as I tried to seperate it from the other two.

So I was able to fix the harness. The knock sensor needs to be looked at I think. I had a check engine light come on after I was done and it was the knock sensor. I started playing with the wire and had a friend in the car and he said the light was going on and off as I played with it. So hopefully that is the only deal.

Thanks guys. I'm hoping this was the problem, but the issue came so randomly that I may not be able to tell for a little bit.

[vrg3]

Ooh. Those signals are very high-impedance so connection quality matters a lot. Yeah, fingers crossed!

[georryan]

<sigh>

Ok, so my knock sensor code keeps being thrown. After I play with the wiring some it will go off. So the other day I pull the harness again to look at the knock sensor wiring. What I found is that a bit down the wire of the knock sensor, just inside the plastic tubing, the wiring gets thicker. It has like a wire inside of a sheeth, with wiring outside that, and then the plastic wrapping for the wire. There is a crack in that section of the knock sensor wire. Is that fixable? I'm almost to the point of wanting to pull that section of the harness out and just replace it with a new section.

Anyways, I didn't know if I should play with that section or not, so I tried to pull it together and electrical taped it for now till I could get an understanding of what is going on there. I don't want to cut and splice there if I'm only going to be causing a bigger problem. I'm guessing the inner wire is the actual wire, and the outer one outside that sheeth is just some sort of interference or shielding like on shielded wire? Anyways, let me know what you guys think. Thanks.

[free5ty1e]

Is it the new knock sensor or the original? If you haven't upgraded to the new one (white connector) then you should do that anyway...

[vrg3]

Are you talking about the wire on the sensor side or on the harness side?

It's better not to try repairing the wire if you can avoid it. As you've found, the single conductor is very well shielded. And like Chris says, if you still have the original type of sensor you should get the redesigned one; its wire is more resistant to cracking.

[georryan]

I'm talking about the wire on the harness side. The sensor side has been replaced already. I got the new version and put it in because my old knock sensor was cracked.

The wire coming from the harness down in the shielded area is cracked. I think that is what is causing my knock sensor code and check engine light to flip on. So are you saying that it is best not to play with that or try to repair it. If that is true I'll have to get the strip of wiring harnes that goes from inside the car and through the firewall.

[free5ty1e]

Ohh... on the harness side. Nevermind then. If that's the case then you might as well try to repair it (and make sure you solder and preferrably heatshrink the finished product!) If you mess it up just keep splicing until either you have to get a wiring harness or you get it buttoned up good enough to satisfy your ECU. Since you are considering getting some harness wiring anyway, couldn't hurt to have it on hand and then try the repairs...

[vrg3]

Oh. Yeah, you gotta try to fix it. Cut away the bad parts of the wire and solder it back together. Insulate the repair with heat shrink tubing and then solder a wire across the two now-separated pieces of shielding.

[georryan]

ok so....I splice in a wire for the center section and use heat shrink tubing. Then I soder the outer wire as well and heat shrink that too?

[vrg3]

Yep!

[vrg3]

Oh yeah, and make the break as small as possible, since it'll be essentially unshielded.

[georryan]

If I'm having this light on, is it bad to be driving it, even if I am driving it real gingerly ( as in staying out of boost as much as possible, and definately not going over 5psi).

[vrg3]

Nope, if the light's on it should be okay as long as you don't get a bad tank of gas or anything.