Idle problems... again!

[mikec]

That's weird that they would come loose. I don't see any problem with a little dab of Loctite.

So how's the idle now?

[vrg3]

Hehehe... So it was my universal bad-idle bogeyman -- a vacuum leak.

The clamps shouldn't back off unless they're not tightened enough... Do you think maybe you're being overprotective of the rubber/plastic/whatever-it-is at the ends of the pipe? Or did you put lube on the threads of the clamp?

I guess Loctite should be okay... Or maybe just a dab of nail polish.

[entirelyturbo]

Do you think maybe you're being overprotective of the rubber/plastic/whatever-it-is at the ends of the pipe?

Yeah, sorta. THAWA said the same thing, that I should put some man in it when I tighten them

I was worried about snapping the clamps in half more than anything else, but screw it, I'll just tighten them with all my big burly manly strength this time

So how's the idle now?

Idle is a lot better now. It still drops a bit low after coming down from 3k, but that's inevitable I guess...

[vrg3]

The factory clamps are actually pretty good... I wouldn't worry about snapping them.

[Legacy777]

I forgot to tighten the clamp at the TB once.....I didn't really notice too much of an issue in how the car ran......or how long it was like that....oops

[entirelyturbo]

Apparently that wasn't it.

It was idling low enough Sunday that it would make the lights on the dash illuminate for a split second intermittently when the ECU thought it was stalled, it was just sitting there bouncing like that pretty much right at 0, then I noticed it and punched the gas real hard to show Patti that I was pissed

Here's what I'm noticing. Let's say I'm doing 40 in 3rd at about 3k, and I just put in the clutch. It will drop to almost 0 without passing Go or collecting $200, it won't even pause at 750rpm. Then it will bounce in between 750rpm and about 400rpm.

Now if I slow down in gear and put the clutch in about 1500rpm, it will just go to 1000rpm and sit there steady for a couple seconds, then drop to 600rpm and stay there, it won't bounce, it will stay steady.

I wanna say this is an IAC problem, but for God's sakes I replaced the IAC valve not 2 years ago!

[vrg3]

Hmm.

Well, the material difference between the declutch-at-3000-rpm and declutch-at-1500-rpm situations is injector cutoff. When the engine's coasting at higher RPMs, the ECU cuts off the injectors and lets the engine just coast on its own. When engine speed drops sufficiently, it starts pulsing the injectors again to make sure it doesn't stall out.

So, hmm. Somehow your engine is getting to the stalling-out point before the ECU can start fueling appropriately again. Right? Maybe?

Complicating the matter is the closed-loop idle control. That's probably what causes it to bounce after recovering from the almost-stall. The ECU ends up cycling around trying to achieve equilibrium. I'd guess if you left it long enough it would quit bouncing around.

So what could make engine speed drop faster than normal when coasting? It seems nonsensical since the injectors aren't operating anyway, but is it possible that Patti just needs a tuneup?

[entirelyturbo]

Spark plugs and fuel filter have maybe 5k miles on em, wires are a tad older but I don't think they're bad, coilpack is only as old as the wires, what else is there?

Maybe I should try running some of that Lubro-Moly FI cleaner again?

[vrg3]

Couldn't hurt, I guess. Do you run that stuff through the intake manifold?

Maybe you could try dumping some Sea Foam or Techron injector cleaner in the fuel tank. It could be an injector that's partially clogged.

[entirelyturbo]

You run Lubro-Moly through the tank like most other FI cleaners. That's the stuff I used in 3 consecutive tanks to get rid of the hesitation...

By any stretch of the imagination, could the O2 sensor be involved in this? I'm still on my original at 139k and while my mileage is good and I've never seen one code for it, I wonder if it's performing as it should...

[vrg3]

Ah. Well, sounds like that stuff couldn't hurt anything.

Yeah, maybe it could have to do with the oxygen sensor. If it's gotten really lazy then the lean reading from coasting could last through into when it starts to idle with closed-loop fueling, causing a temporary overrich condition.

Try unplugging the oxygen sensor and resetting the ECU maybe? Then drive around and see if the problem goes away when the ECU has no oxygen sensor?

[entirelyturbo]

Okay, now that I have a new alternator putting the proper charge to the battery and putting a normal load on its drivebelt, my idle is absolutely terrible.

I mean, if I don't actually heel-toe and keep on the gas a little bit while I'm braking to stop, the engine will go below 500 rpm and will try to stall. This doesn't just happen occasionally now, this happens ALL the time.

I put a used throttle body on sometime ago, because I wanted one that still had its original factory settings (I screwed my original one all up). The throttle body has the proper amount of black goo around the plate, the TPS is at factory settings and seems to be working fine, has never thrown a code, and the car runs simply perfectly when it's ABOVE 1000rpm.

I don't know what the problem is, but I'm really getting sick of fixing shit. I just put a $120 alternator in it, not 2 years after I already put another $120 alternator in it.

[Brat4by4]

I will tell you how my car fixed its idle (it used to idle high). Take dash out, leave it for 6 weeks or so, reinstall.

You can probably skip the taking the dash part out. Maybe go under your knee panel and disconnect all the main harness connections. Pull the battery and let it sit for a week. Reconnect, start up the car and turn on a lot of accessories (lights, fans, compressor, radio) and watch the car learn to idle.

[Legacy777]

have you reset the ECU?

[entirelyturbo]

I've reset it a couple times, but not since I put the new alternator in.

It's acting very much like a vacuum leak, but I don't know where it could be, as everything on the intake hose is snug and tight. It could theoretically be the intake manifold gaskets, but I hope to God that's not what it is.

If I do a vacuum test and the vacuum is lower than normal, that indicates a vacuum leak somewhere right? And by testing it in several different locations, I can isolate what the leak could be, right? If I test it at the PCV valve and it's good, and then I test it at the IAC hose, then it would have to be in between the two right?

[vrg3]

Reset the ECU now that you've fixed the fluctuating voltage problem. That might be all you need.

[Legacy777]

[starsky]DO Itt! Just Do it![/starsky]

[entirelyturbo]

http://bbs.legacycentral.org/viewtopic. ... 9776#99776

There is hope!!

[entirelyturbo]

I'm sick of resurrecting this thread

I still have an erratic idle, it has stalled on me once since my new alternator, and has tried very hard a couple of other times. Most of the time though, it still dips to 500rpm and will go back up occasionally, and still bounces up and down sometimes.

But...

Subyluvr Bad Idle Hypothesis #465,134: Charcoal Canister

We know the EVAP system plays a very small role in the idle performance of these cars. Evidence: those who have a purge valve stuck closed frequently note a poor idle as a symptom.

I dunno the previous owner of my car, so I can't rule out that he could possibly have been in a habit of topping off the tank after the pump has already stopped. I have never done it, but I do recall a certain event where the pump I was filling from was malfunctioning and I didn't know it until gasoline started pouring out of the side of the car

Such antics are thought to ruin the charcoal canister, so since the charcoal canister is part of the EVAP system, could I therefore surmise that my charcoal canister is possibly ruined and maybe causing my idle problems? Is that a proper use of logic?

Please add your useful comments

[vrg3]

What?! Still???

I'm not sure if the canister can cause a bad idle if other things aren't bad as well.. The purge line isn't supposed to be open during idle. There are two things that ensure this: 1) the ECU doesn't open the purge solenoid when the throttle's shut, and 2) the vacuum fitting on the throttle body that's used for purge is ported so that it's upstream of the throttle plate when the throttle is shut.

Hmm, I guess #2 doesn't necessarily preclude the possibility of the fumes getting into the engine.

Maybe you could temporarily try removing the small hose from the canister and plugging it.

[entirelyturbo]

So going by what you said, these claims that people make that a stuck purge valve affecting idle is basically BS.

I'm still wondering about a vacuum leak. What I said here:

If I do a vacuum test and the vacuum is lower than normal, that indicates a vacuum leak somewhere right? And by testing it in several different locations, I can isolate what the leak could be, right? If I test it at the PCV valve and it's good, and then I test it at the IAC hose and it's too low, then it would have to be in between the two right?

Is that true? If there is a vacuum leak, can I find it this way?

[vrg3]

Well, it is possible in principle that the venturi effect could allow the vapors to upset idle mixtures still.

By "vacuum test" you mean just hooking up a vacuum gauge and measuring manifold vacuum, right?

Manifold vacuum is pretty constant throughout the manifold, so you probably won't be able to localize a leak by measuring vacuum at different ports and stuff. You're better off doing the starting fluid test, in my opinion.

[entirelyturbo]

I've never heard of the starting fluid test, can you kindly elaborate?

I'll play with the charcoal canister this weekend and see if I notice anything.

[vrg3]

Go buy some automotive starting fluid. It comes in an aerosol can and is generally used to help start an engine that is hard to start because of insufficient fuel vaporization.

Starting fluid is basically ether or some kind of very lightweight hydrocarbon. In either case, what it boils down to (pun not intended) is a highly volatile combustible fluid.

Warm the engine up and run it at idle. Now, pick a spot where you think there might be a vacuum leak (somewhere at the base of the intake manifold, near a vacuum hose, at the brake booster, et cetera). Give that area a quick spray of starting fluid.

If there's a leak, two things will happen: 1) The vacuum will suck some of the starting fluid in, and it will be ingested by the engine. 2) The starting fluid will kind of soak the area surrounding the leak and temporarily seal it somewhat.

So if you happen to spray at a leak, you'll hear the engine speed change (increase, normally). It might take a second from when you spray for the change to happen, and the change will probably only last a second or so.

Wait several seconds between sprays to make sure all the fluid evaporates. And don't inhale too much of the stuff; it's generally carcinogenic or at least toxic. It's probably not very good for rubber hoses in the long run either, so don't go overboard.

The change in engine speed can be very subtle if the leak is small or upstream of the throttle plate. It might help if you disconnect the oxygen sensor and knock sensor beforehand. Disconnect and cap the purge fitting on the throttle body. That'll help keep the idle as steady as possible so you can notice the change more easily. You also might try one test with the IAC valve unplugged (it shouldn't be necessary to maintain a warm idle), but do a test with it functioning as well so you can test its plumbing.

[entirelyturbo]

Cool deal, thanks Vikash