erratic idle when cold

[Michael503]

I have a 94 Camry LE with the 4 cyl engine and about 180,000, and it has recently developed an annoying problem. Initially, my wife became aware of the issue when she started the car and drove about a mile and stopped at a light. She said the tach started jumping and the car tried to go dead.
Now it does it whenever its started, and the engine is cool.

When started, the idle jumps up to around 2000 and then procedes to dance between 1500 and 2300 on the tach, quite quickly and erratically. The engine idles roughly, but not so abruptly as to indicate a complete ignition malfunction. Putting it into gear does not make it go away, but giving it a small amount of gas does seem to make it get better. It quits completely when the idle in neutral drops below 1000 as the car warms up. In 60F weather, thats about 4-8 blocks. It seems perfect once that happens

My first thought was plug wires since this started about the time things get humid and sweaty around here. They might have been bad, but I had to replace them anyway because one of them tore when I pulled it off of the plug. I inspected the distributor and found it full of oil and the coil I installed 5 years ago was cracked, so I replaced the distributor as well. I also have a spare ignitor, so I swapped it too.

In summary, new distributor, plugs and wires, known good ignitor... no difference.

Inspection of the timing indicates its jumping around quite a bit, maybe 15 degrees while this is happening.

Further inspection reveaded a cracked vacuum line going to an egr related component and a clogged PCV valve... again (I swear this thing makes more sludge than smog) I repaired the line and replaced the PCV valve with the spare that I keep in a jar of solvent (I have 2, just for this reason) and it of course, made no difference.

I think that if the idle air control was malfunctioning, that I'd see a low idle and it would be consistent, even when hot. Likewise with an EGR leak.
I have scanned this forum for hours and the only malfunction description I can find close to mine involves an ECT sensor.

I am completely open to any suggestion... Thoughts?

[BamZipPow]

Did any problems present itself before this? Was this car worked on before this happened?

What did the old plugs look like?

How does the throttle body look? Have you checked fer vacuum leaks?

Did you check the TPS (throttle position sensor)?

[Michael503]

old plugs where white and crusty, prol 100K on them.

I do all the work on this car, last repair was brake related. It has had a consistent issue with sludge, due to the driver and her short route to work. The TB is not sticky, but carboned a bit. As a matter of course, I hit it with TB cleaner and a soft rag.

I have not sprayed the engine down to look for leaks yet, I ran out of daylight. I did find one, but plugged it. no change

I have not done a resistance check on the TPS. I have never seen a TPS get progressively better as the car warms.

A buddy of mine suggested the O2 sensor, its never been replaced, but I'm skeptical. He also suggested the ECT sensor. Its my next victim, after I see if there's a code in the ECM

[Stillrunning]

Erratic idle that improves with gas pedal leads me to leaking EGR, sticking IAC or vacuum leak. All three are easy DIY jobs. When was the last time you cleaned the EGR valve? You may also still have a vacuum leak. Cleaning the IAC just to be sure is also easy. As for the ECT sensor it usually causes the engine to run rich when warm and so would cause bad idle/stalling then, not when cold. If it were running lean when cold you'd get hard start/stalling/no start and no improvement with gas pedal. Did you also clean the PCV hose?

[hill8570]

Originally Posted by Michael503

A buddy of mine suggested the O2 sensor, its never been replaced, but I'm skeptical. He also suggested the ECT sensor. Its my next victim, after I see if there's a code in the ECM

I really doubt a duff O2 sensor would cause those sorts of behaviors. ECT reading too "hot" sounds likely -- certainly worth testing.

BTW, when you changed out your dizzy, did you replace the o-ring on the dizzy shaft? Without that, you're liable to have that dizzy fill up with oil again in short order.

BTW, are you running dino or synth oil? I've found that synth is a lot less prone to sludging in these beasts, although it will ease it's way past the old oil seals more than the dino.

[Michael503]

EGR never cleaned.

IAC never cleaned.

PCV hose is clear

Why would these components only misbehave when cold? A sticking IAC should stay stuck and a leaking EGR should leak when hot. Nevertheless, they will be inspected as a matter of course. Thanks for the tip!

I prol need to have a better go at a vacuum leak that might only appear under cold running conditions.

I am rapidly falling out of love with this car. We've had it for 12 years and it may be time for it to go to a new home. Seems like its always somethin'

----------------------

I can see an intermittent open in a ECT causing a rapid fluctuation in the idle control system, but then again, i could see bad spark doing the same thing.

The dist came as an assy. The seal on the inside is the one that craps out and apparently, isn't available separately and thats why I replaced it as a unit.

I have read lots of raves about synth oil, and I take it with a grain of salt. I run petrol. I know synth can have a hard time staying in an engine, and will pretty much pass right thru a cork gasket.

[Michael503]

Well, I went out this morning and retrieved codes... DTC 13 and 14, cleared them and they came back. These may be codes that stay, I can't remember.

Anyway, I unplugged the ECT sensor and looked at the connectors... eh, they looked like 16 year old connectors. So I started the engine. Poof, starts normally and idles normally,at around 1000-1200 when first started (60F). For yucks and grins, I unplugged the ECT while it was idling and the idle dropped, but smoothly.

All this time I thought I owned a camry, but it looks like I have a gremlin too.

Time will tell, I suppose. I can't get at the EGR or IAC til Sunday, due to work and weather.
I'll post here.

[hill8570]

Originally Posted by Michael503

EGR never cleaned.

IAC never cleaned.

PCV hose is clear

Why would these components only misbehave when cold? A sticking IAC should stay stuck and a leaking EGR should leak when hot. Nevertheless, they will be inspected as a matter of course. Thanks for the tip!

The buildup in the IAC and EGR are both due to crankcase vapors (courtesy of the PCV valve) condensing in the throttle body area. It tends to be somewhat sticky / tarry, and like most oil products it gets less viscous / sticky when hot. Thus you can have valves that work perfectly fine when hot, and do strange things when cold. This is more likely with the IAC, since the hot recirculated gas that goes though the EGR tends to turn the build-up into more of a hard carbon.

Originally Posted by Michael503

I prol need to have a better go at a vacuum leak that might only appear under cold running conditions.

That's a long shot, but you never know.

Originally Posted by Michael503

I am rapidly falling out of love with this car. We've had it for 12 years and it may be time for it to go to a new home. Seems like its always somethin'

I know where you're coming from...my rig is 15 years old, and if I didn't do most of my own work (and generally enjoy it), it wouldn't be cost-effective to hang onto this rig.

Originally Posted by Michael503

I can see an intermittent open in a ECT causing a rapid fluctuation in the idle control system, but then again, i could see bad spark doing the same thing.

Interrmittent open on the ECT usually isn't temperature-dependent. I had a loose wire going to my ECT recently, and it caused funky idle and nasty pickup, but didn't matter if cold or hot.

Originally Posted by Michael503

The dist came as an assy. The seal on the inside is the one that craps out and apparently, isn't available separately and thats why I replaced it as a unit.

Yeah, there are two seals on that dizzy -- one internal, and an o-ring on the stem. The internal one is mostly to keep oil out of the dizzy, the o-ring is mostly to keep oil off the outside of your engine.

Originally Posted by Michael503

I have read lots of raves about synth oil, and I take it with a grain of salt. I run petrol. I know synth can have a hard time staying in an engine, and will pretty much pass right thru a cork gasket.

Yeah, synth gets over-hyped, especially the "better gas mileage". It is less prone to breakdown, 'tho, which is good from a sludging perspective. The trade-off is the increased leaking past the oil seals. After spending an enjoyable couple of days cleaning all the crap under the valve cover and in the pan when I "inherited" my '95 from my wife, I switched to synth and I'm slowly working my way through changing the oil seals. You pays your money, you takes your choice.

[hill8570]

Originally Posted by Michael503

Well, I went out this morning and retrieved codes... DTC 13 and 14, cleared them and they came back. These may be codes that stay, I can't remember.

"cleared them" as in pulled the EFI fuse? That should clear them out. If they're coming back, it's worth tracking down the cause.

DTC 13 is "NE Signal Circuit (No.2)" which basically means the ECU isn't getting the crankshaft / camshaft signals from the dizzy.

DTC 14 is "Ignition Signal Circuit" which means the ECU isn't getting the IGF signal from the igniter.

That would explain the jumping tach -- the tach is based off the same IGF signal.

AFAIK, the only common points between those two error messages are the ECU (which usually doesn't go bad), and the distributor, although it seems likely that the signals both route to the ECU through the same cable bundle, so off chance something has damaged that cable bundle. The connector from the dizzy to the igniter is a different one from the connector from the dizzy to the ECU. The connector at the ECU side looks to be the same one, 'tho.

I'd check / wiggle test the connectors and wires going into the dizzy. If it's an aftermarket dizzy (which I assume it is, since OEM dizzys are pretty spendy), I'd maybe try a junkyard pull to see if things change. If none of that works , physically trace the cable running back to the ECU to look for damage, and maybe check the connector at the ECU (ugh!) for looseness. Dollars to donuts, 'tho, it's a flaky distributor.

I'm somewhat surprised this car is running at all...the ECU failsafe for DTC14 is to cut fuel. Must be really intermittent to be running at all.

Also, is there any chance you're running a California dizzy in a Federal car, or vice-versa? I'd be surprised if that would work at all, but the wiring diagram between the two doesn't look glaringly different, so I suppose there's an off chance it would work, but badly. Certainly far from my first guess, 'tho.

[Michael503]

I should mention that I only drive this car when there's an issue with it, so I'm not familiar with its normal behavior when cold.

The dizzy is new. it was replaced by me as a possible cure for this issue. It comes with all the coils and is a non California dist, as is the car. it is a reman, but all of the coils look new. I had to replace it anyway because it was full of oil (seal failure).

I cleared the codes this morning by turning the ignition off, and unplugging the 30A fuse. The 2 pins were shorted on the datalink connector at the time. The codes appear to be clearing cuz when I unplugged the ECT while running and it defaulted to warm, it kicked out another code. This code cleared later, but the other 2 did not.

About 5 years ago, there was a non starting issue that turned out to be the coil. (weak spark) I can't recall what codes there were, but they would not clear then either. I seem to recall being told they were normal.

using a timing light, I could see the timing jumping a good bit when it was acting up, but the timing light was consistent, meaning the spark never went away. when idling normally (not acting up), with the link shorted to set the timing, the timing fluctuates about 2-3 degrees, well within what I've seen in the past on other vehicles.
In retrospect, I've never seen an ignition misfire cause a high idle.


Its possible the timing jumping could be due to severe variations in engine speed commanded by the ECU controlling the IAC, or an intake leak via the EGR. Logically, it can't idle high without air.

Another remote possibility is the ECT(again). My wifes description of the problem was the tach went nuts and she smelled exhaust in the cabin. If the connector to the ECT was corroded (it didn't look very clean), it could possibly produce intermittent high resistance, but not an open circuit, telling the ecu to idle up and richen up. The only thing I did different this morning was to clear the codes and disconnect/reconnect the ECT before starting, and it behaved perfectly.

If its not raining, Ill try it again tonight or in the morning. If I can get it to act up again, I'll unplug the ECT and settle it one way or another.

Thanks for all the advice. I have a Sentra that I drive and on the nissanforums, I get almost no help at all. You guys are great.
--------------------------------------------------------------

EDIT: Well I got home tonight and after some inspection, I found that in my hurry this morning, I had been removing the wrong fuse to clear the codes After removing the correct one, the codes cleared. They probably occurred when I was testing the vehicle. At one point, I swapped ignitors and did not bother to bolt the new one up. The car would not start because the ignitor is case grounded... go figure.

The car started and warmed up fine, smooth as glass. I wiggled wires and got nothing, no hiccups. It also occurred to me the idle timing fluctuation, may have been caused by slack in the link between the distributor shaft and the cam driving it and was a byproduct rather than a cause.

I plan to reassemble it in the morning and try it some more. I'll keep this post up to date.

[Michael503]

Well it skipped a bit last night, so I pulled the ECT connector. It idled lower, but did not seem to clear up immediately. it only ran rough for about 15 seconds total, that included the time it took me to get out of the car and get under the hood.

I pulled the fuse to clear the codes and went to bed.


This morning , I got up after a sleepless night (the daylight change didn't help) and on a hunch, pulled the upper timing cover to inspect the belt, which I realized was a long shot. It was loose, but not abnormally so. I tensioned it and proceed to recheck the timing, but I got a surprise, something I'd never seen as 20 years as a motorcycle mechanic.

I turned the key to spin the engine, which would not start and when I released the key, the tach jumped to 3000rpm and the distributor made a familiar buzzing noise, like a coil buzzing as it sparked.

Did I mention that the engine was not turning?

I shut it off and did it again... same thing

I checked for codes, and got one 11111 11. the factory manual does not list this code, but my chiltons does... bad knock sensor, or bad ECM, but I think that's for a different engine. The factory manual lists 11111 1 as the knock sensor code. Must be a misprint

Since the car was obviously sparking and not running, I can only assume the ECM has gone south.

I cleared the code and tried again. This time it started, but the engine was flooded. After it cleared, it idled and ran fine.

If anyone has any more info about this phenomenon, I'm all ears, otherwise its off to the salvage yard tomorrow to try to find a good ECM

[Michael503]

Well, I installed the used ECU and BAM! problem solved, I've started the car about a dozen times, warmed it up drove it around, etc. Runs flawlessly! I'm 99% sure this part fixed it, but I will post back if it acts up again.

While I was in car mode, I replaced the timing belt, front oil seals and valve cover gasket. I also sprayed silicone oil on the suspension bushings and wiped the door gaskets with silicone. Now not only does it run well again, it doesn't creak either!

Along the way, I fell in love with the car again... I may have it painted

[denisesballs]

Originally Posted by Michael503

Well, I installed the used ECU and BAM! problem solved, I've started the car about a dozen times, warmed it up drove it around, etc. Runs flawlessly! I'm 99% sure this part fixed it, but I will post back if it acts up again.

While I was in car mode, I replaced the timing belt, front oil seals and valve cover gasket. I also sprayed silicone oil on the suspension bushings and wiped the door gaskets with silicone. Now not only does it run well again, it doesn't creak either!

Along the way, I fell in love with the car again... I may have it painted

Thanks for posting this. I recently bought a used '96 Camry and am experiencing similar symptoms. Funny thing is the guy I bought it from said his father (a Toyota mechanic) told him it was the ECM, but my father in law (a general mechanic) said it was more likely the idle air control or dirty throttle body (a well known issue on these cars which is all over these forums). I gave the throttle body a quick clean with some carb cleaner and sprayed some into the IAC hole and it hasn't really seemed to fix it. I bought some Seafoam Deep Creep and am gonna give it a more thorough cleaning, but if that doesn't help it's good to know the ECM can in fact be the issue like the seller's father said.

[Michael503]

Its true, ECM's hardly ever go bad, and since you can't test one except by replacement, you need to rule out everything else first.

Put a timing light on it., its actually a good diagnostic tool. weak spark and/or a fouled plug will cause the timing light to not flash, or flash intermittently. While your at it, check the timing, it should be steady within 1-3 degrees.
If the spark is strong on all cylinders and the timing is solid, then its probably fuel/air. Unfortunately, both are controlled by the ECM


I should add that I took the liberty of examining my ECM and found that the big capacitor right in the middle of the board took a crap. It wasn't that obvious at first, but it leaked onto the main board. I may replace it just to see if its still good.