My car idles weird. I have a 1991 Toyota Mr2 (non turbo), and when i start her in the morning, when its cold out, i wait for the engine to heat up a bit and then i take off. When i first turn the engine on, it runs at about 2000 RPM, and slowly comes down as it heats up. But lately, it hasn't been coming down past 1500 RPM. The engine will have been running alot, i'll be at a red light, and it'll be idling at 1500 and even 2000 RPM. It rarely comes down below 1100.

I can't understand it. Isn't there a chip that moderates that? Could that be messed up?

Hmm... '91 is pre-OBDII, so the engine management system isn't regulated by any standard. Hence we don't know for sure what is and what isn't monitored or controlled by any ECU that might exist.

But whatever, 7-800 rpm is a typical hot idle speed, 1500+ is definately wrong. Question, does it idle irratically? If it does, then an airhose leak might be to blame. Most cars have crankcase breather hoses that feed crankcase fumes back into the engine so they get burned rather than emitted as pollution, and a leak here could be an issue.
The other distinct possibility is that the choke is never going off - you'd know this because although the idle is high, it is fairly consistent, and the car probably smells a bit of gas. Now if your car is ECU controlled, then I'd expect that this could be an ECU fault, or a temperature gauge fault. (Does your temperature gauge work??)
Oh, and check your radiator fluid level too. (When cold obviously, unless you enjoy getting scalded). If the fluid is too low then it may not be circulating properly, resulting in a hotter than normal engine - which would cause a higher idle. (But in this case, it probably wouldn't be erratic, as the choke would also still be active for the same reason).

The three most likely causes of this in my experience are (in order):

1) Low coolant level

2) Faulty water pump

3) faulty thermal sensor to ECU

My civic suffered from (1), ultimately caused by (2). Same symptoms.

pca

There is also an idle air bypass on most of these types of cars. It is located somewhere on your intake manifold, generally close to the throttle body. Most of these are thermostatically operated by the coolant. As the coolant heats up it causes a diaphram to close, thus, causing the idle to decrease. You might want to check this for proper operation. Good Luck.

What exactly constitutes consistency? Because half the time it idles at just under 2000 RPM but it's also known to stay steady at 1500 rpm. I consider it pretty consistent, however, i don't smell any gas.

The car does have an ecu .

I'll check my coolant level and check back.

do you have a 'check engine' light...and is it on?
you may have a vacuum leak, faulty AIC motor (or throttle switch).
pop the hood and listen for a sucking noise. (i love that).
an engine temp sender (for the injection) could be at fault, or as noted above, a low coolant condition will act the same. the system never knows that the coolant got hot, and keeps your idle high as if it was still cold.
if your CE light is on, i can tell you how to pull the codes.
a vacuum leak is the most common cause.
hth

Wait a minute, what BBS did I sign onto?

No, URL is right, and the logo doesn't show Tom and Ray...

Ahh... short term variations. If the engine is running lean, such as it would with a fuel pump problem, or air intake leak, then the idle wont sit at a nice eg 1000rpms from one second to the next. It will tend to flutter and surge between eg 950-1050 (or even more variation), over a period of a few seconds. Note that variation of 10-20rpms over such a period should be considered normal.

But if I'm hearing you right, sometimes it's steady at 1500rpm, and other times it's steady at eg 1800rpm. Not that it's fluctuating between the 2 within a short time. That would seem to indicate that it's not a lean engine problem, but rather that it's something more intermittent. You've already indicated that you don't smell any gas, and since you've confirmed the presence of an ECU, we can reasonably eliminate 'stuck choke' as a cause - that should all be handled by the ECU.

Time to check your cooling system... Patrick had a very good point about the water pump that I had forgotten about - and has reminded me that exactly the same thing had happened to my old golf when it was about 12 years old - the water pump went, resulting in random overheats and water loss problems... and yes, an increased idle is a symptom. I don't know how I forgot that.

Sounds like the ECU is holding in it's cold start strategy. The comments here about coolant and the coolant sensor are what I was going to suggest, especially if your weather has suddenly turned cold - temp sensor may well have been exposed in the cooling system by a low coolant level which would fool the ECU. You will be burning excessive fuel while this goes on, so don't leave it too long to sort out. One other point of note is that if it keeps holding the engine at high revs while still relatively cold, you may be incurring engine wear. If it had been a carb car, there would also have been the issue of a rich mixture dumping petrol on the bores when cold and washing off the lubricant. Believe me, this is a good way to kill an engine in short order.

You should be able to use the Gunson's Home Diagnosis unit to work out what's up: even if it's not ODB-2, it may have the precursor system, or a proprietary Toyota diagnostic port on the ECU. What's wrong with running it into a Toyota dealer? It would take them only a few minutes to do the diagnostic and this should be relatively cheap.

Good points.

You know, your post did bring to my attention that i noticed this once it started getting cold around here. Is the coolant sensor someting that the shop checks out?

I actually called the guy I take my car to and he said it could be the idle regulator? I thought this was handled by the ECU?
Too bad he didn't have more time to clarify.

An idle regulator is one of the components you would find in something like a Bosch mechanical or mechano/electronic injection systems such as L or K Jetronic. It is an independent subsystem, usually controlled by it's own sensor, which works during cold starts. If you had said you had an early Porsche 924/44 then I'd have said the same thing. The Toyota uses a generic all electronic system developed by one of Suzuki's/Toyota's suppliers. It is non-OBDC, but still has a diagnoistic port. If your mechanic doesn't know this, I suspect it's time to change mechanics

A diagnostic test (depending on the vintage of the ECU) uses one of two possible test strategies:

- switches out the ECU and makes individual tests of all the system sensors (resistivity, conductivity, etc.)
- assumes the ECU does this testing itself and reads out any associated fault codes associated with a sensor test failure.

I would have thought a '91 non-ODBC ECU from Toyota would be the first case (I may be wrong). If so, a garage test will test out the temperature sensor directly. They should then be able to tell you what's up.

I would do two things in your shoes, maybe with the help of a spanner-savvy friend if you are not sure yourself:

- check the coolant level in the header tank, and also the colour. If it is low, top it up with clean anti-freeze rather than just water. If there is evidence the coolant is "muddy" (corrosion) or has an oily sheen (burst head gasket causing oil contamination) then you need to have it looked at. The first case is simple: have the engine flushed out and the coolant completely replaced with fresh. You should be doing this once every 1-2 years in an alloy block engine like yours anyway. In the second, then you really need to get a compression check done and have the head gasket replaced if necessary. A telltale sign of this condition is that you get a lot of steam out of the exhaust during startup and for quite a while afterwards. Don't leave this too long, as it will result in a dead engine.

- if the coolant level and colour looks OK, locate the temperature sensor for the engine. It should be near the thermostat towards the top of the engine, and there will be a lead running to it. Remove it and check if it is coated with residue. Prior to removal, also look around the body casing of the sensor to see if there is any corrosion or anything indicating the sensor casing has failed and it is leaking coolant. Replace it after cleaning and see if it makes a difference.

Either of these is pretty simple and easy to check before spending a lot of money at a garage. Don't be scared of it!

My current bet is on low coolant, with corrosion fouling of the sensor caused by exposure to steam above the coolant.