O1e

Hi!

My 97 STi v3 has a bit bad idle when hot. The engine is just fine when cold. Both driving, idling and when starting. But as fast as the coolant temperatur start rising, it will get a bit bad on idle.

When I stop the rpm drops down, and makes the engine almost die, and then rise to around 1000rpm and slowly falls down to around 800rpm.

This is just when the engine is warm/hot. The car seems to run OK on WOT.

- The MAF is almost new.
- I have cleaned the IACV and tried another used IACV that I also cleaned.
- I have cleaned the throttlebody.
- New TPS, and set it to 0,5V off throttle and 4,3V with full throttle.

Still getting fault code 24 - Idle air controll valve.

But could a bad Coolant Temperature Sensor cause the CEL on the IACV? And could it be this sensor that are the fault instead of the IACV? Because it work great when it is cold, and start to get worse as the coolant temperature rise.

Hope you guys have time to answer me.

Thanks

scoobiewrx555

Air leak possibly that only makes itself known when everything is hot and expanded. If the coolant temp sensor was faulty normally your car will idle and run rough when cold until it warms up. I don't think it's that. IACV error will only occur when the IACV is faulty. Have you tried to reset the ECU??

O1e

Thanks for the reply

If it was an air leak, it would not run good at WOT? It starts as fast as the coolant temperature start to rise.

I have reset the ECU 2-3 times.

I have also looked after leaks on the vacuumhoses and on the hoses on the intercooler. Tried with "Engine Start spray/gas" to try detect leaks. It use to come up with fault code on the IACV when driving also, and when idleing the CEL turn off again.

If you see here, you could see that the coolant sensor signal are working with other sensors to make the IACV justify itself. So if the coolant sensor give a bad signal, it will make the IACV jusify itself wrong and maybe give an fault code because it can't justify because of the wrong signal from the coolant sensor? Just a tought..

And since I have tried two different IACV's and they are just the same. And with an airleak the rpm would drop and rise all the time on idle?

Here you see the image on the function of the IACV:

scoobiewrx555

Granted it's unlikely to be a leak then if the issue starts as soon as the temp gauge starts to move.

A faulty coolant temp sensor normally affects idle when cold and can even give startup problems, but as soon as the engine warms up it runs fine. I don't see how the coolant temp sensor will cause an IACV error code. An IACV error wouldn't cause your car to run rough when warmed up. All it will affect at any temp is the idle under different idle load conditions like normal load, electrical demand load or air conditioning load.

You have a different problem. I would look at the MAF and Lambda sensors. Both MAF and Lambda sensors can be faulty and not show up an error code.

O1e

MAF is almost new, and same with the lamda's. And if these where faulty the engine almost never would have ran clean og good at WOT. Same with a big leak.

The car run all good. It is just the idle when I press down the clutch and stop, that the rpm falls down too low, then it goes up and stay. Not always, but most of the time, when i stop. If I give it some throttle it will fall right down to ~800rpm, without dropping to low. It is just when I have driven for a while and stop.

And after starting the engine again, when it is hot, it run all good at idle.

scoobiewrx555

You reckon you've changed the IACV for another one and it's still doing the same, in which case

After ECU reset do you know if it is learning full timing again? Do you check to see if the IAM is back to 16 or perhaps after reset put the engine under load to raise the IAM purposely?

O1e

I bought a used one from Japperformanceparts. I have cleaned both, and the idle is just the same. Buying used sensors is not smart at all, I know. But the IACV is not free either. And I don't want to buy a new one for 250gbp and the idle stays the same.

How do I know it is learning full timing again, and what is IAM?

When I had reset the ECU, I have used to disconnect the battery for a while, then check the memory that the fault code are gone. Then start it up and let idle a bit then drive with it.. Some normal driving and some small drag with WOT.

Thanks for the reply's mate

scoobiewrx555

IAM is called Initial Advanced Multiplier. When you reset the ECU it goes to an IAM of 8. That means it runs 8 16ths of the base ignition timing map.

As you drive the car around and load the engine bit by bit the IAM rises again and again until you get to a full IAM of 16. When it hits 16 you are getting the full 16 16ths of the base ignition timing map. There's no way to know the ECU has learned full timing unless you know what you are doing and can feel it, certainly when you know your own car, or if you plug a laptop into the comms port hanging down under the dashboard somwhere above the brake pedal.

If the ECU isn't at full IAM it may well affect the idle. That can be the case on newage 2Ltr ECU's, specially WRX's i've noticed, more so than STi's. There is a table in the mapping for Idle timing and that may be affected when the ECU is at low IAM, although i'm not 100% sure.

I'm stabbing in the dark here because your issue seems an odd one when you say there are no air leaks, it drives ok and only runs rough idle when hot.

I can't think what else it might be. Sorry

joz8968

iTrader: (13)

Coolant temp sensor giving out incorrect voltage i.e. mimicking 'cold idle' voltage... thus telling the ECU to overfuel even when up to temp... thus bad idle???

scoobiewrx555

I've not heard of a coolant temp sensor giving a bad warm idle, only bad cold idle and running. When they go wrong the ouput causes the ECU to default to normal engine temp running (Closed loop idle control - Lean) so you can drive the car normally when it's warmed up and it will idle fine too.

If the lambda sensor on a newage car goes down you can map it to run Open loop only in which case it takes idle and drive fuelling straight from the Primary Open loop fuelling map. Doesn't always run the best but idling is fine and it does keep you mobile. I did that a while back when a Lambda sensor went down and it ran fine.

It does sound like the OP's prob is with the IACV but like he says, he's not going to try it when it's so damn expensive to replace but then sometimes you have no choice.

To the OP, try Grade A Subaru on monday and ask them for a price on a used IACV. see what they say.

Ask for Andy. http://www.gradea-subaru.co.uk/4.html
Tel: +44 (0) 1663 744 114

scoobiewrx555

Just as an afterthough....Maybe TPS (Throttle Position Sensor) is on the blink. If it has an internittent fault it might be returning the wrong voltage to the ECU on lifting off hence idle going up and down even when rolling, speed sensor on the blink could cause bad idle, again when rolling or stationary, but it's only happening when warmed up.

What have you got to lose by changing the coolant temp sensor, not very expensive. I'll eat my dog if it's that

O1e

Hi again guys!

As I wrote before. I bought a used one from Japperformanceparts, and have cleaned both mine and the one I bought. I have tried to switch just the electricalmotor at the top, and also tried to change whole the IACV. I have measured both the electricalmotors and they both have the right value.

I have as I said changed the TPS, and set the value to 0,5V off throttle and 4,3V at full throttle. I could see that the voltage seems to drop to around 0,46V when the engine is running. But the workshop manual says that it should be adjusted with just the ignition on.

A completely faulty coolant sensor will also start the radiatorfan and give an fault code. But It could be on it's way to be totally defect, but not right now.

And the engine seems to run a bit rich, because of aftertenning in the exhaust sometimes on normal driving. Should have tried to borrowed a WBL (Wide band lamda) and see the air/fuel ratio.

Thanks for the replys again

scoobiewrx555

Coolant temp sensor will be cheap to change so you have nothing to lose by doing that.

O1e

Just ordered one from japanesecarimports, so we will see when it arrives

scoobyparts2010

if you need any parts pm me breaking all the time

O1e

I have now replaced the old coolant sensor with a new one. The idle seems to be better, as the RPM do not go as fast down as before. When stopping it stay at around 900rpm nicely, but after 2-3sec it falls down to around 500-600rpm. Still getting fault code on the IACV when driving normally.

Seems like I just have to order a new IACV afterall?

joz8968

iTrader: (13)

Yeah I reckon with the new ISCV added to your new CTS, it'll be back to running tickety-boo. Let us know how you get on...

O1e

Hi again!

Got the new IACV, resetted the ECU by disconnecting the battery and it still giving me fault code 24.

The idle was a bit low so I adjusted it a bit to get the right RPM. Still have the factorymark if I have to adjust i back.

What could be wrong then? Are the dealer able to adjust something with the SSM(Subaru Select monitor)?

Thanks!

O1e

Pleease somebody?

The car run fine and the idle seems to be fine also now. It does not drop and says at ~850rpm at idle. But getting the CEL all the time on the IACV. If will disappear if I let the engine idle for 3-4 sec. Have tried some WOT when the CEL is on and then the engine won't go clean. I seem it is the saftymode then.

O1e

Bump?

markjmd

iTrader: (11)

Where have you been measuring the voltage readings for the IACV, at the unit itself or at the ECU? If the IACV definitely is OK, the only other things that could cause the CEL to keep showing up would be a fault in the wiring harness, or a flaky ECU.

O1e

Hi!

I have measured at the connector in the engine bay, and not out from the ECU. So a different sensor as the TPS, can't make the IACV justify faulty and giving a fault code on the IACV then? Or maybe a airleak? Have tried with start spray(?) trying to find some leaks, but seems OK.

Thanks alot for the reply!

markjmd

iTrader: (11)

In theory, if you adjusted the TPS wrongly and then made adjustments to the IACV to compensate and bring the idle speed back into spec (which you would definitely need to do), that might give you a CEL for IACV, but it should also cause a CEL for the TPS. If you have the time to experiment a little, you could try setting the TPS slightly higher than spec (around +0.1v at both ends of the range) and adjust the IACV accordingly, as from my own experience and reading around here and elsewhere, quite a few classic Imprezas appear to have their TPSs set like this from factory, and will begin running erratically when the TPS is later adjusted more strictly to spec.

So far as an air-leak is concerned, I can think of enough occasions when I've forgotten to reconnect one of the smaller vacuum hoses or a hose has blown off at boost while driving because I'd forgotten to tighten the clip properly, and it's never given me a CEL for the IACV (or anything else, for that matter).

In your position at this point I'd definitely be wanting to check the voltages for the IACV at the ECU.

O1e

Thanks again for the reply!

The problem started when I changed the TPS. After changing it, the car runned OK, but after a period the car was running more and more crappy at idle. And after some time and maybe some resets of the ECU the IACV was starting to give CEL's. I tought also it maybe was the TPS. I did'nt measure the old TPS ofcourse. But what would be the difference by justify it up +0,1V? Should I measure at the TPS in the engine bay or should i measure at the ECU?

The funny part is that the IACV newer give CEL's at idling. Just when driving in slow rpm's. If i let the engine idle for some sec. the CEL is turned of again.

Should the voltage at the TPS drop 0,4V when the engine run? If i set the TPS to be 0,5V at idle with just the ignition on, it will drop to around 0,46V when the engine is running. Should i adjust it to be 0,5V when the engine is running?

O1e

Hi again!

I came now from my garage and everything seems good now

I took it out and drove with it, and tried everything to get the CEL again, but seems good now.

And now the idle drops to around 1100rpm as I stop, then nicely drops stable to 950rpm.

The voltage did not fall as I tought. I have measured on a bad earthpoint instead of at the ECU. So I removed a little bit of the isolation on both cables and got a good and right value, with no difference between just the ignition on and with the engine running.

Then i measured the resistance on all three cables from the ECU to the TPS, and all where very good at 0,2ohm.

The voltage was at 0,467V, and it should be around 0,43V +/- 0,3V. So I justified it all the way down to 0,43V. Then the engine ran at around 1800rpm at idle. Justified it to 0,444V and did some small adjustments on the IACV so the idle was stable on 950rpm, and wolahh

Just hope is stays this way.

Thanks for all your help everyone!

O1e

OOOR NOT

Still getting these CEL's. I have now also checked the cabling. It says that I should have more than 10V to the IACV for the power to the motor, and that is OK. And that it should be more than 10V between the two cables and earth(chassi) from the IACV to the ECU. That is also OK.

Have read the workshop manual, and if I get CEL's 24 (IACV), the IACV was broken, or bad wiring. I have now a new IACV, checked the wiring.

The idle is perfect now, but still getting the CEL's when driving. The CEL disapear when idling again. For two-three times the CEL has staid ON for around 10sek at idle, and then the engine has runned almost like it does on 3 cyl, but not that bad. As fast as the CEL disapear the idle is perfekt again. Starting to wonder if the ECU maybe need to be replaced?