70 replies to this topic

[podmak]

I am trying to pin-point an odd problem with how smooth the engine idles. 

 

 

Currently, the idle rpm is not stable. It intermittently fluctuates between ~300 and 1000 rpm (sometimes it goes way past 1000 - really revs) with engine warmed up to ops temp, idling at the lights. Also the fuel consumption is nowhere near an acceptable level; It's currently at about 300-350 km per tank. 

 

 

I've wondered for a long time what it might be, and I already tried replacing a bunch of stuff. I am convinced (amateur convinced) that it has something to do with the air supplied to the engine. 

 

 

1.) Here are the parts I replaced (assuming these parts are now out of equation):

 

Fuel pump, filter and injectors. 

Mass Airflow Meter

Air filter

Spark plugs and leads; 

 

2.) I cleaned the area around TPS (throttle position sensor) thoroughly with tooth brush and some throttle and carby cleaner, and used the Subaru upper engine cleaner on a number of occasions; 

 

3.) I disassembled and re-assembled the piping around inter-cooler, sprayed and cleaned all the parts; 

 - I noticed there is oil film all around the inside of these parts on the passenger side (Australia) air piping around turbo. Is this normal? The oil in there? 

 

 

Now, le question that I have: What is the part that I am missing from the below? 

 

1.) The air comes in through the intake pipe, goes through the air filter. 

2.) Then through the Mass Airflow Meter.

3.) Then it forks to both turbos.

4.) Then it passes through the inter-cooler

5.) Past the throttle position sensor. 

6.) And then it goes through the intake manifold, past the injectors, mixes with fuel and into the cylinder. 

 

Those are the main points - what did I miss? What else is used to measure how much air goes in? Because that is definitely off; Last time I checked the spark plugs, they were all black - dry black;

 

 

Could it be I have an air leak? That would almost certainly explain it BUT (and this is a big but) sometimes, the car runs perfect - it idles and purrs like a proper car - at other times, rpm fluctuates. So I assume this is electrical problem somewhere, one of the electronic gadgets that helps the on-board PC to measure how much air goes in is wrong, but I don't know what else I should try. 

 

Could it be the passenger side turbo is leaking oil, hence the oil inside the air piping on this side, and so it messes up the readings? (Surely how hard can it be to pull out the passenger side turbo and check? Seriously, on the scale from 1 to 10 - where 10 is piston ring replacement...) 

 

 

Phew, long post, thanks for your input. It's always appreciated. 

[podmak]

Symptomatic of a vacuum leak. Are all your turbo and intercooler pipes in good working order and hose clamps tight.

 

Well, that's THE question; 

 

I wonder, and I have for a while, can the system be pressurized - and is it supposed to hold the pressure? That way it should be fairly straight forward to check... otherwise, visually checking all the piping for cracks might be a little tedious. ...  

 

 

What about my other points, do they not count for anything?

[duncanm]

Don't forget the old standby for diagnosis -- idling engine while spraying carb cleaner at likely leak points to see if it resolves a high idle.

 

You didn't mentioned IACV (the idle controller one, not the TT jobbie) -- did you clean that out ?  It is independent of the TPS.

 

Idle variability caused by air leaks would only be those post throttle body -- vacuum hoses, fittings, boost gauge lines etc.  Also make sure that vacuum tank check valve is ok; maybe just start off by clamping off the vac lines leaving the manifold one by one for diagnosis.

[podmak]

Don't forget the old standby for diagnosis -- idling engine while spraying carb cleaner at likely leak points to see if it resolves a high idle.

Tried that arount the IC/BOV area - no change.

You didn't mentioned IACV (the idle controller one, not the TT jobbie) -- did you clean that out ?  It is independent of the TPS.

I think this is the piece I am missing in my original post at the start of this thread. Will get to it over the weekend then I'll get back here with the results.

Idle variability caused by air leaks would only be those post throttle body -- vacuum hoses, fittings, boost gauge lines etc.  Also make sure that vacuum tank check valve is ok; maybe just start off by clamping off the vac lines leaving the manifold one by one for diagnosis.

This is on schedule;

[podmak]

The best advice you could heed is:
Start small.

Thanks I will. I mean I am. I mean I have.

Check your engine codes on the ecu with a scanner, don't waste time doing the black plug dance.
Check your vac lines and their integrity.
Check the condition of your IC pipes and clamps, make sure they are sealing and the whole airflow system post turbos is in in good working order.

Thank you, I will.

[LukeFranky]

A smoke test is actually easier to setup than it sounds. Anytime my engine starts running rough it is the first thing I test now.

 

I use a little portable air compressor. If you only have a big compressor you will want somehow limit the airflow to a nice slow and stead flow.

 

Get a tin can with a lid of reasonable size. A milo tin or a paint can will do the trick. All you need after this is some tubing from the compressor to the tin can's lid, then from the lid to a vacuum line connected to the intake manifold on your car. You want the can to be 98% sealed so that the majority of smoke is pushed into the car, but it doesn't matter if it leaks a tiny bit.

 

Block off the intake pipe near the air filter. This is so the smoke is encouraged to go out any leaks you may or may not have. I use a spray can lid so if the pressure builds up too much it blows off and releases after a certain point.

 

Pull out your dip stick out a little, this is so any smoke that may get pushed into the engine via the PCV gets released and doesn't put pressure on your oil seals. Don't worry if any smoke comes out the dipstick.

 

Now comes the fun part, get a rag that smoulders and smokes well but doesn't burn too quickly, I use a very slightly oily one. Put it in the car and light her up, let it burn for a few seconds then blow it out so it is only smouldering and smoking. Smack the lid on firmly, wait a few seconds then turn the compressor on.

 

Now the air from the compressor should be pushing harmless smoke into the intake. Big leaks will show up quickly, smaller ones might take a while. If nothing obvious happens, turn the compressor off and check there is still smoke in the can. If not, try again.

 

If the only smoke you ever see comes out near the air filter blockage and the dipstick you may not have a leak. Try it a few times to be sure.

 

Final notes:

Keep in mind you want smoke, not fire. No need to burn you garage down, but if you're smart it's safe enough. If you use something that is too combustible, or if you fan the fire too much with the air compressor, things might get a bit too toasty in that can.

[podmak]

Thanks; Sounds like fun. Still haven't had a chance to check anything. Will post here with results once available.

[podmak]

It took a long, long time, but I finally got around to do it.

 

I built a small smoke machine, and found a tiny air leak between the driver side turbo and inter cooler - the plastic pipe did not fit well to the aluminium bit - just used some black silicon and tightened the clamp. 

 

This is the part: https://goo.gl/Sh8Aac

And the plastic bit that was leaking is this one: 

14462AA081 DUCT-AIR INT 14462A 1

 

There is no air leak now, but it still didn't solve my problem with the unstable rpm at idle; 

 

Must. Obtain. Error. Code. Scanner. 

 

There is one at Supercheap auto made by Bosch, goes for about 120AUD - wonder if that's worth it... 

[LukeFranky]

I doubt silicone goo will hold boost. Might be time to replace those hoses with aftermarket ones. There is a thread around here about it.

I don't think a $120 scanner is what you need. I've used the cheap OBDII scanners off ebay.

[podmak]

I replaced the IACV - no change. 

[duncanm]

When you say idle fluctuates -- does it switch between modes (slow, fast), like when the AC turns on, or does it just wander slowly up and down?

If the former -- maybe there's a problem with the AC fast idle circuit.

Other things to look at (easy to check while idling), all can lead to dodgy fuel consumption, especially O2 and coolant:
Oxygen sensor voltage
Coolant Temp sensor voltage
MAP sensor voltage

These can all be accessed with a cheap VAG-com cable and freessm.

[Ordex]

Like the AC compressor, the power steering pump also has a sensor which I belive is used to increase idle speed when under load.

[podmak]

These can all be accessed with a cheap VAG-com cable and freessm.

Feeling a little dumb writing it, but I think this is the single most useful comment I've read since I got the car... can't believe I didn't think of this earlier.

 

Also i don't want to make it sound like all the other guys didn't help nowhere near as much. But mate, this is gold.

 

Will post updates here. Can't wait.

[podmak]

Well... as embarassing as it sounds... cheap vag-com cable arrived allright, but it has OBD2 plug... and I don't know where my OBD2 socket is under the dash... I'm guessing it's not.

Just wanted to make sure.

[Ordex]

What year is your car? If there is no OBD2 then it must be early? Mine is 93 and only has OBD1. It's still possible to read all the parameters, but obviously you need a different cable and it may be tricky to work out the memory addresses to read

[podmak]

Chassis is from 1995; But the twin turbo came off another car along with transmission, what was originally 2.2 auto is now 2.0 twin turbo manual. And I sincerely don't know what year the engine was.  

[podmak]

When you say idle fluctuates -- does it switch between modes (slow, fast), like when the AC turns on, or does it just wander slowly up and down?

If the former -- maybe there's a problem with the AC fast idle circuit.

Other things to look at (easy to check while idling), all can lead to dodgy fuel consumption, especially O2 and coolant:
Oxygen sensor voltage
Coolant Temp sensor voltage
MAP sensor voltage

These can all be accessed with a cheap VAG-com cable and freessm.

 

 

I tried taking out the AC belt, just to rule out a problem with AC, as the AC clutch currently doesn't engage - problem for another topic. Idle RPM still fluctuates. Will post a video. 

 

Also, when you say check voltage, I ask because obviously, if I disconnect the sensor from the harness there will be no voltage, I think this was assumed being measured via the OBD.... but given the situation... can't I check the resistance manually instead?

 

Also a bit of a long shot - do you know, by any chance please, the correct reference values for the above mentioned readout? I assume there will be some kind of a table with these values. 

 

What am I looking for here? 

[podmak]

OK.

 

I am now leaning towards the rpm fluctuation being a result of engine misfire. 

 

I already ruled out the air issue, I measured the impedance of the O2 sensor, I did a smoke test of the air intake system, it all seems ok. 

 

I replaced the fuel delivery system, injectors, filter, pump... It's all working except the fuel gauge is limping behind - I must have disturbed that little electro-magnetic ruler on the fuel pump assembly and now it gives readings that are a little off. Clumsy hands... 

 

 

But, I did notice, after embarrassingly long time, that sometimes, and this is best felt up the hill, the car pulls nicely, then midway through skips a beat, then keeps on going full throttle, then check engine comes on. It has long eluded me, but I started to notice the engine shudders while idling, and that's I think what is happening - misfire. If I had a really tiny camera that I could put in side each cylinder to record this - would be cool, but umm.. this not a fairy tail... 

 

 

So I am now left with the only thing left: sparks. I checked the plugs, they are bone dry, no oil there, but the carbon build up on them is black; I only checked passenger side. 

 DSC_0547.JPG   230.43KB   4 downloads

 DSC_0550.JPG   270.34KB   5 downloads

 

I did notice that my alternator is probably not in the best shape, when I switch on my headlights I see a noticeable drop in RPM. But I imagine this shouldn't have that much of impact on the engine cycle(?)

 

Ordered a new set of coils - once they arrive, the story will continue.