Impreza misfire

[kyletilley]

Morning all, took my impreza out for a spin last night. It had been parked up for about 10 days, all was well for about 5 minutes but then it started to stutter on boost between 3-5k rpm. All was well on the boost and afr gauge's. My immediate thought was the MAF, disconnected it and the car stalled as it should. Left the car whilst I went for dinner, came back and sure enough it ran fine for 5 minutes then the misfire under load came back again.

Anyone have an opinion on this? The car is a 95 WRX import.

Thanks!

[ScoobieWRX]

First port of call is check for DTC's (error codes) and see if it comes up with anything.

Here is the routine....
Error check: http://www.scoobypedia.co.uk/index.php/Knowledge/R...
ECU Reset: http://www.scoobypedia.co.uk/index.php/Knowledge/E...

The MAF sensor could still cause this stutter as it starts to break down reading less airflow over time although it still acts like it's working, and doesn't always show a code either. Bit strange though that it's started after sitting for a while.

If the codes are clear then other possibles may be a MAP sensor or crank sensor, again slowly breaking down over time and not sending a code. A faulty crank sensor might also start to cause start-up problems that get worse over time eventually leading to a non-starter.

You should check spark plugs and check for boost leaks. Just a small leak could cause that stutter although it would carry on past 5000rpm and anytime the turbo is meant to be on boost. This goes for all versions.

If you are still stuck and not too far away from me after having looked at that lot you're welcome to come and see me in Northampton. I can plug in to your ECU to see what's going on. No charge as i'm always happy to help out a fellow PHer.

[kyletilley]

Hello mate, thanks very much for your reply. Unfortunately my car appears to be OBD 1. So I can't just plug the laptop in..... I reasoned (rightly it wrongly!) that it was unlikely to be anything too severe as it boosted ok initially. I was also told it could be a sticky boost solinoid?

Thanks

[ScoobieWRX]

The kit i use can plug into and map your V1-2 ECU on-the-fly. In fact what i would do is swap your ECU briefly as i have one here with an ESL daughterboard fitted (mapped back to standard Z4 calibrations) i got back from a client of mine that reverted his car back to standard a while back.

This eliminates any potential ECU problems and allows me to have a look at what the car is up to via my laptop. I can view timing, fuelling, boost etc... I use a wideband to monitor AFR's and det cans for listening to knock, all in real-time.

The best way to fault find though is on a rolling road. That way you can test for boost leaks visually whilst under boost conditions, and other faults. I don't map using rolling roads but you can't beat them for fault finding.

If after you have checked everything i mentioned in my first post and you're still no further forward then you're welcome to come to me and i'm happy to have a look, otherwise there will be a Suabru specialist near you, likely with a rolling road, but you will be charged for rolling road time and labour. If you don't mind paying then that's probably the better route for you as it will be much closer to home.

I believe you're in Wiltshire??

[markCSC]

Cheack all the stuff ScoobieWRX said

Also have you tried cleaning the MAF with electrical conatct cleaner? Afterwards re-set the ECU

Normally when the boost solenoid is sticking you get overboost and then fuel cut. You WILL know if you have that as it is rather violent.
Cleaning proceedure here anyway http://www.scoobypedia.co.uk/index.php/Knowledge/C...

[TAS1981]

All of the above is true. I found cleaning out boost control solenoid and MAF did the trick for this particular issue and is far and away the cheapest solution! From your description if its spiky boost all over the place under load it is likely this. If that does not work then I have found a whole host of tiny things can make a difference to boost.

If the stuttering is violent and chronic its probably a boost leak or fuel cut similar, have a good look at all the pipelines for leaks and splits.

If its not boosting above a certain level ie less than the normal boost you'd expect and its been left standing a bit then petrol could be the issue. Don't know what ECU you are running but cr@p petrol will be an issue. A touch of Octane booster will tell you if that is a prob.

Next is coil packs, spark plugs which fail under load when they are a bit knackered. When was the last service? A once over all of these things will cover most of your issues I imagine.

[kyletilley]

Well cleaning the MAF helped. A week of fault free motoring and now the problems come back. Only this time, cleaning the MAF hasnt helped. New MAF time?

[ScoobieWRX]

Did you check for DTC's (Error Codes) manually via linked connectors under the steering column??

I posted links for the above procedure in my first post. No OBD1 diagnostic equipment required for this check.

[kyletilley]

42 and 31 were the only codes thrown up. Not sure how either would cause the misfire? As it happens regardless of throttle position and idle seems to be ok.

[ScoobieWRX]

I wouldn't go buying stuff until you know what has gone wrong. You need to get the car to a specialist with a rolling road.

I don't ever use rolling roads for mapping but for fault finding they are the best way forward. That way you can get the car on boost with the bonnet open checking for boost leaks, sparking HT leads, leaking exhaust, fuel and oil leaks under pressure, etc... etc...

This should be the next stage in investigating your scooby problem and it's money well spent. As i mentioned before you're welcome to come to me and i'd be happy to take a look in the first instance.

[kyletilley]

Thats my next port of call. I gave the car a service and reset the ecu yesterday as I already had everything for it. Seemed to cure the misfire but now wont boost past 6 psi. So putting it up on the ramps tomorrow to have a proper look underneath at the manifolds etc.

If no splits are apparent i'll come up and see you.

[ScoobieWRX]

When looking for exhaust manifold leaks it's worth having a white light source to show up exhaust fumes so a White LED inspection lamp is ideal for that.

When looking for boost leaks washing up liquid/water solution sprayed onto all the intercooler pipe joints should show up any leaks. When revving the engine from under the bonnet you will still induce c.0.1bar or so of boost pressure, just enough to see positive manifold pressure and enough to induce bubbles wherever your soapy solution finds a leak.

If all your sensors are OK then you should be good to go. Strange that it's now not boosting over 6psi. Sounds like it's gone into some kind of limp mode or the boost control solenoid isn't functioning properly. Did you remember to disconnect the connectors under the dashboard cowling after you reset the ECU? Did you reconnect anything you disconnected during the service?

[kyletilley]

Just to update you. The problem has been diagnosed as the boost solonoid. Thanks for all your advice. Should I need any mapping etc done I'll come and see you.