wiseyd

I've got a MY99 uk turbo, and recently had the check engine light been flashing up, it was also running poor and idling lumpy, plus back firing. I did the diagnostics on it, and it said it was the knock sensor, so replaced that part and is running at full power again, and light has gone out with no more store codes.

The car is still idling lumpy and backfiring when driving along and taking my foot off the gas, also abit of black around the exhaust, does anyone know what this could be?

I had the car serviced a couple of months back, my mate changed the plugs in that service, and the idling has been lumpy not long after that.

Thanks

wiseyd

Has anyone had a similar problem with there car?

Any Advice would be helpful.

Thanks
Dan

banny sti

iTrader: (68)

Could also be the lambda sensor causing the issue.

Banny

wiseyd

Ok thanks, so you don't think it could be the maf sensor? Its annoying because its not causing the CEL to show up and store the code.

ReggieMY99

You could do a quick test with the maf, with the engine running, should be at operating temp unplug the maf, if the engine stalls it might be ok, if it continues its probably knackered and the ecu has learned to compensate for the missing input or wrong input from maf

does sound like its running very rich with the back fire and black smoke

anyway a maf costs a lot less then an engine rebuild, so just replace it no matter what to be on the safe side, the mafs for the 99/00 models are very fragile and will eventually go bust, result could be a burned piston

also the lambda could be a culprit here as banny says

fireblade37

iTrader: (1)

=================

Is this actually true? i have no running issues, yet, if i unplug my maff it carry's on running.

wiseyd

Right got a new maf sensor for it, and fitted it. Still shakes alittle at idle, least i got that piece of mind. Could it be the plugs or coil pack? Also is it hard to fit a lambda sensor? Its annoying cause it obviously aint that bad cause it would cause the check engine light to come on, so do you think it will be fine to run until that happens?

Cheers

Splitpin

It could be the MAF sensor. There are only two ways to be certain. The best first option would be to connect the car to a Select Monitor (or an SSM emulator) and study the voltage output from the sensor when the car is running, preferably with the help of someone who knows enough about the cars to know what the voltage profile should be.

If you don't have that option, the only other (and still only authoritative) approach is to buy a new one, put it in, reset the ECU and see if the car then runs properly. If it does, then you know that's what the problem was. If it still runs badly with a brand new sensor, put the old one back in, reset the ECU again and check something else.

Wish people would stop posting this well intentioned but ultimately unhelpful bull. This "test" doesn't prove anything (as you yourself admit by only using the word "might") and doing it can actually damage a good sensor.

wiseyd

Thanks for the advice Splitpin, i got the new maf sensor, but like an idiot i didn't reset the ecu, could that of done it then, would the ecu remember what the old maf sensor was reading?

Splitpin

What, exactly, do you mean by "shakes"? Also, did you reset the ECU after fitting the new sensor? If you didn't, do so and see what that does.

Could be, but don't go looking for further problems until you've tried an ECU reset. In any case, the coilpacks are fairly robust and if that was going you'd likely be seeing more than just a slightly rough idle. The plugs you can always remove and check.

No it's easy, 5 minute job.

One of the first things you need to learn is that the check engine light system has limitations, especially as far as sensors are concerned. The MAF sensor, for example, sends a voltage into the ECU that is proportional to the amount of air going into the engine. The ECU will only throw a CEL if that voltage disappears completely. Trouble is that these sensors tend to degrade gradually rather than being fine one moment and completely non-functional the next, and the ECU can't detect that degredation as it is happening.

If the airflow sensor starts to degrade, it will still output a voltage, it'll just output the wrong voltage. So, the ECU will still think there's air coming into the engine and will set the fuelling according to whatever the sensor sends in. If the sensor under-reads, the ECU will under-estimate the amount of air coming in and set the fuelling too lean. This is bad at idle, much worse under boost.

Tell us more about this "shaking", if it still happens after an ECU reset, and we'll go from there.

Splitpin

Yes, see above. The ECU won't know you've changed the sensor, so anything it has "learned" to work around the old one will still be there. Reset it and then see where you're at.

DeanF

iTrader: (1)

Why not start with the Basics, You say the plugs were changed, Were they gapped correctly ? Coil packs have a habit if becoming brittle on the insulator, When removed, bits fall away allowing shorting/tracking..

Has a vacuum/boost hose come off / become dislodged during the work, Fuel pressure regulator is a common one to knock off without realising.


Do you have a mate with which you can do a quick swap of parts to isolate the issue, But i would definately start @ the start As such.

Dean

wiseyd

Thanks for your advice, im not best with cars, but i love giving things ago. I will go reset the ecu now. When i say shakes, i mean its idles lumpy slight up and down not that bad, but annoys me because it never used to. When your sitting in the car, ever few seconds your get a tiny shake rather than idleing smooth. Alot of people keep saying its the lambda sensor, but i just spent over £100 already today, and i had to change the knock sensor a few weeks ago, which sorted that problem!!! Bloody sensors!!!haha. I had my mate thats a mechanic service it a few months back and change the plugs and its been abit lumpy since then.

Splitpin

This is significant info - would have been worthwhile telling us this in post #1. And, it makes what Dean's just said all the more pertinent. If something changed immediately after that service, then it suggests that something done during it might be responsible. Ask your mate what plugs he used and what he gapped them at. If he doesn't remember, take them out and check.

wiseyd

Ok i will do, what should be the gap setting for the plugs? It was a couple of days i noticed it doing it after the service. A few days before that though, i had just put on a prodrive ecu, that i picked up. I have put the old one back on it and that made no difference, so put the prodrive back on. I'll check the plugs, thats not 2 hard to do is it? I wanna try doing these things for myself, as im interested in learning abit along the way.
Thanks for all your advice though, will try the now the ecu is reset.

Proby521

iTrader: (1)

errr... he did. Read last lines of first post! Easy to miss though! I think your right though, my first port of call would be plug gaps. I think they should be about 0.65mm. Do a search and see what comes up.

wiseyd

I've reset the ecu, its actually not running to bad, still pops occasionally, but i do have a scooby sport exhaust on it. Idle is very slightly lumpy, you rev counter hardly moves at idle 800 rpm i think!!! I might check the plugs tomorrow though.

Splitpin

Occasional pop and bang would be considered normal with a standard ECU and a partial/full decat, so it doesn't sound on the face of it as though you're far away from the ballpark here.

"Normal" idle stability, once the engine is fully warmed, should be somewhere around +-50rpm of the target so you'd expect to see a little variation. Assuming your man used the OE replacement plugs, you should find NGK PFR6B's gapped at 0.7mm

Oh - apologies for missing the reference to the service in post #1. That'll teach me to skim!

wiseyd

Cheers, for all your help mate, car seems to be alright, i think im being alittle fussy. Love the car and the way it runs. Actually would love to get a newer scooby with A/C in it!! All this nice weather!!!lol

ReggieMY99

So it was the maf, very nice

Splitpin nice with your advices, but dont trounce others here for helping, not exactly born yesterday and had 3 imprezas

and i used the unplug on my imprezas a couple of times and it surely was the maf, i never reset my ecu for a maf change, seems to have the right readings right away

Splitpin

You were the one posting bad advice, and it's difficult to say that in any other way than just saying it. The trouble with old chestnuts like this is that if nobody questions them, they just get repeated (I daresay you're only saying it here because you saw someone else do it). As I acknowledged in my response you might mean well by it but that doesn't change the fact that the "test" you proposed doesn't prove anything and potentially does more harm than good.

It simply doesn't, can't, prove whether the sensor is "good" or not, as you implicitly acknowledge in your original post by hedging your bets with words like "might" and "probably". It doesn't give the certainty needed. You could try it with a brand new sensor and find the engine doesn't stall, and, worse, as mentioned, it can damage a good sensor due to the sudden cooling that results from removing the element heating with the engine running. It's a plain bad idea, albeit one that has been repeated enough that people regard it as fact.

Okay, so, let's think about this. In your original post you suggest a "test" that, in order to work, relies on the ECU having learned "to compensate for the missing input or wrong input from maf" (your own words).

You then say that when you change your MAF, you don't perform a reset, despite the fact that, if your "test" has any basis in fact, the ECU will have already learned to compensate for the "missing or wrong" input from the previous sensor?

Do you not see a bit of a flaw in your own logic here? If you change a sensor without resetting, the ECU will still be working on learned values from an old, possibly broken or seriously degraded one, won't it?

obwseamus

I may have the same problem with my obw 2.5. I had a cam seal blow and had it replaced with the timing belt and it hasbeen running rough ever since. i have replaced the following.
coil pack
leads
maf
plugs
air filter
no change to running its still rough up to 3000 revs and then it kicks in and takes off with very nearly the power it had b4 all this happened.

any advice ?

DeanF

iTrader: (1)

That could be many things, induction leak, vacuum hose, Faulty O2 Sensor as they are not really used on boost/Wide open throttle Etc, Generally Plugs & or coil packs would give problems under boost/load More

Dean

obwseamus

all vac hoses ok
compression is 125 is this normal?
i havent reset the ecu as don't know how to. any tips

DeanF

iTrader: (1)

What year of car ? Do a search for the compression, 125 is Ok from memory but perhaps a little low, How many miles Etc.. Dependant on year of car, On a classic You can connect the test cables under the dash & follow a driving procedure ( Again documented on here) or most just leave the battery of for an hour, Or alternitavely, Disconnect the Ecu's power lead/plug directly from under the passenger footwell carpet.

Dean

obwseamus

its a 99 2.5 outback with 103000 miles


i will try the ecu plug and let you know thanks

obwseamus

i found and unplugged the ecu and it fixed the problem. thank you !!!!!!!!!!!!!!!!!!!!!!!!!!!!!