s13silvia

Well in one sense it is - it finally proved it was 65% rwd today exiting a wet roundabout and it went sharply broadside across three lanes - a little too 'smokey and the bandit' for my 7am office trip but hey ho.

However I have an intermittent missifre issue and wanted to confirm if anybody else has had this exact issue:

Sitting at 60mph or 70mph its most obvious, comes in an out every 1 min or so. Then sometimes stops for 10-15 mins,then comes back in.

Runs fine at steady throttle then missfires/fires on 3 for 3-4 secs then goes back to 4.

If booted boost sits at 0psi until the missfire stops (usually within one sec) then romps on as usual and revs smoothly on 4 through the range until back on steady throttle.

It's been happening the past few day. I'm strongly leaning towards it being a coilpack issue from previous jap-turbo ownership.

Does this sound right? And is it worth replacing them for newage or are there no real benefits to the newage swap if going for new classic packs.

Does anybody have the subaru part numbers for stock import wrx/sti coils for a 1997 Type R STi?

Thanks,
David

s13silvia

Wait- I have one coilpack. Fantastic- so problem is either a lead or plug - shame it's not a consistent missfire could be diagnosed on idle.

Gavscooby1

try changing your plugs mine did this its all good now

s13silvia

Cheers Gav- yeah I'll be getting them changed and gapped asap - might aswell get some leads at the same time I guess!

s13silvia

Well got the plugs and leads replaced after a solid 35 min session, opted for PFR7B's after all the reviews on here - starts up perfectly but not going to get a chance to drive it until tomorrow when the snow clears.

s13silvia

Okeydokey- no difference with 0.7 gapped pfr7b's and new leads. Had prf6b's already in it. All 4 plugs came out the same, dry and a bit black- pretty normal.
Put some injector cleaner through it too.

Guessing it's not the coilpack as it's the one central pack for all four.

Going to replace the fuel pump/filter and fit an fpr just in case.

I'm concerned about the ECU as it's an unknown mapped 8S, I have a stock 6S ECU ready-to-fit but as the engine has a TD05H turbo it's definitely not mapped for boost (though thinking about it, so long as I stayed off-boost it might be worth a try?).

One other thing I'd like to test is the voltage going across the MAF meter and the O2 sensor - what should they be at if functioning correctly? Also going to give the maf a clean.

To help with the diagnosis in case it sounds familiar to anybody, the revs don't matter (can happen anywhere), it happens off-boost so not turbo/boost related (could feasibly be vacuum related I suppose, but 90% of the vacuum hoses were replaced a couple of months ago). It hesitates (on 3-cyls) on steady throttle, till the mph drops by 4/5 then you feel the 4th coming back in and it eases back up. Putting the foot down doesn't stop the missfire, it usually only lasts for 3-5 seconds though and once it starts it happens every 30-120 secs. It seems as though when the car starts from dead cold it can drive perfectly for up to an hour before hesitation then it's frequent from then on. Oil and coolant levels are good, no boost issues once she boosts up.

Any advice much appreciated fellas.

Sprint Chief

Some things that are worth checking:-

The earth for the engine control electronics is sound. It attaches to the back of the engine just beneath the intercooler on the passenger side as I recall. I think you may have checked this, can't remember! Edited to add: make sure this connection is clean and tight.

Check the cam and crank sensors are good and the wiring of these sensors keeps well away from any HT leads. Sometimes when engines are swapped out wires are not properly routed. The cam and crank sensor wires in particular are prone to picking up noise from these and receiving false triggers.

Did you check for codes? I think you did and there weren't any, but again forgetting where you got to with this...

s13silvia

^ Well Sprint Chief - Looks like the hunch of O2/MAF sensor was on the cards, checked the codes and assuming it's not historic I have a code 32 which points to AF system 1/AF system 2/ O2 Sensor (with a 97 Type R, is 32 specific to one of those things, or could it be either?) - do you know by chance the voltage which should be across the connectors on the MAF and the sensor as I purchased a voltmeter yesterday.

Cheers,
Dave

s13silvia

bump - reset ECU so waiting to see if it appears again.

Sprint Chief

As I recall, MAF should be somewhere around a volt (possibly a bit more, possibly a bit less) at idle, and will increase with load and rpm. The lambda sensor will be difficult to test at idle as it will fluctuate quite a lot - there isn't really enough air flowing over it at idle to get a consistent single reading. Should read between 0 and 0.8 volts under idle, moving to a more stable 0.7-1 volt under load.

The symptoms don't sound like a maf or lambda problem to me but failures don't always behave the same, and those two sensors are always well worth checking as they are both critical to correct running of the engine.

s13silvia

Bah.. cars doing my nut in.

Could somebody advise me the way to check the front lambda sensor? I can't really test it when the cars warm as I'll burn my hand on the downpipe/turbo housing, not to mention the wires are that thin and flimsy the chances of me poking the needle through them are slim (I'm not the best with dexterity). It has 3 wires, 2 white and 1 black - which is the signal wire, which is the live for the heater and which is the earth? Assume when on 2nd ign it should be recieving 12v to the live at the connector?


I pulled the MAF sensor and it cut the engine out as it should and threw up a code 23, but the car's continued to do it's random missfire and it hasn't thrown up any codes since the ECU reset which doesn't help a great deal.

The engine temperature seems significant - it only does it after 30-60 mins of driving easily, or quicker if giving it a bit of welly. Then after it's started it continues to do it even if you stop the car for a while. It only stops doing it once you've left the car off for a few hours to cool down to cold - and then re-starts after a further 40-60 mins. This makes me strongly think it's not a MAF issue, nor plug/ht lead (these have been replaced) - this wouldn't be temperature dependant. The fact it was happening before at around 3k rpm made me think it could be the ECU switchover from Lambda > MAF but it turns out it happens all over the joint on steady throttle, 1.5k>3.6k

So logic rules out >
MAF
Plugs
HT Leads
Coilpack (it's the single coilpack - a) it would affect 2 cyl's at once, and b) it would most likely show on boost, not off-boost)
Boost solenoid/ boost leak (there are no problems once the car's boosting and runs normal pressure 12/13psi)


Logic would suggest:
O2 sensor
ECU (I have a repalcement stock ECU which i'll test off-boost)
Water temp sensor?

Can't think of any other possibilities?

Sprint Chief

Check the lambda voltage at the ECU. Plug B136 pin 5. (B136 is the 22-pin plug, pin 5 in the top row, 5 from the end of the ECU case).

There is no changeover between MAF and lambda. The car runs continuously from the MAF, corrected by the lambda. The only thing that changes is the time constant of the closed-loop circuit on the lambda correction.

Have you checked the engine earth or the cam/crank sensors and wiring yet?

Water temp sensor is worth checking, an intermittent fault at a particular temperature could cause sudden changes to the engine operation.

Remember also that the calibration of the MAF sensor in the ECU may not match the calibration of the MAF sensor in your car, as you have a mismatch of components.

s13silvia

Good morning Sprint Chief, finally had some time to work on the car this morning.
Checked the pin on the ecu, I have a relatively crap digi voltmeter, and not sure if these readings were accurate but seemed to jump about a bit between .000 and .001 . Nothing very exciting. Was in by myself so unable to put engine under load- will get my partner to jump in the car later.

Took off the MAF - holy hell it was FILTHY inside, covered in sludge and dirt. The HKS filter wasnt fitting right, one of the grubs had come out of the plastic on the filter so tapped that back in and used a smaller bolt to tighten that hole. Got it flush so no more big gaps. Cleaned the sensor thoroughly with brake cleaner till it was shiny and cleaned the rest of the inside.

Checked the engine earth - it's good and solid

Checked all the sensors wiring - all good and the crank/cam sensors dont cross near the HT's.

Checked all the vacuum hoses - most have been replaced recently and all are good condition

Checked the rest of the wiring in the bay - only fault I noticed was the driver-side front most injector's yellow wire has the sheath split open as it bends, immediately after the connector. The copper wire is blue underneath so not the best really - but looks to be a bitch to replace and repin.. I guess all four injector cables go into the main loom?

Removed the MAF for inspection - certainly looks to be good from the little you can inspect.

Not switched ECU yet- at the end of the day it WAS running fine for a good few hundred miles before the missfire started.

Ps. I notice there appears to be another o2 sensor coming off one of the manifold branches? Is that an AFR sensor? Would it be as well replacing that at the same time as the lambda behind the turbo?

Sprint Chief

Hmmm, starting to run out of the more obvious problems now.

Engine loom is pictured here: https://www.scoobynet.com/showpost.p...47&postcount=9

The earth used by the ECU is marked "GE" in the above diagram - just wanted to check this was the earth you looked at. I have seen this loose before causing problems.

Other than that, really need to get some diagnostics on it when it is exhibiting the fault and try to trace backwards from there... not much help to you at the moment though I'm afraid

s13silvia

Yep that's the earth - tight as a nut. I'm going to take apart the top of the MAF later today and check the points on it.. might be something as simple as a cracked solder.

Sprint Chief

Oops I didn't read the message you posted up. Those lambda voltages look a bit odd but it is difficult to be sure for two reasons, firstly your meter might have a bit of a slow update rate, and also need to be absolutely sure you're monitoring the right pin and getting a good connection to it.

I would expect the lambda voltage to be jumping up to slightly higher values even under idle.

s13silvia

ok, properly tested the lambda voltage today - assuming the white cable 5th pin up from base is correct (8S ecu so not certain if its same pinout as 6S, or indeed which loom is being used) (didn't realise that there was a flap that lifted on the connecter!) got a bit of an odd reading - could you tell me if I'm on the right selection on the reader? It very slowly moved between 2.6 and 2.5, mostly dropping - revving made no difference.



Since the MAF has been cleaned getting another Code 23 (MAF) so going to reset the ECU once more once it dries up. Cleaned the boost solenoid too since that's all that was changed shortly before the missfire arose.

I checked the injector wire by running the car and giving it a good firm wiggle, cyl kept firing fine

Last night I took it for an hours drive at 50mph and it didn't missfire once. Then gave it a little bit of lick and drove it home after giving it a bit of lick the missfire started coming in and out, continued at 70mph, but once I dropped to 50mph it stopped happening after 5-10 mins then didn't happen again until I got home.

s13silvia

What a state! Just took it out and it's bent to bits with 2 of the small vents crushed shut. Not sure what to make of the outside, very black with sooty deposits - running too rich?

Is this the correct replacement part?

http://www.worldcarparts.co.uk/tabid...S/Default.aspx

s13silvia

bump

Sprint Chief

Oh yes that lambda sensor looks a right mess!

I believe the part you link is correct. The OEM subaru item is about £200+ but that bosch part works perfectly. I think the connector might be wrong and need splicing in but that has to be worth saving £150 or so...

360ste

iTrader: (1)

The question now is how did the sensor get damaged?

s13silvia

TD05H and randomly mapped ecu from an unknown car would be my guess want to get her mapped with the right ECU but it's not cheap will just have to replace the sensor, reset the ecu and hope that's whats been causing the hesitation/missing - then wait to get it mapped properly.

Abbylad

iTrader: (4)

Are you selling the car on ebay now? If it's you you've described the colour wrong, cashmere yellow is a lighter yellow