Botswana WRX

Help, Desperation, any ideas warmly welcomed, are we missing something??
My mid my02 Sti never went well, I know because I had a Prodrive WRX 2002 before and it was significantly quicker. So off I went to Rally Star the South African Subaru tuning specialists.
They recommended and fitted a Unichip and Bosal 76mm decat system.
We should have noticed this at the time but on the 'before mod' Dyno run the ECU was permanently in limp mode, we put this down to poor octane fuel. We only get 93RON on a good day in J'burg.
With the pipe and chip fitted we were getting 180Kw uncorrected-at-the-wheels and all appeared good (that’s 240bhp) running between 1.2and 1.3bar.
On the way home it missed twice and knocked a hole in Number three. By the way this level of tune has been done to many SA Sti's without any problem.
On stripdown No3 appeared lean (lighter colour than the others) and the fuel pump was suspected as we had had pumps fail before due to African dirty fuel, usually after 50k kms or so. The pump was checked and sure enough although it was producing around 4bar it was a little slow getting there. The rebuilt engine went back in, it looked good on the Dyno and a friend delivered the car back to me in Botwana, Next day I gave it a try running 98RON and guess what - a hole in No3.
Since then its never done more than 25miles, once it blew up on the dyno.
To cut a long story of engine rebuilds and replacement subs short, this is what has been done,
Changed injectors.
Fitted second fuel pump -gives up to 5bar.
Changed MAF sensor and both Lambda sensors.
Changed fuel regulator. ECU,
Fitted another Fuel rail with yet another different set of Sti injectors.
Changed Absolute Manifold pressure sensor.
Fitted new Sti inlet manifold.
Heads pressure tested.
Changed Engine wiring harness after finding a slightly loose pin on No3 injector. (thought we found it that time)
Checked over and over for boost leaks.
Ran all the SA available Subaru diagnostics-no fault found.
In fact the only sensor not changed is the cam position sensor- but that couldn't cause lean running could it?
Still even mapped as rich as possible with all the above done it blew no3 again.
Last thing we have done is de chipping and restored original pipe with everything back to standard.
So back to square one, it runs in limp mode all the time even on 102.3RON race fuel and 13.7AFR which seems lean?? with two different standard Sti ECUs but fails to destroy no3 piston like this although it now slaps a bit.
I'd take it back to Subaru under warranty but now they know it was modified, because of the loan of Factory test gear etc. so I don't expect much help there, I hope I'm wrong.
What is this little problem??
We are now changing the ECU to engine harness out of even more desperation.
What is it???? All sympathy and sensible suggestions appreciated.
PS I have to say RallyStar have been absolutly incredible in their efforts to solve this problem even though they haven't.

WUZ

iTrader: (14)

I would have thought 13.7:1 is OK on idle as the Lambda is controlling AFR at idle) but on boost you'll be looking for 11.2-11.5:1

Russell

Botswana WRX

Yep I know its lean but why?? hi pressure pump fitted and checked at the end of the rail and two different sets of Sti injectors but still lean. We even tried a new rail in case the original had a fault.

WUZ

iTrader: (14)

The ECU is controlling injectors pulse at idle to get the AFR that is programming tells it that it needs!

13.7:1 is not that lean at all (when at idle!)

Russell

Botswana WRX

Thanks Russell but its lean on boost as well - this with the standard unmodified ECU
Chris

The Fixer

I once had a problem with the standard subaru fuel rails under the manifold, there were massive fuel pressure differences between one side of the rail to the other, mainly because where the pipes bend they get flattened or could be a fuel blockage. Try parallel fuel rails from MRT or APS. Each fuel rail is fed from a centralised distribution block and the return is T'd up into the FP regulator. At least you will ensure the same amount of fuel is getting to each bank of cylinders.

Hope this helps

Conrad

PS Nice part of Africa

Botswana WRX

Thanks for the suggestion Conrad, We are measuring fuel pressure on the return side so there should be plenty of fuel available. Its always the No3 that lets go so obviously its still that one getting less fuel. But the basic problem remains, why don't we get enough fuel?? Am now thinking to try a separate supply eliminating the tank to engine and return plumbing to see if that does anything for us. After all pressure ain't the same as flow, could be bubbles?

And yep Bots is a well kept secret, not half bad and NO SPEED CAMERAS! although they do have hand-helds.
Chris

Jolly Green Monster

iTrader: (4)

have you used the same set of heads each time it has been rebuilt?

Wondering if there is valve problem on number 3..?

Botswana WRX

Yes same heads but very carfully checked including pressure test. We did try a different inlet manifold though.
Any one ever looked at injector pulses on a scope? I'm thinking of trying that as well.
Chris

Jolly Green Monster

iTrader: (4)

you replaced the engine loom but have you replaced or added extra cable as a temporary measure from ecu to injector number 3?

The Fixer

Measuring at the FPR doesnt show the difference in FP on one side of the engine to the other. Convert to parrallel fuel rails to be sure.

Botswana WRX

No but did check continuity under load, disconnected injector end and put a 55w bulb on connector and applied volts to disconnected ECU connector, nice bright light

Botswana WRX

Agreed Conrad, but keep in mind we tried a fuel rail and injectors off another STi that was doing fine on the AFR front and still no change to mine - lean.
Chris

cong

so after you rebuild the engine you then proceed to run it on the rollers without weraring the engine in? no wonder you blowing a piston

P20SPD

iTrader: (1)

The Fixer

lol

Botswana WRX

Even after 30k miles it blew a piston. we need to concentrate on AFR here, its lean running and detonation that makes holes in pistons

P20SPD

iTrader: (1)

Assuming you have checkd the fuel pressure regulator?
No air leaks post MAF?

Botswana WRX

Yes replaced fuel pressure regulator and checked for air leaks everywhere.
But now you got me thinking (doesn't happen often)
Why 13.7 AFR all the way from idle to 1Bar?? What controls the mixture- the MAF input to the ecu right? So we already changed the MAF maybe its a wiring problem from MAF to ecu. All along I've been looking at why No3 is always the one to go but Conrad is spot on with the fuel rail problem always making No3 leaner than the others so if the whole engine is lean it will always be No3 thatwill go first right?
Chris

Botswana WRX

I should have said that MAF sensor lets the ecu know the effective throttle position etc then with reference to the Lambda sensors the ecu sets either 13.something AFR for idle/cruise and 11.something for boost with variations in between. Anyway thats wot I think is supposed to happen, am I correct?
Chris

Bob Rawle

Seems to me that you have the right idea when thinking its fuel pump related, it probably is BUT not the pump but the fuel pump controller, the oem ecu sets the pump duty at either 33, 66 or 100% dependant on engine load, if the controller is stuck in 33% mode then you would get the results described.

To prove this suggest hard wiring the pump directly to bypass the controller ie connect a switched positive feed in parrallel to the existing cct, that will run the pump at full chat all the time, it will over fuel abit on lighter loads but should fix the heavy load problem, if it does then replace the fuel pump controller and remove the bypass.

cheers

bob

Bob Rawle

BTW on these cars No 3 is not leanest, no 1 is however no3 gets extra heat from turbo and at that afr would be sure to suffer.

FPR is adjacent no 1 not no 3 ie last in line is no 1.

Add to which if the ecu pulls timing big time over this then egt's will rise due to the retard in any case.

bob

Botswana WRX

Thanks for the info Bob, We do see the fuel pressure rising on the Dyno runs as it comes on boost and with the current uprated pump its up to 6bar so I don't think it can be the controller. Its just that FAR stays at 13.7ish instead of 11.5. and we have tried a new original ecu with no change.
Chris

Botswana WRX

OK here's another question, can anybody tell me exactly how the mixture is controlled. Does the M Absolute Pressure control the fuel pump (via pump controller and ecu of course) and the MAF command richer with higher Airflow or is it the other way round or maybe a combination of both? With the Lambda sensor(s) providing the feedback and closing the loop?? Exactly how does the ecu know to command 11.5 AFR when on load and boosting and 13.7 when off serious load.
Chris

johnfelstead

why were you using a unichip when you can get far better control of the engine using EcuTek?

You need someone on that car who knows their stuff, doing live datalogs to see where the problems are. You could have a completely shot or badly remapped ECU.

Do you have a copy of deltadash to do some logging with?

You say you use 93RON, dont you mean 93PON? which is about 97RON.

Botswana WRX

Unichip is available here, and standard gas station fuel on the High Veld is 93RON thats Research Octane Number, non turbos are quite happy as its 6000ft up and the air is 20% lower pressure than sea level. We are now quite good at octane boosting with NF, Toluene, Benzine etc. and we can get 102.3 unleaded race fuel. However a working SA spec STi will run on 93RON without damage, its retarded and low boost but will run OK. But your missing the point a little, Its now STANDARD, no chip no pipe and bog standard oem ECU but still runs lean on boost Why?? I think it has since it was built. Rallystar have chipped 20+ STi's for these conditions without any problems and they do have all the gear,He prepares just about all of the SA rally cars. Just can't find why it runs lean standard and when mapped fully rich with the unichip it still runs lean on boost but why??? And as Indicated above in current standard form we have swapped everything from a working STi and even put the sensors off mine on another on the RR and its been fine with 11.3AFR. So its wierd.

johnfelstead

have you swapped ECU's yet?

Botswana WRX

We tried 2 standard ecus one new from Subaru one off another my02 STi just the same!! 13.7 on Boost.
Chris

johnfelstead

you havnt melted the wiring from the MAF in an engine fire/ let it touch the exhaust incident have you?

Botswana WRX

No, but I guess African rodents could be a problem, what exactly is input to the ecu use to control the mixture apart from lambda to close the loop and injector pulse length and fuel pressure. Is it MAP to match the fuel pressure to maintain differential fuel to Manifold pressure and just MAF to decide if its Idle, cruise or boost, If thats the case I will try tempory wires between MAF and ecu. We may be getting somewhere, thanks.
Chris