hrubago

My spec:
Haltech ECU
TD05/20G turbo
USDM STi 2,5 short engine
Sti 3 heads
550 injectors
custom made FMIC

I mapped my engine till 1bar. (not above because at higher pressures the FMIC piping were too unstable). I work with Motec lambda meter and seems to be enough fuelled (Lambda = 0,88 at 0,9 bar). Higher pressure maps (over 1bar) I did linear aproximation from lower pressure maps. Car goes very quick andwith no power drops.

But the problem is: I checked maps 0,9-1,2 bars on the datalog with serial subaru lambdasenzor. There is lean mixture after 2 second with stable pressure and stable fuelling values. I dont understand.
Possible problems:
1) bad fuel pump
2) bad fuel pressure regulator (bad value for 550 injectors)
3) bad lambdameter (maybe bigger temperature means missreading)
4) engine needs a lot more fuel in over 1 bar pressure.

Bob Rawle

0.88 lambda measured with a Motec PLM is VERY VERY lean and you should richen it up considerably. Don't rely on the factory fit lambda, you have a Motec PLM so you should map it with that.

bob

hrubago

Thanks. But i found recomended values from haltech:

0,2-0,4
0-0,2
vacuum 0,3-0,2 (-4,5-(-3) - lambda=1,01
vacuum 0,2-0,1 (-3-(-1,45) - lambda=1
vacuum 0,1-0 (-1,45-0) - lambda=0,98
0-0,2 bar (0-3PSI) - lambda= 0,92
0,2-0,4 bar (3-6 PSI) - lambda = 0,89
0,4-1 bar (6-14,5PSI) - lambda= 0,88
1-1,6 bar (14,5 - 23 PSI) - lambda= 0.86

It means im in the tolerance range

Have you another better values?

Which is your opinion (recomended) fuel pressure and fuel pump?

hrubago

Motec PLM is not mine. It has my friend from SLovakia (400 kms distance...... :-(

TopBanana

Were those figures intended for a turbo engine? You need to keep the mixture rich to keep the heat down. I'm not an expert but I've read about people mapping to 10:1 (lambda 0.68) under lots of boost on Subarus.

hrubago

I assume that pressure 1-1,6 bars are not atmospheric engine value :-)

But, where is good source for information which lambda at which pressure?

TopBanana

Doh, I haven't had my first coffee yet!

I'm sure the usual suspects will be along shortly to answer your question...

hrubago

Second question is why the A/F falls down when the pressure and fuelling are the same at time (10 seconds?). Maybe Lambda is overheated?

carlos_hiraoka

Does values are waaaaaay to lean, 0.88 lambda which equals 12.9:1 AFR @ 1 bar of boost will definetly "eat" your pistons .
AFAIK the EJ257 has very tight clearances, u should be aiming at least at 0.78 lambda @ 1bar.

Carlos H.

hrubago

I think clearance is 0,02mm same as in 2,0 STi euro version.
I believe low expansion type pistons (hypereutic) are better than euro spec pistons. Also believe that 0,88 at 1 bar is not enough.

hrubago

And where is the source of most popular top tuners recomended fuel mixture? :-)

Bob Rawle

hrubago, as Carlos says ... I've already said its too lean, you will nip a piston and bye bye big ends as a minimum. You need to aim far richer than that. Carlos has given you a good starting point. You need to travel the 400kms and use the Motec to be sure.

0.68 is too rich and wil choke the engine.

bob

Andy.F

Depending on your fuel pressure, even 1 bar on a 20G fitted to a 2.5 this could be maxing out your 550 injectors.
What is your fuel pressure when boost = 1bar and what are your injector dutys at 5-6k rpm ?

I'm with Carlos and Bob on this one, I think you want a maximum of 0.78 Lamda and a minimum of 880mv as an INDICATION on a hot std sensor.

Andy

hrubago

Dutycycle around 80%. (max I have seen was 96% in 6000 in 1,2bar mode)
Fuel pressure (i think) standard 2,5bar (Sti3).
I have to improve little bit to 3,2-3,3 (it could be 10% more fuel than original setup).

My target is safe maps at maximum boost 1,2-1,3 bars.

hrubago

I have no detonation in all maps

hrubago

But:
Please could you suggest me mixture in these pressure points?
-0,3 bar
0
0,3
0,6
0,9
1,2

Thanks a lot

carlos_hiraoka

ok but please only take this as a "guide" .....

<0 bar : lambda = 1 upto 3500 rpm and utpo 0.92 at higher rpms
>0 bar : 0.86
>0.5 bar : 0.82
>1 bar : at least 0.78

even if there is no det, watch out for your EGTs

Carlos H.

hrubago

Iam not starting ECU mapping business :-) like beginner and not hunting for informations how to do it.
I saw few destroyed engines (3rd piston overheating) mapped at same mapper and therefore need verify all that claim what he declared because I dont want to be next one....
Thanks for your help.

TopBanana

Has it now been established that the parallel fuel rail mod fixes the piston #3 fuel pressure problem? You might want to look into that

Andy.F

Don't know that the fuel rail mod fixes it but it is one less variable so should improve the balance.
There is an even bigger mismatch in the airflow per cylinder, there have been a number of bench checks done on this.
I'll be monitoring individual cylinder EGT's and AFR's on my Sti5 engine under live running conditions in the car and I'll feedback on my findings.
The Apexi Power FC ECU will be useful here as it has individual cylinder fuel trim. It would be nice to also have individual cylinder timing control but that belongs to much more expensive ECU's !

Andy

carlos_hiraoka

Andy, take a look at this thread, very very interesting:

http://forums.nasioc.com/forums/show...t&pagenumber=1

Carlos H.

dowser

Andy - you'll do both? I'm getting my 2.5 headers made up with EGT boss on each runner plus the one in the up-pipe. Plan is to map to det point using the up-pipe for safe EGT readings, then check each runner individually for differences, and trim using these readings.

I could fit lambda bosses and use an adaptor for the egt I suppose....but I didn't think the lambda sensor will like header temps?

Richard

Bob Rawle

Richard yes they will if you use the right ones, mine did !! I balanced the cylinders first then removed and mapped from the position after the turbo to get "real" readings.

bob

Bob Rawle

The thing is how do you know the temps you see are safe, egt will depend on mod state just as much as ign/fuel so a corrctly mapped car for ign/fuel with no det running high level mods could be much safer than a car with too much adv and bad fueling running no mods and lower egt's for example.

bob

dowser

Thanks Bob.

I'm assuming that I'll know whether the engine is 'happy' using consolidated EGT and d/p WB along with det monitoring. My stock ej20 has (so far ) survived my learning curve over the last 2 years pretty well.....and I've made some monumental mistakes. All limited to knock, boost and AFR monitoring....plus TB temps. I'll add up-pipe EGT and oil temp/pressure with the ej25, plus temporary use of EGBP an gauge. The EGT per cylinder is intended to fine tune for max benefit.

My builder is absolutely insisting on EGBP monitoring on the new engine ('absolutely' because of my wish to use an 05 hot side intially). Along with EGT, these 2 seem to be his critical benchmarks - but I think it's only 'cos he's used to not being able to play with the ECU.

Digressing - while talking of monumental mistakes, I blew the actuator connection off my TD05 the other day (don't use pliers to remove a stubborn pipe on a faulty actuator ). In 4th on a 6MT it managed 1.7 bar peak and 1.3 at redline - do I have a problem with the turbo, or is this normal EGBP related limits?

Richard

Andy.F

Richard I'm using the same tapping points for the EGT and WB lamda. In order to run the WB unpressurised there will be a 1.5mm feed and a 5mm exit from the sensor chamber, this will lose a little exhaust gas but nothing that will make a difference to the fuel balance.

I would have expected your boost to run higher but the 'blow open' pressure depends on your actuator spring and wastegate orifice area ?

Andy

dowser

Thanks Andy.

Standard TD05 actuator and WG (not ported).

Richard

Andy.F

Sounds a bit low, normally they go 'off the guage' midrange, I have seen 2.3 bar on my TD05 with the chargecooler which increased the size of the C/C a little
I did however run with the actuator rod well tightened in so your result is probably ok.

Andy