Problems with scaling P2s
 

I'm currently remote tuning an NA engine in preparation for FI . The owner has fitted yellow uncapped injectors in the P2 position . These should flow around 750ccs according to the data from the shop that flow tested them.
However if you look at what happened when I scaled the injectors to accomodate 750cc you can see that it went very lean .
To get back to where the stock injectors were I had to scale them by only 12% .

Only reason I can think of for the low flow is that one of the injectors is not working at all . If this were the case how would we go about testing this ?
Can anyone else see any other possibilities ?

Ignore the 8000rpm result - that is another issue to be resolved .

https://www.rx8club.com/attachment.p...1&d=1293744527

Something else is going on with this......you can see where the P2 are effecting the Lamba # in the bottom 3 plots......In the 7000 and above area.

Adjusting the scaling effected the whole thing with the 97% scale...so something else did that :)

Maybe force it into open loop and try the logs again...cause something is very strange

BTW - before 5500 all the logs are pretty much identical . IE P2s cut in at 5500 in NA under load .

Does the 1 injector not working theory seem viable ?

I would doubt it......wouldn't it have made it lean on all of the logs since the re/re? not just on the last one? Unless it crapped out between logs :)

Get him to check the plugs....and re-run with the O2 unhooked and see what you get??

But we can't log without the o2 sensor ?


It only went very slightly rich after the p2s were changed . I would have expected it to go VERY rich . Then when I scaled to 97% it went VERY lean when I would have expected it to be close to what it ran before the swap .

:arcadefre

yep - tried that , then did this :wallbash: so for that suggestion you get :greenchai

You have several issues overlapping each other.

Agreed .
I think I have figured out the 8000 rpm thing .
The fact that stock was not quite aligned with target is minor .
The big issue is the one I asked about .

Well....you really don't have a lot of options...you either believe one thing is valid or the other :)

If you don't know that the injectors are working I guess that's a good place to confirm before you go much further

Not a common failure though.....loose plug would be more likely...or the wire got pulled out when the injector was removed and there is poor contact






What does it run like? You would think that it would run like crap with one rotor very lean...and the other really rich????

For the 97% it ran crap to 8000 then was ok . All the other logs it ran fine (except above 8000 when P2s were first put in and no scaling done)

Go back young man...to a known good map/ configuration :) And see if you can figure it out :)


The modded P2's that I have flowed close to 800 cc for what thats worth :)

fuel pressure?
OD

You think his fuel pressure suddenly went up?

Can't you kick it into OL by setting the load to 0% or setting the rpm switch off to like 1k rpm?

perhaps I need to explain a little better

log 1 : stock P2s - all good but slightly lean of target .
Log 2 : P2s changed out for 750cc uncapped yellows : AFRs go slightly richer . Nowhere near as rich as expected .
Log 3 : P2s Scaled up by 97% . AFRs go horribly lean
Log4 : P2s rescaled to 12% more than stock . Afrs go back close to what they were with stock yellow P2s .

Some math ......

280x2+380x2+380x2 = 2080cc Total

280x2+380x2+750x2 = 2820cc Total


total fueling has increased by 35%

So fitting the 750cc injectors should have produced Lambda of 0.64ish - instead we got 0.85ish

280x2+380x2+750x1 = 2070cc Total

If only one 750 injector was working the total fueling figure is much the same as before the swap was made . So Lambda would be similar Which is close to what actuually happened .

So my best guess is that only one injector is working and flowing a little more than 750cc s.

I think that you have overestimated the contribution of the P2s in an NA application.
What was the load in the aforementioned test?

Probably - have to start somewhere though .

100% load on all runs

You were at 100% load above 6800 RPM? What was the airflow?

yes 100%

Flow :
200 @7000
202@7500
219@8000

i didnt explain myself well--i need more lessons from Swoope:)

The fuel pump does go to the high out put mode at a certain tps %/load, like you were referring too--ifs there is a sudden change in pressure wouldnt bigger injectors respond in a stronger manner? I know the fp regulator comes into play here but if its low to begin with.....?
Really just guessing--but dont you need to know for a fact the fp pressures you are dealing with before any scaling is done? Or does that matter?
OD

Fuel Pressure does matter.

However, it is an expensive mod to an NA vehicle, so I could see where you wouldn't want to do it.

In that graph, add in duty cycle....if your in transition, funky things can happen if your scaling or latency is off. Or hell, for that matter, the Fuel VE% map can mess with it too.

Without further information I fail to see the problem here. You figured out a scaling that worked. It wasn't what a simple calculation said it should be, but so what? It sounds like there are no other signs of anything else being wrong.

Here's an example. I was scaling some 740cc injectors on a 2005 STi not too long ago (stock size is 550) with Cobb AP. It was a matched set from Deatchwerks. I had flow test and latency data from the manufacturer and another set of numbers to try from Cobb. I put the data into the AP. With both set of numbers it ran like shit--stupid rich. I ended up taking the stock scaling and made an across-the-board correction to the latency and all the trims fell in line.

Even though they didn't behave as I expected, there was nothing wrong with the injectors. The turbo was upgraded and the boost increased. The car runs fine and the trims are all within an acceptable range.

Wondering about this latency thing and what effect uncapping the stock injectors has on it . Ayone else had scaling issues after fitting the uncapped P2s ?

Latency will change for sure......I think that Paul Yaw might have the data..he seems to have the data for most injector combinations :)