High VE and strange Injector flow numbers

Hey all,

I've recently finished the understanding EFI, and Practical Standalone + AFR courses, and am starting to tackle the base map on my new build (After doing the engine building courses of course, all fantastic).

I'll give you a rundown on the setup, then the questions:

Link G4+ WRXLink V1-3 plugin.

EJ20G Closed deck, 1mm oversized bores, forged etc.

Ported + polished heads with 282 cams, non AVCS heads (v3 STI, Shim under bucket).

VF22 turbo. stock BCS, with Link 4 Bar MAP sensor

Bosch 720cc injectors (brand new this past week, from a reputable NZ motorsports retailer), FPR set to 3 Bar.

Link CanLambda

A brief log & the tune file are attached.

This is a brand new map, setup from scratch after following the HPA courses, the previous tunes were done by someone else on a traditional fuel model, so I wanted to start again with the new injectors and setup, and using the modeled fuel model.

The questions:

With the base FPR set to 3 Bar, and the injector dead-time enter that came with the injectors, and set to 728cc, it runs stupidly lean at idle, with anything under 80 entered in the idle area in the fuel table, this seemed odd to me, considering everything I've understood in the course.

Setting the injectors to 550cc (as per the attached tune), I can get it to idle around 60-70 in the table.

as it enters low boost, it's already nearing 90 or so. Does this seem right to any of you?

This is my first time putting these courses into practice, so I just wanted to sanity check.

Should I set the injectors to their actual cc, and just run really high VE numbers? I'm unsure if the large cams would have anything to do with this, as most of the info I've found online has been for more standard flowing heads.

I do have a FP sensor setup in the ECU, but had it disconnected during this log file, with a gauge installed to sanity check the pressure.

The store I purchased these injectors from are known to be very reputable, so I'm not suspecting fake injectors.

Thanks, and I appreciate any help.

I had a quick look over your tune and most of the fundamentals seem correct. I would be leaning towards maybe the deadtimes not being accurate. Supplied deadtimes are not always relevant depending on the driver used to generate them.

A quick reality check you can do is as follows:

Set the lambda target table to 1.0, run the engine at fast idle around say 2000rpm (just set idle target to 2000), then adjust all fuel table cells around the idle area so that the measured lambda is 1.0. Then change the lambda target to 0.9 and if the deadtimes are correct the measured lambda should also change to near 0.9. Try the same test in the other direction - say 1.05 lambda. If it under or overshoots the target change then adjust the whole deadtime table up or down and try again.

Hey Adam,

Thanks for that, much appreciated.

I'll give it a go this week and let you know how I get on.

Also glad to hear the fundamentals are mostly there, it's a good sanity check!

Thank you.

Hi, I have same problem. So problem was with deadtimes? Thanks!