Port injector scaling
 

I am wondering if anyone can help nail down what the 2 values defined for port injector scaling are. They both seem to be used to calculate pulse width in the ROM disassembly, I have found that the "Injector Flow Scaling BRZ" parameter affects final fuelling a lot more.

Stock values in the ROM:
"Injector Flow Scaling" = 12176.55 (this is scaled by 2707090/x to get approx 222.32 cc/min)
"Injector Flow Scaling BRZ" = 263.10 (this is scaled by 60000/x to get approx 228.05 cc/min)

If anyone has upgraded their port injectors and has appropriately scaled them in balance with DI, can they let us know:

a. Nominal Injector size
b. Which tables were scaled
c. What the final value in the ROM reads (or nominal value and scaling used)

If it's any help, in ECUtek there is only 1 PI scalar value which ties in with the Injector flow scaling BRZ value. I get the impression the other is a port over from older ROMs.

Probably not much help but only the "Injector Scaling BRZ" value is changed in the OFT E85 maps by aprox 30% over petrol maps, suggests it alters scaling of both pi/di injectors as they did not appear to alter the pi/di ratios.

think @D-VO is doing fi on opensource he may have changed injectors

just guesiing though :iono:

I'm fairly certain the injector scaling brz parameter only affects the PI. You need to adjust the DI fuel pressure compensation tables to adjust flow for the DI.

With EcuTek, we just create a custom map to increase total fuel by x%. For flex fuel we create a custom map that increases total fuel by x% based on ethanol content, along with custom maps to adjust timing, AFRs, and other items, also based on ethanol content for a dynamic tune that always adjusts in real-time for the actual ethanol content in the fuel, which can vary greatly even with "E85".

- Bob

Shiv's OFT E85 maps adjust PI with the "Injector Scaling BRZ" value and DI using GDI Pressure Multiplier A and B.

Looks like there is a lot of redundant code from old Subaru implementations, but how much is used, or not used in different subroutines, we won't know. The ROM is also difficult to untangle due to a lot of indirect referencing going on. The MAP sensor scale is in there 3 times, all referenced by different bits of the program, the Front O2 sensor scaling is in there 2 times.

I changed both flow scaling tables just to eliminate the trial and error aspect.

Slightly off topic but it strikes me this may be an issue for those of us that have been working on our MAF scale with Romraider/OFT. Are the different parts of the program significant ie. one for fuel mapping and the other for ignition?

Or one for normal running, one for cold start or tipin... Or just there to fake people out and not used at all.

I also wonder about the front O2 sensor scaling - it is in the ROM twice, but some people rescale one of them, not having a definition for the other. A lot of people have commented on the inaccuracy and unreliability of the front wideband at the rich end.

If anyone has a wideband installed (I don't yet), I've got some scalings for the stock sensor that I'd like tested -pm me.

VDR only. =) Base on VDR was 147hp. I'm at 290 - 310 depending on the barometer.

only one is needed for scaling. All of my personally developed E90 roms use the BRZ one. I exchanged mails with TD-D way back when the open source defs first came out about the originally defined flow scaler not working correctly so he dug up the second one that is actually used.

Cliff's notes,
Use "injector flow scaling BRZ" to adjust the port fueling.

is this mean ecutek could not work easily with DI?

Is it just me or is the value for port scaling in ecuflash different than the value ecutek? open the same ROM in both and the number is not the same. I thought is was just the OFT ROMs but I opened up the stock ROM and found that it is not the same as will. If I open the OFT e85 tune in ecuflash the port scaling is 350.67. That same ROM opened in ecutek it is 171. Also the older A01C ROM doesn't have BRZ scaling at all.

The 350.67 value displayed will be the raw ROM value.

RomRaider definitions scale as 60000/x which approximate ml/min. If the number shown in RomRaider corellates to ml/min (171 ml/min), then x corellates to msec/ml (350.67 msec/ml).

After a bit of fiddling around, I dialled in my DW450 injectors to match my DI scaling before turbo install.
Latency was quite a bit higher than advertised on the DW site, I changed DI:PI ratios under 0.3 g/rev load to full DI - I was getting DI:PI switchover where PI operation was under the linear range giving me all sorts of hesitation issues.
My final settings as below: