RomRaider

Fresh copy worked, only thing that happened was a crash when I used my scroll wheel. I never saved over my original.

Now I see that dip I felt... Not sure where that came from, boost was steady 11psi. When I was done I immediately checked learning view, IAM 16, no learned knock.

I forgot to log rpm this morning. DOH! I was using the new "engineering" params too. Damnit.

I only read the first page of this thread, so if what I say has already been added, sorry...

Be sure you understand what injector latency is. This is the time it takes for the valve armature to stroke from the closed position to the open position. Voltage, inductance, line pressure, and backpressure all have bearing on this time constant.

If you increase fuel pressure, it will take longer to open the valve, but the valve will shut quicker. These are pressure assisted shut valve configurations.

Modding your injectors won't affect the time it takes to open the valve, but it will cause the valve to shut quicker because you are decreasing the back pressure.

Backpressure? If you do not change out the FPR, the pressure at the rails will be the same. A fuel injector just opens and closes. If the same pressure is in the rail, then it will take the same time to open and close. The only difference in the injector is the cap. Subaru uses the same exact latency for the same type of injector. That is my argument.

I think the backpressure he is citing is between the cap and the actuator. Technically if a device flows X and a downstream device flows Y, device X experiences a backpressure and the total volume of the combined devices is no greater than device X.

But I really don't think there is enough volume between the cap and actuator to really make this a concern. I would think with such a small space and lack of room to expand, any such backpressure would be gone in well under enough time for anyone to need to worry about.

Ahh, I see.

It should not effect latency however. Proof is in the stock ROMs (compare an 04 WRX to a v7 or v8 JDM STI that has pinks).

I was refering to the pressure drop between the actuator and the injector cap. Theoretically, this would change latency a bit, but I imagine its so small its neglegable...

Anyhow, I need some help with this latency vs MAF scaling at the moment. I just installed modded injectors (I had Witchhunter mod them for me so I got flow data and clean injectors).

According to their flow data, these should be roughly 825 cc/min @43.5 psi injectors. Well, I set my injector flow scalar at 825, and it ran super lean (18/19:1 AFR, verified by my front o2 and my wideband). Only messing with the injector scalar, I had to knock it down to 550 cc/min to get it to idle around 15:1 and run around 14.5-15:1 at lower loads.

But when I got on the throttle, my AFR's crashed well below 10:1 (that is as low as my wideband reads, I'm sure it was much richer than that because the motor was stumbling). Before, I was running a steady 10.8-11:1 at wot and higher load. All I have changed recently is the injectors. I'm assuming I need to mess with injector latencies or tip-in.

It would stand to reason that the latencies shouldn't change. But at the same time, I'm running the stock MAF, so I don't believe I should be messing with the MAF calibrations.

Help? What have others experienced? It would be nice if someone has changed there latencies and you can provide that info...

You kinda missed my point. In order to get the injectors to work correctly, you must scale the MAF. You can increase latency and get it all lined up, but that is not the way Subaru does it is what I was saying...

No, I got it. And I found my anwsers on the latency. What follows has to do with modded stock injectors...

I would have to disagree with you though on your view of the MAF scaling. Ultimately, either way you can get to the solution. My hang up with messing with the MAF scaling is that you are changing the injectors, so why wouldn't you change the parameters that pertain to fuel? Your MAF is calibrated, and fairly accurate, Why mess with something that is already dialed in?

I say change the latency, because the problem is that with very short pulsewidths on the injector, it has so little time to flow fuel that it appears like a stock injector. Once you get into higher pulsewidths, then it appears to be acting like a 820 cc/min injector. Just bump the latency up to compensate for the low end, and at the high end, the latency is such a small time period in the pulsewidth event that you can nearly ignore it.

Either way will ultimately work, but for me, messing with the latency is far easier and seems more correct. Messing with the MAF just seems to be messing with the wrong variable.

My 2 cents...

The problem lies in that once your load drops in the higher rpm, latency becomes a larger percentage of PW, so you must compensate in your o/l fuleing map... Modify the maf scaling and you get it "more" inline and this is the way subaru does it...

Either way works, its just the correct way is to input the real world latency IMHO...

If the MAF scaling is too high, and the injector flow settings are correspondingly too high - so the fuel trims are all small - what sort of side-effects would this cause? Or, put another way, what sort of driving/logging tests could be done to indicate that the MAF & injector settings are off?

I ask because I suspect my car may be in that state right now, but if I can't demonstrate any negative side-effects. Or any side-effects at all, really.

I have tuned a few cars with modded stockers and did adjust the latencies. All the cars idle rock steady, have no driveability issues, and AFRs perfectly match the table. No MAf scaling and only minor adjusting of the fuel table (During spoolup). Long term fuel trims after 6 months are less than 2% across the board.

I see where Gabe is coming from but I think either way works. Rescaling the MAF does require also rescaling the engine load axis on the load-based tables (i.e timing, fuel, etc).

That is the same way I tune cars with aftermarket injectors (keep the MAF scale stock and modify latency). Usually they won't have a bigMAF or anything like that, or it is already scaled. MAF scaling is a PITA and I try and avoid it. On my car however, I had to scale the MAF anyway because I changed out all of the piping. I figured I mine as well use "real world" latency.

hi gabedude
ive removed my fuel regulator from my car as i could not get latency and fuel scaleing in line with the aftermarket induction kit.
ive just about finished scaleing the cl maf, do you have an example of what figures i could change my ol fueling table too so it makes it easier to tune ol maf?

thanks
neil

Flat 10.5:1 AFR above 1.6 load is what I use. After the scaling is done, then I dial in the curve. This takes care of any interpolation the ECU would do.