RomRaider

I always had a problem with this idea of changing injector dead times to bring fuel trims inline. I had my reasons, but wanted some solid evidence. Of course I have tuned cars both ways and some responded just fine to minute latency adjustments to bring trims inline without touching the MAF scale. Ok so here is my reasoning on using the "correct" advertised latency for injectors.

Assuming stock fuel pressure (increasing fp will increase real dt):

1. Look at Subaru injectors of the same type and different flow ratings. Take a 2003 JDM STI for example (pinks) and compare the latency values to a regular WRX. They are the same. What differs is the MAF scaling. This goes for any stock rom where the injectors are of the same "type" but have different caps on them to flow more.

2. The STI has a higher flowing pump but fuel pressure is the same. If you do not change fuel pressure real world dead time remains the same.

3. At low rpm, dt becomes more of the total PW. At peak TQ/EL, dt takes up the least amount of total PW. As RPMs rise though, PW decreases as EL drops so dt is more of total PW again. What does this mean for dt? Well, if you have too much dt, you may be ok at idle, but you will end up running rich as the rpms rise and EL drops. Moreso, the fueling will be all over the place in other areas too.

I have always studied the OE mapping to get a good idea of what these engineers are doing and why they tune things a certain way. The fact is dead time is dead time and part of the injector and fuel pressure system (it has nothing to do with what cap is on the injector). You don't know what dt your car is going to need when the battery is almost dead as well.

So, what do you do? Modify the MAF scaling, leave dt the same as the type of injector you bought. This is what Subaru does to their factory cars.

This was the only way I was able to hit fueling targets and have the car run nice and smooth. I used 2004 WRX latency as I have that type of injector (same latency and type as pinks as well).

I never agreed that a 2X4 table (based on battery voltage, so only one cell is really being used) was good enough to bring trims inline and not throw fueling off in other areas.

I have had a problem with the previous explanation as well. To me, injector latency is the delay from when the ECU sends the 'fire' signal to the injector to when the injector actually fires. If the latency value is too low the injecotr will fire too late and vice versa.

I had a terrible tip in stumble that is getting better and better as I raise the latency. I run what I was told was Helix 660s(but mine look identical to PE650s). I used the latency for Helix injectors and have a stumble. I increased the latency to PE650 specs (1.18@14v for Helix vs. 1.5 for PE) and the stumble is gone at initial tip-in from idle.

I think we should start a database of recommended values for various injectors. It is a topic that is constantly revisited with very little consensus.

Assumptions would have to be made, like perfect MAF scaling, stock fuel pressure, good battery voltage, and so on. But I think this is one of those "right off the bat" topics that is gray and fuzzy and it ends up scaring people off.

Even then, my point is that the DW750's that I have tuned all behaved differently on different cars. Adjusting latency was what I did on most and left MAF scaling alone. Of course I am splitting hairs here on my own car, but the principle is the same. DT is DT if fuel pressure is the same regardless of whether or not the injector has a different cap on it. The injector scaler is easy enough to adjust, so what I would do is do that first then go and bump latency up. This always made the high airflow (at high rpm) fueling shift richer than the target AFR after increasing DT. It is because it is using the wrong DT. I am only talking about 0.5 AFR points, but hey, I am picky.

Keep in mind as well that MAF housings vary on different makes/models/markets. I think a database would be great but will only get you close. You must get the latency close and then tweaking the MAF scaling will help dial it in further. I would suggest this database include the MAF housing information, as well as airbox configuration.

forum31.html

You kept telling me to just modify the MAF scaling so I went back and looked at several tunes I did with DW750's. Each and everyone, I did the same way. I did not touch latency (these were all 16 bit cars) and got the scaler dialed in so the line was a perfect 11:1. After that, I adjusted latency for trims and idle. Then I tuned the car. On every one, they would deviate about 0.5 AFR points rich as airflow increased. I just leaned the O/L fueling map out or left it flat because I like to richen up towards redline anyway. On my car though, I want it perfect, just like stock so my targets match. So on the light blue DW injectors, the best latency is stock 02-05 WRX values.

Thanks for the tip on the MAF scale Braden. It was really pissing me off. We need to get dinner sometime together.

-Gabe

Cameron is in Austin right now and we're meeting up tonight at Iron Works if you aren't too busy come on by. I almost always dial in a little MAF scaling even after fussing with latency forever. Most of the time it's minute adjustments in the A-D range, and some larger adjustments at the top of the scale. On larger turbos I will extend the MAF out to 4.92v and smooth the curve back down to allow headroom and fine tuning.

I doubt I will be able to come out tonight, but it would be cool to see Cameron again. My upper MAF scaling is pretty close with stock values, it may just need a little minor tweaking. I doubt I'll be pushing over 350 G/S on the 20G, but ya never know. I hit 300 already at 16.5 PSI @ 6k rpm.

but what hapens if your running sti pinks at 3.75 bar, what would this be scaled to and more import what would it do to the latency?

Upping the fuel pressure makes it so the injector takes a longer time to open and close so dtime has to increase. That was what prompted the original discussion by John. He upped his FP and was able to bring everything inline with only latency and scaler. I am saying that latency running the stock FPR should be the same as the "type" of injector. Subaru does this on their cars with the same "types" running the same fuel pressure. What they change is the MAF scaling (the whole table differs in G/S values).

Ideally, you would take that injector on some test bench and apply that amount of FP and figure out latency. Since that is not realistic, just up it a bit if you are having issues. Upping latency too much will make you go rich when load drops, but rpms increase as PW is a direct relationship to EL. When you add a large static offset like latency that is incorrect, it shifts rich in the upper rpms.

ive upped my fuel pressure to 3.5 bar and i used romraider injector scaler tool which varies on how you actually drive the car, anyway i rescaled my injectors to 615cc and used the latency it told me to use which was about .10ms on top of stock values(first value from stock). when i flashed it to the car it took a while for the iam to come upto 16, but when it finaly did i had no knock what so ever as i was getting it before scaling the injectors due to overfueling.
but then on a long drive down the motorway whilst it was chucking it down with rain my iam went down to 14 with no knock what so ever? so i reflashed the latency times back to stock and the iam came up straight away but know im getting knock again.
so now im stuck and dont know what to do, could it be that my induction kit sucked in some of the rain and caused this to happen or is there something else that could cause this to happen.
sorry for the high jack
thanks

I doubt adding latency caused knock. Latency just adds extra fuel (static IPW offset).

sorry if i was not clear in the last post
adding latency didnt cause knock it removed knock but when the iam dropped i put the latency times back to stock (removed latency)which then caused knock.

i think ive jumped into this tuning thing feet first as ive done lots of mods before fully learning how to map the car with the std parts

Adding a ton of parts, especially intake hoses and injectors at the same time makes it very hard to tune for. You will have to add some latency in so the fueling is correct, then modify your MAF scaling for the rest. It is very hard to tune a car so it matches the O/L fueling map when many things have changed. A quick and dirty though is just to modify your OL fueling map if fueling is off. The O/L fueling map actually stores enrichment value offset from stoich.