Definitely try it out.. I think that 1C entry is just the same calculated load that you would see in OBD-II also..

Knowing the current formula correlates perfectly for values at low boost before it clips..

Thats why I want to see something that could go over that magic number.. if there is a value that correlates at higher load (if it starts at 0 again or some other value and continues to read something and not stop at 255 and stay there, at least it can be calculated for)

 

It won't matter if its labeled incorrectly anyway... your correlating values mathematically, hitting 255 places the value at its highest point, regardless of the axis.. The secondary calculated load value that we have been using correlates to the current formula and value.. But when I take the value 255 and put it in there, I get the max value it will calculate.. Its not going to matter if its Load site value, Grams/rev etc..

If the axis in ECUEdit was in Grams/Rev and your formula calculated Grams/Rev, it would still hit a max when that value hits 255.. UNLESS there is more than one request ID represeting an overflow byte, or some sort of indicator that the map shifted.. But logically if your giving two bytes.. once you hit 256, your two bytes would then be 01 00, 01 01, 01 02.. all the way to FF FF (65535)

 

I went out and logged a few sessions. Like chrisw's results, the ECU load value (1C) compared very well to the calculated load value at constant speed (off by less than 1 unit), but during accel, it was off by as much as 10. I suspect its due to the IPW, AFRMAP, and Battery values being scattered around in the EvoScan "Data.xml" file. I've rearranged those items to all be in sequence, and I'll log a few sessions tomorrow on the way to work. I'm calculating load from

load = 5*injscaling*(IPW-IL)/AFRMAP where

IL = -0.1026*V+1.8741

 

Have you guys actually gotten to load sites over 150 yet? I think that is where it may stop working..

 

Yeah, the raw value, "x" maxes out 255. My first run was WOT 3rd gear from 2500 rpm to 7000 rpm, and the raw load value was 255 for the entire run. :-)

 

I kinda figured that, I also think the value doesn't update frequently either..

 

In the new logger I'm working on, I won't need the log converter tool anymore, but I plan on using this formula, only it will actually have all the injector voltage latency values and it will correctly interpolate between them.. This was actually something that began to concern me when I was seeing other peoples logs and noticing the voltage was different than on my car, which was always consistently 13.8-13.9v, where I see others at 14.3, but goes down to 13.5 or so in some.. I can see this affecting things in some circumstances on some cars.

 

What does IPW stand for? Injector pulse width?

Is AFRMAP simply the AFR?

 

IPW is the logged injector pulse width.. AFRMAP is not the actual AFR, but the value in the fuel table cell its on at that moment.. The actual AFR will likely be different..

Its really just a method of calculating load from available data where you might not be able to otherwise.

 

gottcha...thanks.

 

I logged some more load data on the way to work this morning. Again, during constant load, calculated load agreed well (within 1 unit) with ECU load, but during accel, calculated load exceeded ECU load by as many as 15 units. This is with ECU load being read by EvoScan directly after IPW, AFRMAP, and battery voltage. I'm going to put ECU load first in the lineup, and log some more data on the way home from work.

 

But are you getting any situations where the ECU load is over 150? Mathematically the value doesn't seem to work over 159 if you use that formula... Under boost my car is rarely under the 200 load site.. And this wouldn't seem to work.. I'm going to test it the next time I get to take my car out..

 

What I've said so far applies to ECU loads up to 150 (below 159.375). When I rearranged the logging sequence to ECU load, IPW, AFRMAP, and Battery V, then calculated load did a pretty good job of tracking ECU load. During steady loads, calculated load was usually a few units below ECU load, and then during rapid load change under accel, calculated load was generally either equal to ECU load or no more than 5 units above ECU load. So it seems to me that the calculated load is doing a pretty good job.

 

Thanks.. I wanted to make sure someone was clear about it so everyone knew that both the calculated values and the ECU values did correlate, and that you still needed the calculated load anyway..

At idle and cruise, the Load value (1C) is of value though where the calculated load may not work right.

 

Is there a way to read the injector scaling at runtime?

d

EDIT: probably isn't clear. What I meant was, is there a way to read it off the ECU without putting it into boot mode and dumping the ROM?