Difference in observed injector latency vs maps?
Feb 19, 2009, 08:37 AM
#1
Newbie
Thread Starter
iTrader: (1)
Join Date: Apr 2008
Location: Missouri
Posts: 8
Likes: 0
Received 0 Likes
on
0 Posts
Difference in observed injector latency vs maps?
I've been playing around with dissecting ROM files for the 3000GT and 3G Eclipse, and I was wondering if someone could explain the injector latency values a little bit. Using the Evo 8 as a guide, I'm seeing a discrepancy between the latency values in the maps and what people have seen in real world testing from injector rebuild shops and AEM.
It's not quite an apples to apples comparison since this chart uses even voltage numbers and the ROM images use those oddball voltages, but I still think there's a pretty good gap.
I'm basing my numbers on the chart here: http://injector-rehab.com/kbse/lag.htm
For the Evo 8, they list the following latencies:
v / ms
10 / 1.13
11 / 0.97
12 / 0.85
13 / 0.75
14 / 0.66
15 / 0.58
Whereas the 94170008 ROM lists them like this:
4.69 / 3.312
7.03 / 1.680
9.38 / 1.032
11.72 / 0.672
14.06 / 0.432
16.41 / 0.264
18.68 / 0.144
The closest overlap we have is at 14v, which is different by about 23 ms.
Then we look at my 3000GT ROM. It uses the same battery voltage scale as the Evo. So I'm taking a guess that the scaling factor is the same for the values since so many other calculations match up as well.
Injector rehab's values for the 3000GT injector:
10 / 1.07
11 / 0.94
12 / 0.81
13 / 0.73
14 / 0.64
15 / 0.54
And the ROM:
4.69 / 6.120
7.03 / 2.832
9.38 / 1.560
11.72 / 1.128
14.06 / 0.720
16.41 / 0.480
18.68 / 0.480
At 14V it's closer than the Evo values, but the whole scale still seems way off.
So what I'm really getting at is, am I missing something about how the ECU interprets the latencies? How do you get from lab results to the values you enter into the ECU? I'm going to be recalibrating my 3000GT ECU for my 550cc injectors (Rochesters) and I want to make sure I understand what it is I'm actually entering.
Thanks!
It's not quite an apples to apples comparison since this chart uses even voltage numbers and the ROM images use those oddball voltages, but I still think there's a pretty good gap.
I'm basing my numbers on the chart here: http://injector-rehab.com/kbse/lag.htm
For the Evo 8, they list the following latencies:
v / ms
10 / 1.13
11 / 0.97
12 / 0.85
13 / 0.75
14 / 0.66
15 / 0.58
Whereas the 94170008 ROM lists them like this:
4.69 / 3.312
7.03 / 1.680
9.38 / 1.032
11.72 / 0.672
14.06 / 0.432
16.41 / 0.264
18.68 / 0.144
The closest overlap we have is at 14v, which is different by about 23 ms.
Then we look at my 3000GT ROM. It uses the same battery voltage scale as the Evo. So I'm taking a guess that the scaling factor is the same for the values since so many other calculations match up as well.
Injector rehab's values for the 3000GT injector:
10 / 1.07
11 / 0.94
12 / 0.81
13 / 0.73
14 / 0.64
15 / 0.54
And the ROM:
4.69 / 6.120
7.03 / 2.832
9.38 / 1.560
11.72 / 1.128
14.06 / 0.720
16.41 / 0.480
18.68 / 0.480
At 14V it's closer than the Evo values, but the whole scale still seems way off.
So what I'm really getting at is, am I missing something about how the ECU interprets the latencies? How do you get from lab results to the values you enter into the ECU? I'm going to be recalibrating my 3000GT ECU for my 550cc injectors (Rochesters) and I want to make sure I understand what it is I'm actually entering.
Thanks!
Feb 19, 2009, 09:14 AM
#2
The values in ECUFlash aren't actual ms. So for 660ms it has 432ms value in the ECUFlash. So what some people do is take a measured injector ms, lets say for 14v and lets say it's 700ms and recalculate it using the above example, i.e. 700x (432/660) = 458. So they would put .458 in the ECUFlash instead of 700... You'll still have to fine tune it using trims. Most info you need is here though: https://www.evolutionm.net/forums/ec...injectors.html
