Injector Scaling Formula
Jun 10, 2009, 03:32 PM
#17
Evolved Member
Join Date: Apr 2008
Location: Sydney, Australia
Posts: 604
Likes: 0
Received 0 Likes
on
0 Posts
You should note that most injectors always need the pressure to be quoted at the same time as their flow rate. One without the other is meaningless. So if you blindly buy 650cc injectors the manufacturer could have tested at 100psi for all you know, and they really flow alot less and similar to stock. Then if you run a different fuel pressure you also effect the latency, so any quoted values will be different. Setting this is up never easy for the DIY's. That is why there are professionals out there. 
Very interesting to see if a correct multiple is found. In the end it does not really matter that much, just so long as you tune the car and not the numbers.
Very interesting to see if a correct multiple is found. In the end it does not really matter that much, just so long as you tune the car and not the numbers.
Jul 30, 2009, 01:49 PM
#18
This seems to be the closest for me:
31275/x
Its not perfect, since I think there is an adder of sorts that may be required for this formula, but it gets you in the ballpark I think. I personally think the EVO injectors seem to react closer to 550's than 560's and this scaling kind of depicts that.
.
31275/x
Its not perfect, since I think there is an adder of sorts that may be required for this formula, but it gets you in the ballpark I think. I personally think the EVO injectors seem to react closer to 550's than 560's and this scaling kind of depicts that.
.
Last edited by Jack_of_Trades; Jul 30, 2009 at 03:08 PM.
Jul 30, 2009, 02:45 PM
#19
Evolving Member
Thread Starter
And the two formulas are not inverse to each other, they don't match.
Jul 30, 2009, 03:11 PM
#20
And YES, they DO INDEED match.
ROM to Display f(x)= 27440/(57-8) =560
Display to ROM f(x)= (27440/560)+8 =57
The problem is thats only accurate between 450 and 560 but its doesn't hold true throughout the entire range,which was my bad.
Last edited by Jack_of_Trades; Jul 30, 2009 at 03:15 PM.
Sep 7, 2009, 07:06 PM
#21
Evolving Member
Thread Starter
One of the first fuel injected, ecu controlled cars mitsu made used 260cc injectors at 38 psi. So they arbitrarily set this flowrate to be 100%.
260cc @ 38 psi = 243 cc
31050/128 = 243 cc
(128 is from 100% trim=0x80=128)
Sep 8, 2009, 11:15 AM
#24
Interesting, as the numbers don't line up for the EVO based on Ceddy's post above.
560cc @ 43.5 PSI = 523cc @ 38 PSI
31050/57 = 544.7cc
If it was correct, either the factory injectors are 582cc @ 43.5 PSI or Mitsubishi should have used 59 instead of 57. It could also just be non-linearities within the ECU/drivers.
Having a solid connection between reality and the ECUFlash software is a good thing. I paid extra for fully charted Injector Dynamic Injectors that I have a list of very accurate numbers for size and latency. If the numbers in ECUFlash meant anything, I could scale for the injectors perfectly with just a few strikes of the keys. After that, logging would just be to verify and not tune.
Yes, all that matters in the end is how the car runs, but if the numbers are right for things like injector sizing, MAF scaling, etc, it means you can make accurate calculations for things like brake specific fuel consumption and engine volumetric efficiency. Being able to accurately calculate this stuff can improve tune quality substantially.
560cc @ 43.5 PSI = 523cc @ 38 PSI
31050/57 = 544.7cc
If it was correct, either the factory injectors are 582cc @ 43.5 PSI or Mitsubishi should have used 59 instead of 57. It could also just be non-linearities within the ECU/drivers.
Having a solid connection between reality and the ECUFlash software is a good thing. I paid extra for fully charted Injector Dynamic Injectors that I have a list of very accurate numbers for size and latency. If the numbers in ECUFlash meant anything, I could scale for the injectors perfectly with just a few strikes of the keys. After that, logging would just be to verify and not tune.
Yes, all that matters in the end is how the car runs, but if the numbers are right for things like injector sizing, MAF scaling, etc, it means you can make accurate calculations for things like brake specific fuel consumption and engine volumetric efficiency. Being able to accurately calculate this stuff can improve tune quality substantially.
Last edited by 03whitegsr; Sep 8, 2009 at 11:18 AM.
Sep 8, 2009, 04:14 PM
#25
Evolving Member
Thread Starter
Interesting, as the numbers don't line up for the EVO based on Ceddy's post above.
560cc @ 43.5 PSI = 523cc @ 38 PSI
31050/57 = 544.7cc
If it was correct, either the factory injectors are 582cc @ 43.5 PSI or Mitsubishi should have used 59 instead of 57. It could also just be non-linearities within the ECU/drivers.
Having a solid connection between reality and the ECUFlash software is a good thing. I paid extra for fully charted Injector Dynamic Injectors that I have a list of very accurate numbers for size and latency. If the numbers in ECUFlash meant anything, I could scale for the injectors perfectly with just a few strikes of the keys. After that, logging would just be to verify and not tune.
Yes, all that matters in the end is how the car runs, but if the numbers are right for things like injector sizing, MAF scaling, etc, it means you can make accurate calculations for things like brake specific fuel consumption and engine volumetric efficiency. Being able to accurately calculate this stuff can improve tune quality substantially.
560cc @ 43.5 PSI = 523cc @ 38 PSI
31050/57 = 544.7cc
If it was correct, either the factory injectors are 582cc @ 43.5 PSI or Mitsubishi should have used 59 instead of 57. It could also just be non-linearities within the ECU/drivers.
Having a solid connection between reality and the ECUFlash software is a good thing. I paid extra for fully charted Injector Dynamic Injectors that I have a list of very accurate numbers for size and latency. If the numbers in ECUFlash meant anything, I could scale for the injectors perfectly with just a few strikes of the keys. After that, logging would just be to verify and not tune.
Yes, all that matters in the end is how the car runs, but if the numbers are right for things like injector sizing, MAF scaling, etc, it means you can make accurate calculations for things like brake specific fuel consumption and engine volumetric efficiency. Being able to accurately calculate this stuff can improve tune quality substantially.
Also heard the JDM Evos have a higher injector number, but EcuFlash didn't have XMLs for any of my JDM roms so couldn't check.
If the JDM is 55(0x37), it would be dead on for 31050 @ 43.5 psi.
Sep 8, 2009, 06:22 PM
#27
I also heard that the non-JDM roms had the injector scaling altered so they ran a bit more rich. It was either for warranty precautions or something about the gas offered in other parts of the world,can't remember. I'd LOVE to confirm or deny that rumor though!
Sep 9, 2009, 08:33 AM
#28
My mind was actually going a different direction. Not even sure what I was thinking now. I was just relating the numbers to the 260s@38 psi, which wouldn't make sense. Disregard. 
It still doesn't line up with the 560s that are in the EVO from the factory as the scaling is saying they are 545s.
If 31050 turns up in the fuel cals though, it would definitely make sense that it is correct though and it's just an issue of non-linearites in the drivers/ECU.
It still doesn't line up with the 560s that are in the EVO from the factory as the scaling is saying they are 545s.
If 31050 turns up in the fuel cals though, it would definitely make sense that it is correct though and it's just an issue of non-linearites in the drivers/ECU.
Sep 9, 2009, 05:59 PM
#29
Did anyone actually have stock injectors cleaned and flow tested? I tried to search some data, but couldn't find any. Although many people say stock EVO injectors are 560 cc/min, but this site says EVO injectors are actually 550 cc/min?
http://injector-rehab.com/kbse/lag.htm
Then 545 cc/min is very close to the actual value?
http://injector-rehab.com/kbse/lag.htm
Then 545 cc/min is very close to the actual value?
Last edited by ace33joe; Sep 9, 2009 at 06:25 PM.
