Methods for finding K value
What are you guys methods for finding K value?
I am running a basic daughterboard with the SST 27sf512 chips and have the ability to erase and burn bins as needed.
The issue I have been having is that my car runs horribly rich at idle, high 9's low 10's very unsteady. What would be the best way to get the idle steady?
Here is my setup:
S13 SR20DET stock bottom end
Cometic HG
BC 264
Z32 Maf
BC Valvesprings
Top Feed Low impedance 75lb injectors
50 trim t3/t4
Just changed the plugs to bk7re, timing was set to 15* and TPS set to .5 volts closed.
Not sure if my idle air speed needs to be adjusted but the car idles around 750-800 rpms.
I am currently using tuner code and Nismotronic for logging and making changes to the BIN's. I just purchased a Innovate MTX-L wideband and took out my old Gloshift wide band, which btw is for sale.
I was on ECCS and stumbled upon these posts which gives a method for finding your K value below:
Credited by Iforgot on ECCS (Not trying to steal his info just get information)
Basically, the point of the proceduce was to set the K value for some 'known' A/F ratio. Now, the entries in the fuel table are multipliers to the basic pulsewidth. Without even looking at the fuel table, with the correct K value, the engine knows the correct injector pulsewidth to obtain a stoich (14.7 AF) ratio. The entries in the table modify this injector pulsewidth by the following formula (with DATA being from 0 to 255):
if DATA > 128 then MULTIPLYER = (DATA - 64 / 128)
else
MULTIPLYER = (DATA + 128 / 128)
new A/F ratio = 14.7 / MULTIPLIER
This is the multiplier to the basic pulsewidth. So, if we set the DATA to 0, the multiplier is (0 + 128 / 128) = 1. So the injectors will be pulsed with a pulsewidth which should produce an A/F ratio of 14.7. I know I said to use the DATA to 0 here. That is the theory. However, badbiki said to use 1 for some reason. One explanation I can come up with is that we DON'T want the ECU to go in closed loop mode. That will defeat the purpose. It will adjust the injector pulsewidth on it's own to hone in on A/F ratio of 14.7
The reason why I went thru all of this is to generalize the method. Essentially, you can set the fuel table up with data of your choice, calculate the A/F ratio that it should be, and then tune the K value on the wideband to hone in on it! If you happen to choose 1 as the DATA, then you should tune for 14.59 (do the calculation). If you choose the DATA to be 30, then you should tune for 11.9 A/F ratio. Perhaps the rationale behind reving the engine up to 2000 rpm under no load is just to be safe. We know that even turbocharged engines which need rich cyllinder burn conditions, are happy with 14.7 A/F ratio under low load/idle. Hence, if we rev it to 2000 rpm under low load or in nuetral, there is no chance of damaging your engine under this radically different temporary fuel map.
****************************************
Let me add another tuning tip if you're using the above method. Let's say you've set your DATA to 1 everywhere in the fuel map. Now, according to the formula 14.7/(DATA + 128/128), we should be seeing an A/F ratio of 14.59. Great, but the whole reason why we're going thru this exercise is because we don't have an A/F of 14.59! So, let's say the measured A/F is X. Do the following:
1) Calculate what the A/F SHOULD be using (for DATA < 128)
AFR = 14.7 / (DATA+128 / 128)
2) Rev engine up to 2000 rpm and note the A/F ratio on your wideband.
Let this A/F ratio = X
3) compute Y = X/AFR
(the ratio between the actual A/F ratio and desired)
4) multiply the existing K value with Y
This should bring you very close to the actual K value in just one step. Obviously, in some cases, you may need some fine tuning. Also, before doing any of this, you should already have a ball park figure of the K value by following the procedures for changing the injectors and MAF in the following thread: hybridka.com .
I am running a basic daughterboard with the SST 27sf512 chips and have the ability to erase and burn bins as needed.
The issue I have been having is that my car runs horribly rich at idle, high 9's low 10's very unsteady. What would be the best way to get the idle steady?
Here is my setup:
S13 SR20DET stock bottom end
Cometic HG
BC 264
Z32 Maf
BC Valvesprings
Top Feed Low impedance 75lb injectors
50 trim t3/t4
Just changed the plugs to bk7re, timing was set to 15* and TPS set to .5 volts closed.
Not sure if my idle air speed needs to be adjusted but the car idles around 750-800 rpms.
I am currently using tuner code and Nismotronic for logging and making changes to the BIN's. I just purchased a Innovate MTX-L wideband and took out my old Gloshift wide band, which btw is for sale.
I was on ECCS and stumbled upon these posts which gives a method for finding your K value below:
Credited by Iforgot on ECCS (Not trying to steal his info just get information)
Basically, the point of the proceduce was to set the K value for some 'known' A/F ratio. Now, the entries in the fuel table are multipliers to the basic pulsewidth. Without even looking at the fuel table, with the correct K value, the engine knows the correct injector pulsewidth to obtain a stoich (14.7 AF) ratio. The entries in the table modify this injector pulsewidth by the following formula (with DATA being from 0 to 255):
if DATA > 128 then MULTIPLYER = (DATA - 64 / 128)
else
MULTIPLYER = (DATA + 128 / 128)
new A/F ratio = 14.7 / MULTIPLIER
This is the multiplier to the basic pulsewidth. So, if we set the DATA to 0, the multiplier is (0 + 128 / 128) = 1. So the injectors will be pulsed with a pulsewidth which should produce an A/F ratio of 14.7. I know I said to use the DATA to 0 here. That is the theory. However, badbiki said to use 1 for some reason. One explanation I can come up with is that we DON'T want the ECU to go in closed loop mode. That will defeat the purpose. It will adjust the injector pulsewidth on it's own to hone in on A/F ratio of 14.7
The reason why I went thru all of this is to generalize the method. Essentially, you can set the fuel table up with data of your choice, calculate the A/F ratio that it should be, and then tune the K value on the wideband to hone in on it! If you happen to choose 1 as the DATA, then you should tune for 14.59 (do the calculation). If you choose the DATA to be 30, then you should tune for 11.9 A/F ratio. Perhaps the rationale behind reving the engine up to 2000 rpm under no load is just to be safe. We know that even turbocharged engines which need rich cyllinder burn conditions, are happy with 14.7 A/F ratio under low load/idle. Hence, if we rev it to 2000 rpm under low load or in nuetral, there is no chance of damaging your engine under this radically different temporary fuel map.
****************************************
Let me add another tuning tip if you're using the above method. Let's say you've set your DATA to 1 everywhere in the fuel map. Now, according to the formula 14.7/(DATA + 128/128), we should be seeing an A/F ratio of 14.59. Great, but the whole reason why we're going thru this exercise is because we don't have an A/F of 14.59! So, let's say the measured A/F is X. Do the following:
1) Calculate what the A/F SHOULD be using (for DATA < 128)
AFR = 14.7 / (DATA+128 / 128)
2) Rev engine up to 2000 rpm and note the A/F ratio on your wideband.
Let this A/F ratio = X
3) compute Y = X/AFR
(the ratio between the actual A/F ratio and desired)
4) multiply the existing K value with Y
This should bring you very close to the actual K value in just one step. Obviously, in some cases, you may need some fine tuning. Also, before doing any of this, you should already have a ball park figure of the K value by following the procedures for changing the injectors and MAF in the following thread: hybridka.com .