S13 CA18DET AFM plug harness have 4 wires......

[louiswun]

I'm going to help my friend to tuen his S13,
but I have found that the CA18 AFM harness have 4 wires,
except +12V, signal, GND, there is one more green wire with yellow strip.
what to do for the extra wire ?
I use RB20 AFM and left that extra wire unconnected now.

[skylinegtrhr]

It is different in CA18det with catalytic converter and without:

[louiswun]

Yes, I know that that wire is for emisssion control (somewthing like idle mixture),
I have been reading a post in old ECU froum, if chaging AFM, this wire must have a resistor connected,
to control the voltage as a sihgnal to ECU, but I have no idel what ohm of resistor to use.

[Matt]

on RB30 AFMs this was variable and beteeen 0-1Kohm. Factory emissions setting was 370ohms for this ECU. You need to measure the pins on the S13 AFM to check for sure

I just leave this disconnected these days on the RB30 AFMs like on my GTS

[louiswun]

I have check from a CA18DET 4 wire AFM, the extra pin is a variable resistance around 1Kohm, but it seems ok without connecting.

[KillerB]

Hi there,

Im also wondering about this idle mixture signal. I have a problem running rich at idle/low load (like 1:12/13) and a bad mixup of JDM ECU on a Euro spec car. Pin 37 of the ECU is just left floating now.
Id like to try the resistor, see if it helps to get the mixture back to normal. So if one side of the resistor is attached to pin 37, where does the other side go? I suppose gnd...? I cannot check since i dont have a 4 pin AFM over here...

[Matt]

yes the other side is against ground (black wires typically on AFM plug. measure against battery ground to verify this)

[KillerB]

OK. Ive attached a 1k adjustable resistor between pin 37 and AFM gnd. That did not help. Whatever resistance I set, nothing changes. The problem actually gets weirder, because since yesterday instead of running rich it is running lean (like 16's when cruising)

So i started fiddling around with the flag bits in nistune, and for sure yes when I disabled O2 sensor the mixture returned to normal Doing 14.5-15 while idle/cruising and 11.5-12 on full load.

I still dont quite understand whats wrong here, but I do know for sure the narrowband simulation for the ECU is working as it should. Is it possible that the JDM ECU lacks the hardware/firmware for O2 feedback, and enabling the O2 sensor flag just corrupts fueling calculation?

[Matt]

JDM ecu should be using same O2 fueling. What does your O2 sensor look like when logging it. Does it bounce up/down or does it get stuck high or low?

[KillerB]

A closer look to the O2 sensor shows something interesting. Below is a log from the first time i started modifying the O2 feedback flag. In the beginning of the log the flag is enabled (this is the normal burned state) and right after the red cursor i disable the flag:



Wideband is purple, u can see the lean conditions 16Afr i was experiencing at the cursor. O2 sensor (green) is showing nothing. Then after i disabled the feedback flag (see just after the red cursor), the O2 sensor comes alive and the AFR lowers to normal 14.7 and starts oscillating. The next closeup confirms this:



So good thing is, feedback and narrowband simulation are working OK!
What is strange is that the feedback bit seems to be inverted?

[Matt]

so for S13 CA18 this flag is actually backwards in the ECU?!

[raddy]

I never experienced that with CA18DET....with default state (feedback ON) I allways get usual AFR 14,7:1 and mixture is for sure driven by OEM lambda sensor. Actually, I was never able to switch feedback OFF by switch in nistune... Only way how to use values directly from fuel map is only physicall disconnecting of OEM sensor. Your experience looks strange for me, as it should mean that OEM lambda sensor doesnt work in standard ECU?

[KillerB]

@Matt: Yes it seems to be inverted

This is a JDM ECU which normally doesn't have a lambda sensor. So i suppose it could be inverted for JDM ecu's only?