You should be able to bypass the IAT sensor using a resistor in series. The resistance is TBD, but I'll check out my Helm manual tomorrow to see. Several possible effects, none of which should be harmful and might even be beneficial, especailly in hot weather.

UL

 

3.2K = 104 deg F
2.7K = 115 deg F
2.2K = 124 deg F
1.8K = 135 deg F
1.5K = 145 deg F

Enjoy

BTW, engine doesn't seem to start dialing back the timing until IAT is somewhere > ~ 160 deg F.

 

IS, I've had this problem since the 1st week I bought the Stook. It is extremely intermittent and can happen 6 times in one day and then not for 3 months. It doesn't seem to be specifically temperature dependent either as I have had it happen on cold winter days. I will admit that it does seem to happen more frequently in summer though. Needless to say Honda is in hurry to solve the problem for me.

I hope you are correct in your diagnosis. I am very interested to hear how it all pans out - if it solves the problem I'll certainly be putting a case forward for the same fix for mine.

 

What I've begun to do at autocross sessions is that just before taking off, just rev the engine a few times to suck cooler air into the engine bay from the outside - hopefully, the outside air is not anywhere near 160 degrees!

 

2kTurkey,

This morning our head mechanic tested the IAT sensor when the engine was still cold. It gave an accurate reading of the surrounding temperature. The thing is, the check engine light that came on pointed to the IAT sensor as malfunctioning.

The computer was resetted to remove the check engine light, the car was started again and the light never came on. The problem is quite obvious, the sensor malfunctions intermittently and probably causes the intermittent stalling as well.

Anyway, there was yet another trouble code that came out and that trouble code does not appear on the service manual of the S2000. For now, the faulty IAT sensor seems to be causing the stumbles but until our mechanic is able to determine error code 5A, then we can never be too sure.

I'll just post whatever develops within the week. Nevertheless, you might want to bring your car to your dealer and have the IAT sensor changed under warranty.

 

Thanks RT, I didn't even need to go to the garage and get the manual out of the trunk :-)

Now, here's what you can try. Put a 3.3kOhm resistor in series with the IAT sensor (you will have to cut a wire). No matter what the IAT sensor reads, the ECU will think the intake air is below 100 F.

What will this do?

1. Prevent dialing back of timing due to excess temps
2. Cause the car to run a bit richer (perhaps with more power).

Potential drawbacks?

1. Running richer could cost a bit of power.
2. Not retarding intake temps could increase chances of detonation. But since the Mugen ECU doesn't retard timing and we're only talking about 50-60 F, I wouldn't worry. Also, you have a knock sensor to protect you.

I think I'll try this next time I can dyno the car, put a switchable link in there, or maybe a potentiometer. I'd prefer a more elegant solution, but its a worthwhile experiment IMO.

UL

 

Hold up!
Sorry UL, the values I posted were not temp offset values.
They were what the ECM would see if you substituted the resistor in place of the thermistor (IAT sensor).

If going with the series plan (UL) instead of the substitute plan (RT), I'd use something on the order of 1.5K ohm for a NA car and maybe a 1K ohm for a FI non intercooled set up. Keep in mind the offset will not be constant since the thermistor is not linear (the attenuation effect gets larger as temp rises).

I think


[Edited by RT on 05-23-2001 at 11:21 AM]

 

Ahh, sneaky! I take it you've been substituting resistors and looking at what the OBDII scanner outputs?

The basic intent with the series play is to not completely fix the temp reading, but to make it read lower than actual. A 1.5k resistor would have an average effect of about 20-40F in normal operating ranges (seems that on a 70F day, intake temps underway are in the 110F range, +/-10F).

BTW, where did you get a 3.2k resistor? I seem to remember 3.3k as the next standard value (red/red/red!), but it has been a while. Or are you using a potentiometer?

UL

 

It's marked orange/orange/red/gold (3.3K) but it measures 3.2K (RadioShack and all )

 

Gold = 5% tolerance, so it's still "right"

-koejing