As you may know, the ECU uses the CHT resistance to determine how much to increase the injection pulse width (i.e. mixture enrichment) when the motor is cold. I have some charts on this on my ECU web page, and I simulated the action of the circuit, which showed that once resistance of the CHT dropped below about 300 ohms, the ECU no longer keeps leaning out the mixture - this is the "stead state" or "fully warmed up" mixture.

I never actually verified this behavior with an ECU, so I experiemented with three ECU's this morning (037, 043, and 052), and verified that what was simulated is how it actually acts. Independent of engine speed or load, once the CHT drops below about 330 to 375 ohms, the injection pulse width stays the same.

OK, to complete my comments on this - The next step I need to do is to get a CHT gauge on my car to observe the head temp through warm-up and various drive cycles. From characterization I did on some CHT's, I have a decent correlation of resistance to temperature. My goal is to see if during normal running, and after the engine is fully warmed up, to see if the head ever cools off enough so that the CHT starts richening the mixture. I have a suspicion that it does, and it may explain certain driveability issues.

I also plan to get an oil temp gauge, and to see if it helps support my theory that the "warm start" problem with the 2.0L is due to the head cooling off more than the core engine temperature, leading to rich running and poor starting until the head warms back up.

Cool stuff! Thanks for sharing your experiments. I love it!

the head is probabably the best of choice if only other choice is oil temp for feul injection cold enrichment. that is were all the compbustion is happening, regardless if the oil is warmed or not. in cold climates the oil may take near forever to heat up much, yet the heads get to operating temp fast even in cool weather. there is probably tooo much lag in the oil temp rise to be useful. however it may be useful as you suggest to even out the swings of the head temp if they are with in the control range of the sendor.

do you have correalation of pulse time verses total flow?

keep up the interesting research, i am on the edge of me seat!!!

Thanks Brad for all of your research, you are doing something most of us have no clue how to do. You are the Man!
Tom

Re: realred914's comments - What I'm learning is that determining the combustion chamber temperature of an air cooled motor isn't a simple thing. Add in that the cylinder is cast iron and the head is aluminum, with different thermal properties, and it makes it worse. I agree oil temperature is a poor indicator, too. And for various reasons, I don't think placing a sensor on the outside of the cylinder is any better.

funny/interesting that you would be observing this now. I have taken measurements in the course of troubleshooting that support your theory. the case and cylinders stay hot/warm long after teh heads have cooled. I have measured the change in resistance on teh CHT while teh oil remains hot. I look forward to reading your findings

This is really interesting stuff. I am a real believer in the fuel injection system originally designed for these cars. What about some type of timed interpolation based on both oil and head temp?