I installed my e440 Select ECU (MSM version) a while ago (see http://mx5cartalk.com/forum/viewtopi...41091&start=75
and onwards) with the help of prospero's installation notes: 440 Select from e420c install
(Thanks very much for that! I've added an "Excellent" props to the thread since it helped me out a lot).
Andy from Adaptronic had provided me with a base map that he'd quickly whipped up since the ECU had only just been released and I used this as a starting point. I was quite wary of his initial sensor values since the graphs weren't very smooth so decided to try using e420c values instead since they have been "nailed down" for quite some time now. I have the MAZDASPEED graphs that map the ECT (Water Temp), IAT (Aux Temp - @ air filter) and IAT2 (Air Temp - @ throttle body inlet) thermistor resistances vs temperatures which, when used with the ADC count formula defined in the Adaptronic documentation, mapped pretty well against the Boundary Engineering e420c values.
I've done the linear temperature gauge mod so my dummy temperature gauge does actually work. Usually it sits at about 1/3 if the car has been well and truly heated up but all of a sudden it was sitting at 2/3rds and when I got stopped at a red light, steam was coming up from the front of the car. When I pulled into a petrol station (what you guys call a "gas station"), there was a popping noise and coolant was squirted from every orifice of the bonnet (what you guys call "the hood"). When I opened the bonnet/hood, the radiator cap was sitting upside down on the radiator and there was coolant everywhere.
I topped up the coolant and turned on the air conditioner, which activates the radiator thermofans, and the temperature dropped back to 1/3rd.
Andy subsequently told me that the formula does NOT apply to the e440 ECUs (actually when I went back and read some of my initial correspondence with him, I found that he did state that earlier :( )
So I decided that I needed to calibrate all sensors by chilling and boiling them to determine what values they should actually be.
Here's my "high-tech" test lab (required 3 USB cable extensions to reach the netbook):
Once I'd heated the first sensor up to 100C and turned off the burner, I went inside the house to let it cool down and heard an "explosion". The pyrex dish had exploded and there was water and glass everywhere - including water in the netbook which had shut down and I had to pull apart to "dehydrate". Luckily it started again but I still wasted about 2 hours doing this and cleaning everything up.
Test setup 2:
In the bottom photo, I'm validating the digital thermometer against a regular one.
This worked MUCH better - damned pyrex, not quite the same quality as we used back in high school chemistry.
Anyway, here's the calibrated ECT (Water Temp) graphs:
That last one shows e420c Calculated values, Andy's initial values, measured values, and Boundary Engineering values. As you can see, the e440 and e420c values are quite different - no wonder my radiator boiled, the thermofans were never activating!
Andy's initial values weren't too bad!
Calibrated IAT2 (Air Temp) graphs:
In the bottom graph, you can see that the calculated values are completely screwed.
Once again, Andy's initial values were pretty close from 25C onwards.
Calibrated IAT (Aux Temp - air filter temp) graphs:
Andy didn't put much time into this since it's not used by the e440, it's merely reported.
Interestingly, the calculated values match up pretty well (which is because the e440 doesn't actually do anything with the value other than to report it).
* I only calibrated from 5C..100C since I was calibrating the sensor in water; other values were sourced from Andy's initial values.
* I should probably smooth these out a bit manually.