Braineack

iTrader: (62)

only takes 4 screws to open the ecu up. you know what lead the input comes in

paul

iTrader: (21)

the trick is just to change the .inc files each time since all they do is change how MS sees the temps. then burn to ECU the .s19 when you are happy with results. after burning the s19 file to the ms this method proved accurate for both my 94 and with the numbers we did on jason's 01 with my MS. his MT figures couldn't have been closer to perfect.

paul

iTrader: (21)

setting we used for jason's 01 were clt 1500, iat 1300. he is using the stock 01 sensors for both.

Saml01

iTrader: (9)

Smart, I didnt think to do that.

supersuk

iTrader: (5)

I'm using values that are similar to Pauls and I have a 90 with the GM sensor. I had to use a bias resistance of around 1500 to get the damn sensor to read right.

paul

iTrader: (21)

so when looking at STOCK ECU 5th pin from the left along the top row of the bigger connector is CLT?

and 6th from left on bottom row on same connector is IAT?

now what pins are ground?

paul

iTrader: (21)

WOOT! +1 for paul edit: "'s stupid approach"

Braineack

iTrader: (62)

paul

iTrader: (21)

hmm, testing resistance at the pins out of the ecu yields
~675 for the CLT
~1500 for IAT

for 2 94 ecu's i had in the house

paul

iTrader: (21)

so by everyone else's rationale i should be using 675 for the CLT and 3990 for IAT, correct?

even though those are most likely gonna give me some wildly wrong numbers.

paul

iTrader: (21)

enough theory, i'm a hands on kinda guy. someone running those numbers go stick their GM IAT sensor in ice water and then boiling water and see what megatune reports.

Saml01

iTrader: (9)

By everyones rationale you should be having no issues using 2490 for both.

paul

iTrader: (21)

No. the way i understand it is:

the MS has a bias resistor at R4 for IAT. i believe it's value is 2490. Therefor since it's wired in parallel with the stock ecu you need to add the bias resistance of that which is 1500 per the way i just measured. So 2490 + 1500 = 3990 for the IAT.

and the MS doesn't have a resistor at R7 for CLT so thats 0 but the stock ecu is 675 so i should use 675 for CLT.

or am i completely missing the boat here?

paul

iTrader: (21)

yo, i just thought of something. parallel. that means the MS should see the same whether the stock ecu is in place or not. it's tapped off the line BEFORE the stock ECU. WTF.

so with stock ecu not connected to the parallel harness does megatune show a reading for CLT and IAT? it should by my calculations. and it should not change once the stock ecu is plugged back in. someone please confirm or refute this. i can't test as i lent my MS out. **** **** ****

paul

iTrader: (21)

no answers yet? damn you ******* are slow. it's been almost a minute!

paul

iTrader: (21)

the MS is just reading the voltage coming from the sensor right?
which is exactly what the stock ecu is doing.
both are reading the voltage and through whatever calculation comparing the returned voltage with the 5volts that was sent to the sensor and determining the temp based on that and what it knows the resistance curve of the sensor to be.


****! but from what i read of the MS-I that doesn't make sense because without a resistor at R7 the MS can't do **** with the reading.

"In order to measure these two temperature inputs, MS is configured to use thermistors that vary in resistance with temperature changes. The microprocessor cannot read variable resistance inputs directly, as it is designed to measure inputs as variable voltage signals between 0 and 5v. In order to convert the thermistors variable resistance into a variable voltage signal, the thermistor is connected in series with a fixed bias resistor as shown in the graphic below. The free end of the bias resistor is connected to the regulated 5V supply on the MS circuit board, with the free end of the thermistor is grounded. These two series resistors (the fixed bias resistor and variable thermistor) create a voltage divider circuit. The voltage at the junction varies from 0V with the thermistor at 0 ohms, to nearly 5V when the thermistor resistance is very high with respect to the bias resistor."

my head hurts more now.

Braineack

iTrader: (62)

hahah


you just need to know the resistance of the resistor the oem ecu uses.


if you take a DDM to the ECU itself at the AIT signal and ground, that should tell you.


worse comes to worse, you unscrew 4 phillips head screws, find the resistor and measure right on the damn thing.

paul

iTrader: (21)

yeah, read above. i did that. have 2 94 ecu's in the house. measured at the corresponding pins. got 1500 for IAT and 675 for CLT https://www.miataturbo.net/forum/sho...1&postcount=29

so the idea is to add them to whatever the MS resistors are at R4 and R7, correct? makes no sense in my head but then again i am getting old and have the flu and have been staring at this monitor too long today

Saml01

iTrader: (9)

Wrong. The stock ECU's bias is not 675 for CLT.

Why are you sharing your GM AIT with your stock ECU?

As for the ECU not getting readings without R7. Thats only if you dont have a bias resistor in the circuit. Since you are getting yours from the ECU, you dont have to install one on the MS. It doesnt work with the stim plugged in, but works with the car plugged in. I dont know why or how though.

Braineack

iTrader: (62)

sounds really low. measure a known resistor to verify. 675 would be a .675K