Page 1 of 1

Temp sensor calibrations - are they in the msq?

Posted: Wed May 20, 2009 1:44 pm
by RedRocketRally
Over the weekend, I ran my Microsquirt-ed rally car in the Oregon Trail rally. The weather was the hottest we've had this year, and I discovered that the car runs really poorly when it gets hot outside. I also noticed the Intake Air Temp reported by Megatune, first thing in the morning when the motor was still cold, was higher than both the coolant temp and the outside air temp. Surely I need to recalibrate. When it worked, it was a screamer, but as the day got hot, it didn't want to idle or start (it pretty much only worked well at half- to wide-open throttle when it was 80+ outside).

Just now I was digging through my msq file in Notepad, and can't find the temperature calibration settings. So my first question: are they in there, and if so, what's the variable called?

I also found "thermistor.log", and the info in that file ("this is what was sent, and you can delete this file...") makes me think the calibration info is only stored in the MS flash memory, and probably is wiped out with a firmware update.

I've updated the Microsquirt firmware a few times since I first set the IAT and CLT calibration settings (using room temp, ice water, and boiling water per the tutorial on diyautotune.com). I was assuming the calibration would load when I reloaded the msq, but now I suspect it isn't, and is using some default settings instead. Do I need to re-enter the calibration info anytime I load new firmware? Do I need to recalibrate when I load an msq file, or just on firmware updates?

Are there any other configurations that I need to re-do after a firmware update that aren't stored in the msq?

Thanks,

brian

Re: Temp sensor calibrations - are they in the msq?

Posted: Wed May 20, 2009 7:00 pm
by TheMonkey
Brian-

yes, it's burned into flash memory using the thermistor thing. tools -> calibrate thermistor tables...

inputs are bias resistor value, and 3 data points. the bias resistor value is a fixed value on the board. the 3 data points are not linear, so there is some kind of interpolation when the curve is burned in.

default values are GM sensors. i was recently creeping around this info because i'm working on a BMW which has different values.

too bad you found the problem on a day that really counts.... otherwise, how did it go? hopefully didn't ruin all the fun.

Re: Temp sensor calibrations - are they in the msq?

Posted: Wed May 20, 2009 8:42 pm
by RedRocketRally
It didn't ruin any fun, but it did make getting there a little more of a challenge. We ended up 3rd place in class on Saturday and 2nd on Sunday. Overall it was a great weekend.

Its all a learning experience. At least now I know what I need to do each time (that's half the battle! ;). I'll have to measure the sensors again, since I didn't write the numbers down, but that's not too big of a deal.

Learning all this stuff is fun, but sometimes a little quirky, i guess.

Thanks!

brian