Megasquirt Support Forum (MSEXTRA)

I would really like to use the same input for both a instrument cluster gauge as well as the ECM. I have a K type thermocouple for the dash gauge, so can I use the conditioned and linear T/C amplifier output to feed into a spare 0-5 volt input and reassign the gauge to use this for the CTS reading?

TIA

I don't think it is reasonably possible because the CLT and MAT cal curves have to be the same. Maybe that can be handled with different LUTs in the .ini files, but I don't think so.

I am pretty certain this isn't worth the bother, even if it can be done. GM-type sensors can be bought for under $10, and as small as 1/8" NPT; so easy to locate almost anywhere. Look for a Standard Motor Products TX-73, for example.

Is that really true when Tuner Studio has separate calibration tools for each sensor?

The problem I've been fighting is having two different readings that can be off by 20F or more. Right now I'm using thermistors for both and I never know which to trust until I get out an IR gun and check actual head temperature. I've come to the conclusion that I hate thermistors and their non-linear & less than completely predictable response curves.

billr wrote:I don't think it is reasonably possible because the CLT and MAT cal curves have to be the same.

Nope. There are two different calibration lookup curves.

The only difficulty is that the lookup curve generator in TunerStudio is designed for thermistors and generates a curve. For a thermocouple you would need to hand generate a linear transform and import that file.

James

I looked for the correct table (there's an awful lot of tables in there...) and wasn't able to find the one I'm looking for. Any clue what it might be called, even a partial name I could do a string search on?

frontrangeFJ40 wrote:I...The problem I've been fighting is having two different readings that can be off by 20F or more. Right now I'm using thermistors for both and I never know which to trust until I get out an IR gun and check actual head temperature. I've come to the conclusion that I hate thermistors and their non-linear & less than completely predictable response curves.

You want something accurate? Use an RTD. The tradeoffs are a lower response time & greater expense.
RTDs rely on certain metals' (usually platinum, nickel, or copper) thermo-resistive behavior to detect changes in temperature. RTDs respond to temperature changes very linearly and can achieve good accuracy.

A thermistor is a temperature-sensitive resistor, whilst a thermocouple generates a voltage proportional to the temperature.
Resistance Temperature Detectors (RTDs) are sensors that measure temperature by correlating the resistance of the RTD element with temperature. Most RTD elements consist of a length of fine coiled wire wrapped around a ceramic or glass core.

https://cecas.clemson.edu/cvel/auto/sen ... ature.html

Thanks, but I started my professional life as an EE and I'm pretty familiar with the technology.

frontrangeFJ40 wrote:I looked for the correct table (there's an awful lot of tables in there...) and wasn't able to find the one I'm looking for. Any clue what it might be called, even a partial name I could do a string search on?

You would need to go to Tools -> Calibrate Thermistor Tables. There's a button there for uploading custom INC files.

FANTASTIC!!

Any idea where I can find an example of how that file should be formatted??

If you look in the "inc" subfolder of the folder you have TS installed in, there's a file called "thermfactor.inc" that I think is in the right format. But I'm not sure what the numbers in it mean... I think there are 256 entries representing A-D values, but the actual numbers in the file I'm looking at are kind of weird.

I'm not actually sure of the format that TunerStudio expects, best to ask in that sub-forum.

Inside the firmware, the table has 1024 entries for MS2/3, these correspond to ADC count. ADCcount = Volts/5.0 * 1023. The temperature is in degrees Fahrenheit, * 10.

James

Thanks guys!