Hi

Your equation should just be 'A-40' and your units should be in C. Torque converts the units for you, and if you convert them in the equation then things will end up being double-converted.

Put 'A-40' in the equation, and C in the units, and everything will start working. I may visit this in a later version.

Also i suspect your current F to C conversion may be incorrect -Where are you adding '32' to convert to C as well as the multiplication part (your x/y)?

Thanks for the quick reply!

Yes, going with "A-40" instead of "A*(9/5)-40" works. Curious, how does Torque know it is a temperature, and so it should apply the conversion to Fahrenheit? If I leave the unit type as F it doesn't seem to care. Is it the ODB2 mode?

Quote from admin on April 26, 2011
Also i suspect your current F to C conversion may be incorrect -Where are you adding '32' to convert to C as well as the multiplication part (your x/y)?

Not sure what you are asking? The "A*(9/5)-40" formula is supposed to be for Fahrenheit. Taking that, subtracting 32, and multiplying by 5/9 gives "A-40". Which seems to give quite accurate displays now, within a degree or two F. At least, it agrees closely with engine coolant and intake air temperature, when the engine is just started and the fluids should be at ambient.

Hi

It knows it's a temperature as you have put 'C' in the units! 🙂

The temperature from the ECU is in C, not F. Your formula does not do the offset (-32) from the temperature in your original conversion, therefore you will be significantly out in the calculation

"It knows it’s a temperature as you have put ‘C’ in the units!"

But originally I had "F" as the units, and Torque was still treating it as a Celsius temperature. Which makes me think something else is going on.

"The temperature from the ECU is in C, not F. Your formula does not do the offset (-32) from the temperature in your original conversion, therefore you will be significantly out in the calculation."

I got the formula "A*(9/5)-40" by starting with a ScangaugeII custom PID reported to give good data in Fahrenheit with my vehicle, and then using Aberk's tutorial http://torque-bhp.com/forums/?wpforumaction=viewtopic&t=352.0 to convert to Torque extended PID. Then, assuming "A*(9/5)-40" is giving good Fahrenheit data, the same data in Celsius units should be

((A*(9/5)-40)-32)*5/9 --> A-40

... which is what you suggested above. Now, that formula "A-40" indeed seems to be giving very good Celsius numbers. At least, the ATF temp from this extended PID, and the engine coolant temp and the intake air temp from Torque's standard PID's, all agree within a degree or so when the engine is cold. (I don't have a good way of independently checking ATF temp when things warm up, but it seems to be about what I would expect.)

So I'm confused why you say I will be significantly out in the calculation. Should I be using something different from "A-40" now?

Hi

No, your original formula "A*(9/5)-40" is wrong (as you don't put the 32 degree offset into the equation).

The A-40 is correct, as the units from the ECU are in C.

To convert this to F, (ignoring the fact torque would double-convert) then you would need to add the 32 degree offset into the equation, otherwise you would be 32 degrees off the real reading - ((A-40)*(9/5))+32 should give you F.

For example, try:
http://www.wbuf.noaa.gov/tempfc.htm

with values of 0 and 10 in the celsius box.

But for Torque, A-40 in the equation, with C as the units, and you should get the correct reading.

Hope this helps

Ian

"((A-40)*(9/5))+32 should give you F"

Right. But then multiplying out,

((A-40)*(9/5))+32 =
(A*(9/5)-40*(9/5))+32 =
(A*(9/5)-72)+32 =
A*(9/5)-40

So, if "A-40" is correct for C, why wouldn't "A*(9/5)-40" be correct for F, since it is numerically equal to "((A-40)*(9/5))+32".

--Paul

Yes, you've managed to totally confuse me and I got the formula in the previous post wrong 🙂

I believe I meant to say: ((9/5)*(A-40))+32 which would convert from celsius direct from the ECU to fahrenheit