Toyota RAV4 Forums banner
21 - 40 of 43 Posts
Discussion starter · #22 ·
I was discussing on TorquePro group about the "known" custom PID that can help determine transmission fluid temps in units of Fahrenheit and it was identified by another that the formula that I've been using is over complicated.

((((A*256)+B)*(7/100)-400)/10)

When really a shorter formula works quicker.

(A-40)*1.8+32

When I run a test on both formulas within the TP app, you can see what the app has to do to solve the equation for both.

Image


I'm all for the "KISS" or "Keep it simple stupid" method so I'll be using this going forward. 😁

Image
 
If you can "live" with that simplified error, that great!
My S22+ just shrugs off those calculations :ROFLMAO:

PS: That "simplified" formula is not simple enough. Tell those guys that don't know basic arithmetic:
(A-40)*1.8+32 = A*1.8-40*1.8+32 = A*1.8-72+32 = A*1.8-40
 
Discussion starter · #25 ·
If you can "live" with that simplified error, that great!
My S22+ just shrugs off those calculations [emoji23]

PS: That "simplified" formula is not simple enough. Tell those guys that don't know basic arithmetic:
(A-40)*1.8+32 = A*1.8-40*1.8+32 = A*1.8-72+32 = A*1.8-40
Well shoot, guess call me stupid I guess! [emoji1787]

Have you tested that formula in TP yet?
 
I didn't tested it. The calculation is done by the Torque itself and that relies on the phone's CPU. Pooling the sensors probably takes more time than the calculation. I have an OBDII that can read Toyota's CANBUS at 500kbps, that's faster than normal OBDII protocols.

I wish I would find a document that says what the actual formula needs to be....
 
Discussion starter · #27 ·
I didn't tested it. The calculation is done by the Torque itself and that relies on the phone's CPU. Pooling the sensors probably takes more time than the calculation. I have an OBDII that can read Toyota's CANBUS at 500kbps, that's faster than normal OBDII protocols.

I wish I would find a document that says what the actual formula needs to be....
What are you currently using? Just bought the obdlinkMX+ it's soo much faster than my Bafx elm327

Image
Image
 
Discussion starter · #31 ·
[emoji1787]

That coupon helps but it is expensive. For how much I log, it was worth it to me.

What's this canbus stuff you speak of? What are you logging?
 
Discussion starter · #34 ·
If you can "live" with that simplified error, that great!
My S22+ just shrugs off those calculations [emoji23]

PS: That "simplified" formula is not simple enough. Tell those guys that don't know basic arithmetic:
(A-40)*1.8+32 = A*1.8-40*1.8+32 = A*1.8-72+32 = A*1.8-40
I stand corrected again.

The absolute simplest way is to use just A-40, and set the unit type to °C, then in Torques settings menu under the units heading set Torque to display values in Fahrenheit. The test result in the PID editor will display in Celsius but gauges in the Dashboards will be converted to Fahrenheit.

The "*1.8+32" is a just a conversion from Celsius to Fahrenheit.
Using A-40 as the formula appears to work correctly and I don't think it gets any simpler than that! I used both formulas for comparison.
Image
Image
Image
 
I am not sure if that's the correct formula. It "works" for that register value (A=111), but I can't tell if that's the exact value of the temperature. If the number it was 32 it would make sense, because 0 in C is 32 in F. The 40 seems to be random.

I wish it was something else besides trial and error.
 
Discussion starter · #36 ·
When I test either formula, the output is the same. I suppose you could be right and we've been using the wrong formula all along but at least the ouput is the same and the formula is shorter.
 
I've got a 2009 Rav4 Sport 2GR-FE V6 with the U151F AWD transmission.
Same here, and my daughter has an identical 2011.

1) For pre 2010 models
21D9
...
2) For 2010 and newer
2182
Only PID 2182 works for my 2009 and her 2011. IDK what is in register B and have not confirmed with a thermometer but values computed as A*1.792-40 seem plausible (example).
 
The formula: ((((A*256)+B)*(7/100)-400)/10) means that the B factor has a 256 times less influence than A (0.5%) so that's fine.
I just use the whole formula because I feel that my phone can handle a few calculations with integer numbers.
Note that using 1.792 means working with real numbers and that usually takes more processing that integers... IDK if there is a way to test it and why it would matter. The actualization is not happening faster or slower than the register read rate.
 
21 - 40 of 43 Posts