RomRaider

[EDIT]

I'm an idiot...lol I read the formula for the chart completely wrong in the spread sheet, disregard, I should only attempt to read things when I'm not falling asleep

[/edit]

I realize that CarBerry is a 1 man operation, and does not have the same size of staff of Cobb...

Using the GM IAT seems to be a very popular choice, however, the issue I have is because of the current limitations of open source 16-bit roms, it is not possible to edit the voltage reference points that the ECU uses to determine temperature, this unfortunately makes the use of the GM IAT inaccurate in some areas.

Now the people at Cobb have figured out how to edit the voltage reference points for 16-bit ecus, so I figure it should be possible for open source guys as well

Credit for this graph goes to ryanr, who came out with the amazing spreadsheet found here: viewtopic.php?f=32&t=7797&start=30

Looking at the chart here:



We can see that because of the way the 16-bit OS roms require the fixed voltage that is pre-determined in the ECU, the way the curves go are different, and the accuracy becomes quite poor below 100F (37.78C) and above 140-160F (60C to 71C) Now I don't think there is much of an issue for the above the 140-160F accuracy, as if your car has IAT that high, it's time to consider better intercooling solutions

However, below that could be a problem for cold starts or those who have severe winters (at worse it only gets down to 0C where I live...but I've lived in -40C (also -40F) climates)

Now I can't see this being too catastrophic for people, but I would imagine it would cause the car to run quite rich, at these lower temps, or even running boost at lower temps with very efficient IC systems, which isn't a bad thing and is much better than running too lean, but it would reduce power at these lower temps, and it isn't optimal.

Of course I could just be over thinking this issue, and there may be no real issue, and that even with the odd scaling, the GM IAT could still be more accurate than the stock IAT (in blow through config) due to other variable factors.

Dude, those curves are different because the temp sensors used are different! LOL.

The GM = GM sensor with calculated scaling from supplied resistances
Cobb = GM/AEM sensor with their modified scaling
Carberry = JDM Spec C sensor (most likely)

I have Bosch combo IAT+MAP sensor installed in manifold custom hole.
http://www.bosch.com.au/content/languag ... cation.pdf
How can I calibrate IAT Sensor?
2.5 K ohm / 68 degree is only data in Bosch data sheet.

Just did a bit of looking around for you and tried to adapt my spreadsheet to the Bosch TMAP like you have and the resistances that I could find are VERY high. This puts the entire scale of the sensor off (ex. at 4.74v it would be at 175deg F) This is the only scaling that I could find and I found it in 3-4 places but this very well may be for the LS application.

http://www.edelbrock.com/automotive_new ... tmap.shtml

Are you able to read the temperature off of that? It looked to me that the temperature was for pressure compensation vs. temperature?

Crap, nm, I read the excel formula wrong...wow...no more late nights for me after studying trying to decipher excel spread sheets lol...

Yeah, the sensor is too sensitive below 150F even after adding a 10Kohm resistor in parallel. What we really need to do is find a way to change the pullup resistor in the ECU to something in the 10K range. Then this will work perfectly.

Thank you for scaling spreadsheet.
I hope my SD compensation table will be returned on previous scaling. It has been totally rescaled after installing this TMAP.

So has anyone actually got this working? I used a gmiat sensor at one point post intercooler and it read off but i didn't have anything else.

Did anyone resolve editing of reference voltage ranges?

Why do you want to edit them? They are scaled pretty even throughout the entire 0-5v range, you just need to edit the temperatures they correspond to.

As I mentioned above, I have Bosch TMap sensor and IAT sensor scaling is totally differ from subaru IAT sensor, so if change only temperature values, it will be match only 5 points of voltages and most part of temp ranges will be out of table.

Hi all

I know this is an old thread, but im looking at the 3Bar Bosch TMap sensor now.

It seems there's new sensor calibration data for the temp sensor, than what is shown in here though.

Based on the Calibration chart from Bosch (https://www.efisolutions.com.au/assets/ ... 31-030.jpg) the voltages (based on the spreadsheet and formulas posted above) are much more reflective of accurate temps...

I get these voltages for the supplied range of temps/resistances, assuming a 5v supply and 1229ohm internal resistor.


I think I'm going to give it a go and see if I can get it working irl.