new coolant temp and intake air temp entries for EvoScan
Mar 7, 2008, 06:23 PM
#1
new coolant temp and intake air temp entries for EvoScan
Hey Folks,
After Hamish included revised scalings for the coolant temp and IAT entries in EvoScan 2.1, I decided to do a bit more sniffing through the ROM to see if I could find the reason for revised scalings. Hamish was onto something. As it turns out, for both IAT and coolant temp, the raw temperature sensor data and the scaled temperature data are both available in the stock MUT calls. Hamish used 1.8*x+32 for the scaling in deg F which is not quite correct. The scaled values, expressed in deg C are "x-40". To log in deg F, the scaling would be "1.8*x-40". The MUT addresses Hamish gave are incorrect too. Raw coolant temp is at MUT 07, and raw IAT is at MUT 3A. Scaled coolant temp is at MUT 10, and scaled IAT is at MUT 11. So, the correct entries for the EvoScan 2.1 "Data.xml" file are:
<DataListItem DataLog="Y" Color="" Display="Air Temp" LogReference="AirTemp" RequestID="11" Eval="1.8*x-40" Unit="deg F" MetricEval="x-40" MetricUnit="deg C" ResponseBytes="1" GaugeMin="-40" GaugeMax="260" ChartMin="-40" ChartMax="260" ScalingFactor="1" Notes="" Priority="1" Visible="false" />
<DataListItem DataLog="Y" Color="" Display="Coolant Temp" LogReference="CoolantTemp" RequestID="10" Eval="1.8*x-40" Unit="deg F" MetricEval="x-40" MetricUnit="deg C" ResponseBytes="1" GaugeMin="-40" GaugeMax="260" ChartMin="-40" ChartMax="260" ScalingFactor="1" Notes="" Priority="1" Visible="false" />
The old setup was accurate from about 60F to 250F. This new setup will be accurate from -40F to 250F, so for logs done under normal operating conditions, there won't be any significant change in logged temperature. The nice part about the new setup is the easier formula, and the also the values that are logged will be the exact values that the ECU is seeing.
After Hamish included revised scalings for the coolant temp and IAT entries in EvoScan 2.1, I decided to do a bit more sniffing through the ROM to see if I could find the reason for revised scalings. Hamish was onto something. As it turns out, for both IAT and coolant temp, the raw temperature sensor data and the scaled temperature data are both available in the stock MUT calls. Hamish used 1.8*x+32 for the scaling in deg F which is not quite correct. The scaled values, expressed in deg C are "x-40". To log in deg F, the scaling would be "1.8*x-40". The MUT addresses Hamish gave are incorrect too. Raw coolant temp is at MUT 07, and raw IAT is at MUT 3A. Scaled coolant temp is at MUT 10, and scaled IAT is at MUT 11. So, the correct entries for the EvoScan 2.1 "Data.xml" file are:
<DataListItem DataLog="Y" Color="" Display="Air Temp" LogReference="AirTemp" RequestID="11" Eval="1.8*x-40" Unit="deg F" MetricEval="x-40" MetricUnit="deg C" ResponseBytes="1" GaugeMin="-40" GaugeMax="260" ChartMin="-40" ChartMax="260" ScalingFactor="1" Notes="" Priority="1" Visible="false" />
<DataListItem DataLog="Y" Color="" Display="Coolant Temp" LogReference="CoolantTemp" RequestID="10" Eval="1.8*x-40" Unit="deg F" MetricEval="x-40" MetricUnit="deg C" ResponseBytes="1" GaugeMin="-40" GaugeMax="260" ChartMin="-40" ChartMax="260" ScalingFactor="1" Notes="" Priority="1" Visible="false" />
The old setup was accurate from about 60F to 250F. This new setup will be accurate from -40F to 250F, so for logs done under normal operating conditions, there won't be any significant change in logged temperature. The nice part about the new setup is the easier formula, and the also the values that are logged will be the exact values that the ECU is seeing.
Mar 8, 2008, 03:49 AM
#3
Evolving Member
mrfred - nice catch. I did some logging and can confirm that it works 
For IAT MUT 3A (x) gives the same values as MUT 11 (x-40).
For coolant they seem to be more apart - MUT 07 (x) is higher by cca 6-7 degC than MUT 10 (x-40).
This was tested during on a warm engine and outside temp was cca 8 degC.
For IAT MUT 3A (x) gives the same values as MUT 11 (x-40).
For coolant they seem to be more apart - MUT 07 (x) is higher by cca 6-7 degC than MUT 10 (x-40).
This was tested during on a warm engine and outside temp was cca 8 degC.
Last edited by evo828; Mar 8, 2008 at 05:34 AM.
Mar 8, 2008, 06:04 PM
#5
Just some corrections on the above information...
EvoScan values will currently show -104 to 158 on CoolTemp and -59 to 184 on AirTemp (degC)
x is not what you think it is. x is from my internal lookup table using the RAW value using MUT07 and MUT3A. which should be more readily available than MUT10 or MUT11 across more vehicles. Do you know which vehicles have MUT10 and MUT11 in them? I'm thinking that not all vehicles support MUT10 and MUT11, hence I think evoscan has it pretty much right in v2.1, I understand that other mitsi loggers are still behind, and using the old inaccurate polynomal function.
Interesting that by adding the lookup table to EvoScan v2.1 instead of that polynomial function, I have got it pretty much spot on, and allowed you to easily find the Scaled version of this value. nice work. If people want to take a look at the scaled versions in MUT10 or MUT11, I'd be weary.
If you can find out how far reaching the MUT10/11 is on vehicles, I may consider making it a more permanent fixture of EvoScan.
EvoScan values will currently show -104 to 158 on CoolTemp and -59 to 184 on AirTemp (degC)
x is not what you think it is. x is from my internal lookup table using the RAW value using MUT07 and MUT3A. which should be more readily available than MUT10 or MUT11 across more vehicles. Do you know which vehicles have MUT10 and MUT11 in them? I'm thinking that not all vehicles support MUT10 and MUT11, hence I think evoscan has it pretty much right in v2.1, I understand that other mitsi loggers are still behind, and using the old inaccurate polynomal function.
Interesting that by adding the lookup table to EvoScan v2.1 instead of that polynomial function, I have got it pretty much spot on, and allowed you to easily find the Scaled version of this value. nice work. If people want to take a look at the scaled versions in MUT10 or MUT11, I'd be weary.
If you can find out how far reaching the MUT10/11 is on vehicles, I may consider making it a more permanent fixture of EvoScan.
Mar 8, 2008, 08:30 PM
#6
Just some corrections on the above information...
EvoScan values will currently show -104 to 158 on CoolTemp and -59 to 184 on AirTemp (degC)
x is not what you think it is. x is from my internal lookup table using the RAW value using MUT07 and MUT3A. which should be more readily available than MUT10 or MUT11 across more vehicles. Do you know which vehicles have MUT10 and MUT11 in them? I'm thinking that not all vehicles support MUT10 and MUT11, hence I think evoscan has it pretty much right in v2.1, I understand that other mitsi loggers are still behind, and using the old inaccurate polynomal function.
Interesting that by adding the lookup table to EvoScan v2.1 instead of that polynomial function, I have got it pretty much spot on, and allowed you to easily find the Scaled version of this value. nice work. If people want to take a look at the scaled versions in MUT10 or MUT11, I'd be weary.
If you can find out how far reaching the MUT10/11 is on vehicles, I may consider making it a more permanent fixture of EvoScan.
EvoScan values will currently show -104 to 158 on CoolTemp and -59 to 184 on AirTemp (degC)
x is not what you think it is. x is from my internal lookup table using the RAW value using MUT07 and MUT3A. which should be more readily available than MUT10 or MUT11 across more vehicles. Do you know which vehicles have MUT10 and MUT11 in them? I'm thinking that not all vehicles support MUT10 and MUT11, hence I think evoscan has it pretty much right in v2.1, I understand that other mitsi loggers are still behind, and using the old inaccurate polynomal function.
Interesting that by adding the lookup table to EvoScan v2.1 instead of that polynomial function, I have got it pretty much spot on, and allowed you to easily find the Scaled version of this value. nice work. If people want to take a look at the scaled versions in MUT10 or MUT11, I'd be weary.
If you can find out how far reaching the MUT10/11 is on vehicles, I may consider making it a more permanent fixture of EvoScan.
MUT 10/11 holds true for the 94170008 ('03 USDM Evo 8 ROM). I'm sure it holds for everything in between. Bez' Evo 7 disassembly shows it holding true there too. If you are going to continue using the internal scaling, could you let us know what wording or syntax triggers the use of the internal scaling for coolant temp and air temp (and any other internal scalings)? For the time being, will these definitions not invoke the internal scaling?:
<DataListItem DataLog="Y" Color="" Display="Air Temp Scaled" LogReference="AirTempScaled" RequestID="11" Eval="1.8*x-40" Unit="deg F" MetricEval="x-40" MetricUnit="deg C" ResponseBytes="1" GaugeMin="-40" GaugeMax="260" ChartMin="-40" ChartMax="260" ScalingFactor="1" Notes="" Priority="1" Visible="false" />
<DataListItem DataLog="Y" Color="" Display="Coolant Temp Scaled" LogReference="CoolantTempScaled" RequestID="10" Eval="1.8*x-40" Unit="deg F" MetricEval="x-40" MetricUnit="deg C" ResponseBytes="1" GaugeMin="-40" GaugeMax="260" ChartMin="-40" ChartMax="260" ScalingFactor="1" Notes="" Priority="1" Visible="false" />
Last edited by mrfred; Mar 8, 2008 at 08:43 PM.
Mar 8, 2008, 09:10 PM
#7
I am supporting about 80 different ecus from 1989 thru to 2006, so I'm aiming to keep it as accurate as possible. Evo7/Evo8 are closely related, I will get some feedback from other makes and models, I'll let you know the outcome. At this stage I am correctly reporting AirTemp and Coolant temp in EvoScan v2.1 for both degC and DegF.
the internal scaling is invoked based on the request id 3A and 07. I'll put a note in the notes box regarding this. If you see a Temp that doesn't seem accurate, send me the ECU ID displayed in EvoScan, and I'll provide an update.
in the future I may change the "x" to something like "AirTemp[x]"
would that be better? so that you can continue using the raw value for developers?
the internal scaling is invoked based on the request id 3A and 07. I'll put a note in the notes box regarding this. If you see a Temp that doesn't seem accurate, send me the ECU ID displayed in EvoScan, and I'll provide an update.
in the future I may change the "x" to something like "AirTemp[x]"
would that be better? so that you can continue using the raw value for developers?
Trending Topics
Mar 8, 2008, 09:36 PM
#8
I am supporting about 80 different ecus from 1989 thru to 2006, so I'm aiming to keep it as accurate as possible. Evo7/Evo8 are closely related, I will get some feedback from other makes and models, I'll let you know the outcome. At this stage I am correctly reporting AirTemp and Coolant temp in EvoScan v2.1 for both degC and DegF.
the internal scaling is invoked based on the request id 3A and 07. I'll put a note in the notes box regarding this. If you see a Temp that doesn't seem accurate, send me the ECU ID displayed in EvoScan, and I'll provide an update.
in the future I may change the "x" to something like "AirTemp[x]"
would that be better? so that you can continue using the raw value for developers?
the internal scaling is invoked based on the request id 3A and 07. I'll put a note in the notes box regarding this. If you see a Temp that doesn't seem accurate, send me the ECU ID displayed in EvoScan, and I'll provide an update.
in the future I may change the "x" to something like "AirTemp[x]"
would that be better? so that you can continue using the raw value for developers?
Where did you get the scaling curve that you are using? The polynomials that were used with earlier versions of EvoScan diverge greatly from the scaled temperature in the ECU for temperatures below about 60F. The scaled temperature range the ECU reports can be as low as -40C and as high as 125C.
Last edited by mrfred; Mar 9, 2008 at 12:30 PM.
Mar 9, 2008, 12:22 PM
#9
I used a scaling table from mmcd in the Pocket PC and my test logger and it works well over the entire range of temperatures I could test. I just did an integer lookup/interpolation. Better if we use ECU values that are already linear if they are universal.
Thread
Thread Starter
Forum
Replies
Last Post
GST Motorsports
09+ Ralliart Engine/Turbo/Drivetrain
113
Feb 14, 2017 06:09 AM
