Reading Tire Pressures Real Time via TPMS Sensors

[Pat]

While that might explain some of the discrepancies, it doesn't explain all of them. I've still seen meaningful differences between sensors on the same vehicle at a fixed altitude.
Also, explaining the difference is altitude doesn't really fix the problem. It's still a problem many people have to account for if they want accurate readings. I don't understand why people are so reluctant to check tire pressures manually. It takes almost no time or effort.

[NoHaveMSG]

Originally Posted by Pat

While that might explain some of the discrepancies, it doesn't explain all of them. I've still seen meaningful differences between sensors on the same vehicle at a fixed altitude.
Also, explaining the difference is altitude doesn't really fix the problem. It's still a problem many people have to account for if they want accurate readings. I don't understand why people are so reluctant to check tire pressures manually. It takes almost no time or effort.

I do check them manually. The problem is I would like to have an idea of what my tire pressures are peaking at as what I am reading in the pits isn't 100% accurate either because of a long cooldown lap at the track I frequent the most. I am interested more for the data and to have a live reference while driving.

[timurrrr]

Originally Posted by geraldjust

i actually got TPMS out of the can data from these cars. But its not as simple as sending some bytes and getting the return back. I can give ya a hint. its basically how Multi-Frame Message. If you google that you will get a semi aquatineace on how it works.

...

BTW responses should be in 0x758

This is a great pointer!
We would all really appreciate it if you could just share the exact request / response protocol.
I'm going to add it to my GitHub anyways, but you can just save everyone some time

I promise to read more into the details of how Multi-Frame Messages work!
I find "learn by studying an example" to be more effective than
"learn by first solving a puzzle that requires you to already know what you want to learn to solve it".

Also curious if you've confirmed if the same message format is working on the 2022 cars.

Originally Posted by Pat

Have you guys ever compared TPMS readings with readings from a high-quality traditional gauge? [...] But I think it's important to recognize there can be significant inaccuracies in them.

Totally. I definitely wouldn't rely solely on them, but they should give you a ballpark trend how much your tires gained in PSI since the start of the session.

Originally Posted by geraldjust

Keep in mind that the sensors are Absolute pressure

Interesting data!

Originally Posted by geraldjust

yeah i was worried about that too. but the sensors have a .5psi resolution

I have seen two different sensors at the exact same ambient conditions reading 2–3 psi off each other,
even though the same pressure gauge measured the same pressure from both tires.
So "0.5 psi resolution" doesn't necessarily mean "0.5 psi precision".
But again, as long as you can see the general trend...

Originally Posted by geraldjust

So i made mine also check temp too.

This is perhaps even more interesting to log!

[timurrrr]

Originally Posted by Pat

While that might explain some of the discrepancies, it doesn't explain all of them. I've still seen meaningful differences between sensors on the same vehicle at a fixed altitude.
Also, explaining the difference is altitude doesn't really fix the problem. It's still a problem many people have to account for if they want accurate readings. I don't understand why people are so reluctant to check tire pressures manually. It takes almost no time or effort.

I think doing both is better than doing either of them individually.
If only doing one, using a high quality pressure gauge is probably more precise,
as long as you fly into the pits at full speed, and your assistant quickly checks all four tires while they're hot.

[geraldjust]

Originally Posted by timurrrr

This is a great pointer!
We would all really appreciate it if you could just share the exact request / response protocol.
I'm going to add it to my GitHub anyways, but you can just save everyone some time

I promise to read more into the details of how Multi-Frame Messages work!
I find "learn by studying an example" to be more effective than
"learn by first solving a puzzle that requires you to already know what you want to learn to solve it".

Also curious if you've confirmed if the same message format is working on the 2022 cars.



Totally. I definitely wouldn't rely solely on them, but they should give you a ballpark trend how much your tires gained in PSI since the start of the session.



Interesting data!



I have seen two different sensors at the exact same ambient conditions reading 2–3 psi off each other,
even though the same pressure gauge measured the same pressure from both tires.
So "0.5 psi resolution" doesn't necessarily mean "0.5 psi precision".
But again, as long as you can see the general trend...



This is perhaps even more interesting to log!

Interesting thing to find about two tires being at different temps / pressures but same ambient temp. it was a conundrum that only happen at daytime but not at night. i ended up finding out with this that a parked car whos tires face the sun warm up so much it can use a super noticeable difference in PSI. by resolution i mean yes by .5. but so far ive tested with a backup manual check of my in the valve stem and for me they been all accurate with 1psi. But i do have the oem sensors. im not 100% sure if the accuracy is maintained if using aftermarket sensor though as its just a added variable in accuracy.

also for what use is for simplicity. not saying you should rely on the sensors. but knowing the temp and psi help out a ton. i even did some measurements when heating the front tires when braking repeatedly with heavy breaking. Looking at the sensors and sure enough you can see the difference.

let me dig up the code and and il ask my business partner if i can release the info!

[timurrrr]

I FIGURED IT OUT!

To read tire pressures

Request 1:

ID: 0x750
Data: 2A 02 21 30 55 55 55 55

Provides a response:

ID: 0x758
Data: 2A 10 07 61 30 XX YY ZZ

The highlighted "1" means that this is the first frame of a multi-frame response.

To get the second frame, we need to send a "Continue to send request":

ID: 0x750
Data: 2A 30 00 05 55 55 55 55

Which provides the second frame of the response:

ID: 0x758
Data: 2A 21 WW 00 00 00 00 00

The XX, YY, ZZ and WW bytes in the response will contain the tire pressures for sensors #1, 2, 3 and 4.
I believe the encoding is:

pressure_in_kPa(x) = (2 * x) - 100
pressure_in_psi(x) = (0.29 * x) - 14.5

where "x" is the value of XX, YY, ZZ or WW.

However it's worth noting that ScanGauge seems to use a different formula:

pressure_in_psi(x) = (0.2977 * x) - 14.7

Given that the car is Japanese (metric), the "(2 * x) - 100" seems like a simpler / more plausible encoding than what ScanGauge uses.

To read tire temperatures
Similar thing, just slightly different bytes:

0x750 | 2A 02 21 16 55 55 55 55
0x758 | 2A 10 07 61 16 XX YY ZZ
0x750 | 2A 30 00 05 55 55 55 55
0x758 | 2A 21 WW 00 00 00 00 00

and formulas:

temperature_in_C(x) = x - 40
temperature_in_F(x) = (x * 1.8) - 40

Enjoy!

Thanks so much @BRZyJ for letting me play with his ScanGauge!

[geraldjust]

Originally Posted by timurrrr

I FIGURED IT OUT!

To read tire pressures

Request 1:


Provides a response:



The highlighted "1" means that this is the first frame of a multi-frame response.

To get the second frame, we need to send a "Continue to send request":



Which provides the second frame of the response:


The XX, YY, ZZ and WW bytes in the response will contain the tire pressures for sensors #1, 2, 3 and 4.
I believe the encoding is:

where "x" is the value of XX, YY, ZZ or WW.

However it's worth noting that ScanGauge seems to use a different formula:


Given that the car is Japanese (metric), the "(2 * x) - 100" seems like a simpler / more plausible encoding than what ScanGauge uses.

To read tire temperatures
Similar thing, just slightly different bytes:


and formulas:


Enjoy!

Thanks so much @BRZyJ for letting me play with his ScanGauge!

hey you gut it! congrats! for PSI its: (X+32)/4

And to get relative pressure then X+32)/4 - 14.5

so you understand how extended frames work!

[22R]

Originally Posted by NoHaveMSG

https://www.geraldjustprojects.com/p...category/cars/

I really do not need this but Dayum that it too neat. I have a Scangauge II but this is the shizzle. Ima put it on my wishlist.
Nice Website and equipment.

22R

[timurrrr]

Originally Posted by geraldjust

And to get relative pressure then: (X+32)/4 - 14.5

Looks like Autel is using ((X * 0.25) - 7.5) for psi,
which maps to ((X * 1.724) - 51.7) for kPa.