S1 CAN PIDs
 

Preface - I searched.

Anyone ever get around to logging the CAN PIDs from the S1?

I may be able to work on this in a month or two. I have the parts to build an Arduino CAN logger but I lack the time.

If you use OBD Fusion on your mobile device, you can buy the $10 add-on pack to get access to all of Mazda's private PIDs. They're listed by name so you'll have to jump through some packet sniffing hoops to match up names to actual PIDs

What are you trying to do? I have some of the CAN stuff.. Dash and wheel speed and stuff to make the power steering work. and have a CAN sniffer.... but don't have a functional car right now to do any logging

I hope to use a buddies car and do some logs at some point 5o see if I can figure out more stuff

Trying to get some useful information and a better throttle trace from RaceCapture MK2.

Just get a Reader that reads the Mazda extended PID's. Doesnt sound like you need the actual CAN info for what you want.

I’m sure many are common, but there are a few unique ones for RX8 not on here; they got a few from this forum:

https://forum.miata.net/vb/showthread.php?t=631418

Looks like the mother lode. Thanks~!

2004-

https://cimg9.ibsrv.net/gimg/www.rx8...aeacf28bf.jpeg
https://cimg6.ibsrv.net/gimg/www.rx8...3bdae1048.jpeg

2006-2008 w/changes-differences from 2005

https://cimg4.ibsrv.net/gimg/www.rx8...8245f4cd7.jpeg
https://cimg2.ibsrv.net/gimg/www.rx8...e55b9480c.jpeg
https://cimg3.ibsrv.net/gimg/www.rx8...fccb3fe92.jpeg

This link has proven to be accurate for my purposes:
Mazda CAN Bus ? Madox.NET

I don't have a Renesis or stock ECU, but I'm logging the individual wheel speeds to my RaceCapture. I'm also using the RaceCapture to send CAN data to the OEM gauge cluster via a Lua script - https://www.rx8club.com/series-i-eng...luster-268611/

Throttle position appears to be 0x201 byte 6. Scale is divide by 2.

i have AIM solo DL 2 and have tried different protocol to log the brake pressure but none seem to work.

I don't think the S1 ECU will communicate that (steering and break pressure) data over CAN. I am going to see about RPM, Brake (just on or off) and Throttle tonight.

Well maybe you’re confused; the only CAN data in a displayable format are the items listed in the Mazda PID/DATA list per the Definition and Unit/Condition parameters as indicated in my posts above.

Which neither of those are listed or formatted for what you’re seeking. You’d have to install the devices and setup the software parameters for both on your own to have that data.

Pages 20 and 22 of the electrical FSM show the steering angle sensor connected to the CAN bus on DSC cars. I have a non-DSC car, so I can't test it. But the link I shared above ID's the steering angle data @ 0x4DA bytes 0 and 1. I would bet that it's available there.

Just because data is not available via OBD, that doesn't mean that it's not floating around on the CAN-bus somewhere.

Well that’s assuming angle is actually being measured rather than some other value. It still might serve as a reference value, hard to say without the actual data. I’m aware of it for the ND MX5; not the NC/RX8, but then I don’t claim to be fully informed on the subject.

I’ll concede that it’s more likely for there to be a steering angle value than brake pressure value by virtue of the devices being used. It’s possible that some of the data didn’t exist until later in the production cycle, either during the S2 or after RX8 production ceased. The NC thread stated that they couldn’t get those to work.

I didn't see this thread.

I have a RaceCapture Pro Mk3, and I have the throttle position sensor mapped. If you like, I can send you the channel info when i get home tonight.

Incidentally, I've had nothing but trouble getting OBDII channels working. If i map even one, it messes with the car. Are you mapping any OBDII?

I did map OBDII but ended up gimping my racecapture somehow over the weekend.

OBDII maping works if the PIDs are in passive mode. I presume you had the gnarly engine cut outs above 8k?

Would love to see what you have for throttle trace on CAN.

need to improve those GoogleFu skills:

https://forum.autosportlabs.com/view...fb7e14105f3335

.

1. On that forum, Dadspeed is me.

2. It did not work, and Brent has spent the last year and a half not solving the problem.

I actually forgot entirely that I posted on that forum. I didn’t find it useful.

Is the throttle trace you get useful?

I use CAN for throttle, RPM and wheel speed. They are all very useful. This is what my CAN Mapping page looks like in Imax:


https://cimg2.ibsrv.net/gimg/www.rx8...37e4d9af12.jpg

Also note that the CAN channels are set to big endian.

One little, two little, three little endians...

That's fantastic. I had a thought that maybe some kind of math channel for steering angle could be made by calculating the difference between the two front wheels.

Steering angle has got to be available via CAN.

Interesting, this document shows what CAN PIDs AiM can log with their devices:

https://www.aimtechnologies.com/supp...08_105_eng.pdf

I wish I could get those parameters. I would really like to be able to display fuel level grrrrrrrr.

I'm going to try to get brake on/off working this week, hopefully it isn't too hard

I would be super happy to hear about it if you do. I have a brake pressure sender pipe, I just haven't gotten the sensor yet.

It's in the link I shared originally, just need someone with a DSC car to confirm it.

In racecapture format, steering angle in degrees should be (CAN ID, Offset(length), formula):
1242, 0(2), x1 /10 +(-3277)

Individual wheelspeeds that I'm logging are: (these come directly from the non-DSC ABS unit, I assume the DSC unit has the same IDs and scale. Also the formula is for km/h)
https://cimg1.ibsrv.net/gimg/www.rx8...badaecac50.jpg

Brake on/off is listed in a couple of places; 0x190 and 0x205. Looks like it should read 64 when the brake is on, and 0 when the brake is off. So in racecapture format it would be one or both of the following (CAN ID, Offset(length), formula):
400, 2(1), x1 /64 -0
517, 2(1), x1 /64 -0

I ask this because I'm a dummy, not because I'm being a smartass:

If the PIDs are the same between the DSC and non-DSC, then why are your CAN IDs different? Or am I logging an entirely different 'wheel speed' somehow? Honestly, I don't know. A vendor set up my PIDs for me because I couldn't figure it out.

The wheel speed PIDs that I listed are the individual wheel speeds that come directly from the ABS unit (CAN ID 1200). The factory ECU takes those values and does some averaging between multiple wheels, and maybe some scaling and smoothing, before sending a 'conditioned' wheel speed back out on a different ID (CAN ID 513). This conditioned wheel speed is what is used for the speedometer and to control the electric power steering. I know this because I had to re-create this conditioned wheel speed and send it out from the racecapture to get my EPS and speedometer to work (I've got an REW swap and no factory ECU).

The individual wheel speeds are really helpful to look at in the data - you can see when you're getting wheel spin, how the diff is working, when you're getting into the ABS under braking, etc.

Here's what I am able to log with an AIM solo on a 2004 grand touring (with traction and stability control).
I don't get any data for yaw rate, lat or long acceleration, hence I've crossed those out, but the rest give me valid data. There are more channels I can pick, but again, don't get data for.
The same channels also work for a 2010 s2 sport, except steering angle (I'm assuming that was because it didn't have traction and stability control)

TPS is the most useful to me for driver analysis. Really just seeing where the driver got on the gas and whether there is any lift or not. I would love to get brake pressure as well, but that is not available via the aim solo and I use longitudinal acceleration data as an alternative.

AIRT good to understand track conditions (higher temps result in slower laps as the engine makes less power).
MAF and AFR are helpful for tuning, or just making sure it is running safely (along with ECT).

Fuel level bounces around a lot. I have not found a good use for it, particularly since the car fuel starves in turns way before it is empty.

https://cimg3.ibsrv.net/gimg/www.rx8...09513659de.png

Yeah, I'll positively be adding individual wheel speed logging to mine. Thank you for sharing that, and for explaining the difference between them and the 'conditioned' speed. :)

Honestly, I just want to be able to read fuel level when the car is sitting still, or down a straight. Due to some reclassification I had to do, and by adding aero to the car, it's heavier. I'll probably go back to running it as empty as I can.

I'm sure with some packet sniffing I could find CAN fuel level. I just haven't had the time.

I agree that throttle position is super important. Lat, long and total G are the story. Throttle and brake pressure are the ones telling it. I'll be adding brake pressure to my RaceCapture sometime this season.

Thanks for your input. :)

Wish me luck
https://cimg2.ibsrv.net/gimg/www.rx8...479f1ddf85.png

No go on the brake, but the TPS and RPM and working wonderfully. It's nice to have a real time trace.

I know this works with the NC Miata saddle tank and I bought one for my RX8 to tie in with the AiM Pista dash. I’d have to go back to see what the S1 RX8 ohm limits are

ICM Fuel Sender

.

Try this for the brake:

https://cimg9.ibsrv.net/gimg/www.rx8...c62146e206.png

I did some CAN sniffing on my car tonight and found that guy. Works perfectly for me, shows a value of 1 when the brake pedal is pressed, and 0 when it's not. As mentioned, I have a non-DSC car with no factory ECU, so YMMV. I'm surprised that I still have this data without the ECU, so I assume the brake pedal switch is read by the ABS unit directly before being output to CAN.

Also, there is some sensor reading on ID 1072 (has a little bit of noise like a typical sensor reading would). I suspect that this could be fuel level, I'll put some gas in the car while logging this ID and see if it changes sometime.

I tried this guy this morning - works like a charm. Thanks for doing the dirty work! I can't sniff my canbus because for whatever reason that script gimps my racecapture MK2 track.

what protocol did you use? i have 06 sport with traction and stability control but i couldnt log individual wheel speed with aim solo DL.

I'll give that PID a try.

I noticed you have the sample rate pretty low. At 120mph, you cover 33.5ft in .1s. I'm refining my braking within a pretty narrow window on some tracks, so 10Hz wouldn't be nearly enough. 25Hz is still 13.4ft at 120mph, but I might start there.

Thank you!

Good point, I was thinking that an on/off didn't need a fast log rate - but paired with GPS data, it can show the brake initiation point really precisely.

It looks like the ABS unit is sending data from that PID @ 100hz. If I log for around 2 minutes, ID 530 shows up 12000 times. I'll set the racecapture to 50hz, no reason not to I guess.

Did some more CAN sniffing tonight and managed to find the fuel level output. I started with the fuel gauge @ 60%, started logging raw CAN data, dumped a 5 gallon can of gas into the car (first dumped ~1.4 gallons, then 1.6, then the last 2 gallons). Then I pasted the CAN data into Excel and filtered out everything but the first two bytes of ID 1072.

Ignore the "bytes 0 and 1 combined" graph. I thought that this was a 2 byte value initially, but it's not.
https://cimg6.ibsrv.net/gimg/www.rx8...f176c8749f.png

The scale for fuel level from byte 0 seems to be "full scale." I.E. 1 byte of data can represent a number from 0 to 255. So, it looks like 255 represents a full tank, and I'm assuming 0 represents an empty tank.

I'm not 100% sure what byte 1 is, but I'm going to log it from now on and see what it does. It looks like a delta between the two sides of the fuel tank. So 255 would represent the left side full and the right side empty, 0 would be both sides at the same level. Just a guess right now. It definitely has something to do with the fuel level because it responds to me putting gas in the car.

I put it into racecapture with the following format.
https://cimg8.ibsrv.net/gimg/www.rx8...84d2ab9b51.png
FuelGal is the fuel level from byte 0, I'm confident in that. Confirmed that it's reading 15.98 gallons with this setting as my car sits.
FuelDelta is experimental. Right now I'm treating it as the difference in fuel level between the two sides of the tank; with 100% being one full, one empty, and 0% being both at the same level.

You can change the fuel level scaling to view it in different units or whatever else. Some examples:
Fuel level in %: x100 /255 +0
Fuel mass in lbs: x102.6 /255 +0
Fuel level in liters: x64 /255 +0
Basically just change the multiplier to whatever a full tank is equal to in whatever units you want.

Disclaimer: I only tested it by dumping a bit of fuel into the car in the garage. I haven't confirmed that the scale works when the tank is empty.

This is amazing! Thank you! If it works, it's going to be a huge help to me. Right now fuel level is total guesswork! If I can start to log how much fuel I use per session at each track, I won't pace around wondering if I should put fuel in...and then put in two gallons...and then two more...and the fuel gauge still hasn't moved, so I put in one more, and it goes to 3/4. Dammit!

Does anybody know how to set this custom PID to a dashboard display? It doesn't show up in the list for, say, the tach/predictive timing display I use 99% of the time.

Thank you again!!!

Yeah I'm pretty excited about this one too. It seems to respond immediately to the fuel level changing, it's not laggy like the OEM gauge, so you should be able to key the car on and watch the fuel level in real time as you dump fuel into the car.

I was able to try it real quick this morning. I couldn't help it. And it appears to work! Thank you so much! And you're right - it appears to be instantaneous.

But I do want to change it to read tenths, instead of just whole gallons. I was able to get 7.4 to read as 74. Just need to figure out how to get the decimal in the right place. I may be a software developer, but I'm an artist. I am no programmer. :p

There is a field in the channel setup called "precision" that sets the number of decimal places. I set mine to 2 decimal places.

There it is. Thank you!

Are you going to tell Brent how to do this since he hasn't been able to figure it out?

What Brent hasn’t been able to figure out is getting OBDII PIDs to work. If we had known how to get fuel level via CAN, we would have just done that.

Does anyone have the individual fuel tank level sensor readings? The fuel level sensor outputs are wired into the dash and are then calculated by the dash display computer into an output for the dash level gauge indicator. I’m pretty sure the ohm output readings for both are on the CanBus though because you can display the current reading of each one on the dash check mode (Check Code 22).

I'm almost positive that the raw sensor readings are at CAN ID 1072, bytes 1 and 2 (byte 0 is the filtered overall fuel level, I.E. the fuel gauge reading). The raw sensor readings in CAN seem to be high when the tank is empty (close to 255) and low when the tank is full (close to 0). I think byte 1 is the drivers side sensor, and byte 2 is the passenger side sensor. I haven't had a chance to do more experimentation with these yet to fully confirm them, but I'm mostly sure that this is how it's reporting those values.

No, that sounds like the overall output reading from the dash display microcomputer to the fuel level gauge. There are two individual ohm readings from each level sensor to the dash display per the Code 22 check mode. The dash reads the two signals, calculates the fuel in each side, then displays a single output to the level gauge. The two level sensor ohm readings might not be on Can, but I was thinking possibly they were since the dash display can indicate each ohm reading individually uding the Check Mode dash display procedure. Honestly it doesn’t matter now. I already figured out how to get what I need using the Iron Canyon Motorsports ICM, which converts the sensor to a 0-5V output. I’m not using the RX8 dash display going forward; AiM Pista dash tied into an Adaptronic modular ECU CanBus. Thanks just the same.