XGauge (CAN-bus) for Dummies...

EDIT:

• Excel file with all the codes I hacked together.
• [POST="115258"]Jump to post 21[/POST] where I explain how to program the XGauge.

Ok... I feel pretty stupid but I finally get it now. I've been trying for weeks to get my head around what XGauge was doing. Took WFEC and the revelation that we were flying blind without Battery Amps on the Prius to focus my efforts.

Well here's what I found. First let qualify that this is CAN-bus only. Since I'm a Prius Pilot that's all I've been focused on. So what I've found out is that there are two types of CanBus requests. Passive (snooping the bus) and active. Passive is kinda like ease-dropping on a conversation... the data is just there. Passive is like asking a direct question and expecting a direct answer.

Now after doing a bit of digging there is lots of data out there for the taking... Here's what I "think" we can get:

Passive
L/R GForce = TXD { 0x002A }
F/B GForce = TXD { 0x002B }
Steering degrees L/R = TXD { 0x002D }
Brake engagement % = TXD { 0x0038 }
Speed (decameters / hour ) = TXD { 0x00BC }
Gas pedal position = TXD { 0x024C }
Battery Temp = TXD { 0x03C3 }
Ignition timining = TXD { 0x052E }
EV button (on/off) = TXD { 0x0521 }
Gas Fill Level = TXD { 0x05AC }

Active
Battery SoC = TXD { 0x07E321CE }
Battery Temp = TXD { 0x07E321CF }
MG 1/2 RPM = TXD { 0x07E221C3 }
MG 1/2 RPM = TXD { 0x07E221C4 }

Now there is some other stuff like Headlights (on/off) and stuff like that that I'm not that interested in, but this data all looked pretty cool. I haven't gotten that deep into it so don't pepper me with Q's yet, but please help my "Teach Dan XGauge" thread.

If you look for the Passive commands and you can't find them, it's because the passive TXD is derived by taking the ID, left padding it with zeros and XOR'ing it wish 0x0008. Now all this is Attila Vass data, so I don't know if Ron has already tried and failed to get these to work, or if he just figured, no prius driver in his right mind would be interested in how many lateral G's he pulls in a turn. I'm obviously not in my right mind

GForce sensor is the first one I want to try to get working!

Now once I get the TXD working we can work on the rest... Here's my current reference set:

Attila Vass Can Dump (Excel)
2way's XGauge Coding walkthru
PriusChat XGauge Coding walkthru (Passive Can)
Prius Tech Info
PriusChat Known SGII Glide Bug​

11011011

 

Great thread dan. Can we also write to the can bus. What I am thinking is, can we command engine off when in Neutral. Thus having a glide above 42 MPH.

 

Alright... First new XGauge is tested and Verified.

EV Button ( On / Off )
TXD: 05 21
RXF: 01 05 22 29 00 00
RXD: 31 01
MTH: 00 01 00 01 00 00
NAM: EVb

What you'll see on your Gauge is
EVb ON
EVb OFF

And the fun part... It's real-time. No polling delay since it's packet sniffing. I leave it as a homework assignment to decode the logic I used to arrive at the XGauge coding.

11011011

 

Looks like your doing what I had planed to do. I just got my SG yesterday. After working with the RAW can-view data and the can-decoder software I'm now familiar with the raw CAN hex data and I know how to convert it.

I'm Not familiar with the way the xgauge does the converting yet. I think I will need an example of how you get from Vass's info to what the SG displays. I'm sure its simple.

I'm really interested seeing the Gforce in real time.

 

Look through the refereces I listed in the OP and all the XGauge stuff is pretty well explained.

Here's the next dump of XGauges. These are blind codes (untested) so I don't know which work:

Battery Temp (Max)
TXD: 03 C3
RXF: 01 03 02 3B 00 00
RXD: 38 08
MTH: 00 01 00 01 00 00
NAM: BTc

Battery Temp (Min)
TXD: 03 C3
RXF: 01 03 02 3B 00 00
RXD: 30 08
MTH: 00 01 00 01 00 00
NAM: BTc

Left / Right GForce
TXD: 00 2A
RXF: 01 00 02 22 00 00
RXD: 10 10
MTH: 00 01 00 01 FE 00
NAM: LRg

Front / Back GForce
TXD: 00 2B
RXF: 01 00 02 23 00 00
RXD: 10 10
MTH: 00 01 00 01 FE 00
NAM: FBg

Alignment (stear left / right)
TXD: 00 2D
RXF: 01 00 02 25 00 00
RXD: 10 10
MTH: 00 01 00 01 00 00
NAM: Slr

Break Pedal Position
TXD: 00 38
RXF: 01 00 02 30 00 00
RXD: 30 08
MTH: 00 64 00 80 00 00
NAM: BPs %

11011011

 

I just tried out the G force setting while sitting still and here is what I seen.

LRg stays at about 3 but if I shake the car the number jumps around a bit.

FBg stays at 1 even when I shake the car.

I will be leaving for work in about an hour and I will report back with my trip results.

 

Yeah, it's just a flat hex value. I haven't scaled it to actuall GForce. It should go way positive on a hard left and way negative on a hard right.

11011011

 

OK.. got another two coded up but untested....

Speed down to the hundredth place. I always disliked the fact that the speedo doesn't have a decimal point. Well now it does. The MTH was tricky on this so I'll actually be surprised if I got it spot on with the first attempt.

kph xx.xx
TXD: 00 BC
RXF: 01 00 42 B4 00 00
RXD: 38 10
MTH: 03 E8 04 00 00 00
NAM: kph xx.xx

mph xx.xx
TXD: 00 BC
RXF: 01 00 42 B4 00 00
RXD: 38 10
MTH: 08 D8 0E 93 00 00
NAM: mph xx.xx

 

Gforce does work. Only problem is that the SG display rate is way too fast. You might have to divide the data by something to get it to slow down or to get an actual 0.00 gforce value.

 

All these passive commands are gonna refresh wicked fast. Can't be helped... that's how many canbus packets are flooding the bus. You could do some creative binary shift operations to clip off the last 4 bits or something. That should slow the jitter.

11011011

 

OK... have three left to do but

Left / Right GForce
TXD: 00 2A
RXF: 01 00 02 22 00 00
RXD: 10 10
MTH: 00 01 00 01 FE 00
NAM: LRg

Front / Back GForce
TXD: 00 2B
RXF: 01 00 02 23 00 00
RXD: 10 10
MTH: 00 01 00 01 FE 00
NAM: FBg

Alignment (stear left / right)
TXD: 00 2D
RXF: 01 00 02 25 00 00
RXD: 10 10
MTH: 00 01 00 01 00 00
NAM: Slr

Break Pedal Position
TXD: 00 38
RXF: 01 00 02 30 00 00
RXD: 30 08
MTH: 00 64 00 80 00 00
NAM: BPs %

Battery AMPs (passive)
TXD: 00 33
RXF: 01 00 82 3B 00 00
RXD: 14 0C
MTH: 00 01 00 01 00 00
NAM: bta

kph xx.xx
TXD: 00 BC
RXF: 01 00 42 B4 00 00
RXD: 38 10
MTH: 03 E8 04 00 00 00
NAM: kph xx.xx

mph xx.xx
TXD: 00 BC
RXF: 01 00 42 B4 00 00
RXD: 38 10
MTH: 08 D8 0E 93 00 00
NAM: mph xx.xx

tps %
TXD: 02 4C
RXF: 01 02 82 44 00 00
MTD: 40 08
MTH: 00 0A 00 02 00 00
NAM: tps

Battery SOC (passive)
TXD: 03 C3
RXF: 01 03 02 3B 00 00
RXD: 20 10
MTH: 00 0A 00 02 00 00
NAM: SoC

Battery Temp (Max)
TXD: 03 C3
RXF: 01 03 02 3B 00 00
RXD: 38 08
MTH: 00 01 00 01 00 00
NAM: BTc

Battery Temp (Min)
TXD: 03 C3
RXF: 01 03 02 3B 00 00
RXD: 30 08
MTH: 00 01 00 01 00 00
NAM: BTc

Battery Volt
TXD: 03 C5
... todo

IGN (passive)
TXD: 05 2E
... todo

WTc (passive)
TXD: 05 25

Gas Gauge
TXD: 05 AC
... todo

EV Button Status
TXD: 05 21
RXF: 01 05 22 29 00 00
RXD: 31 01
MTH: 00 01 00 01 00 00
NAM: EVb

More later....

11011011

 

Thanks for all the hard work, Dan!

 

ARGH! I can't take it anymore, I need a SGII so that I can play with this myself.

 

After looking at the g-force data in real time could the actual g-force readout be the number currently displayed divided by 100 or 200? Dividing by 10 on the values I seen and have captured with my can-view would result in a high number for such low actual g-force. Dividing by 100 gets us to what looks like a real number.

According to Vass's data 0x200 is neutral. If going to a negative value of 0x000 that means there is only room for 0x200 of play in g-force. Converted to decimal that would be 512. If you divide 512 by 100 you get 5.12 or a max reading of 5.12 g-force in each direction. This number seems to be about right for measuring the g-force a car will experience. Actually if you was pulling 5g in a car that's a pretty good indication your currently in a crash.

I'm still looking into understanding the xgauge so Dan if you want to you may want to divide the g-force data by 100 unless you think it needs to be another divisor. I think I will update my can-view decoder program to do the division of 100.

 

Cool... Thanks for the test data. I keep thinking of Prius Pilots doing donuts in a mall parking lot trying to calibrate the G-Force gauge

Here's the updated gauge that just thunks a decimal where you suggested.

GForce to Right x.xx
TXD: 00 2A
RXF: 01 00 42 22 00 00
RXD: 10 10
MTH: 00 01 00 01 FE 00
NAM: gfr

GForce to Back x.xx
TXD: 00 2B
RXF: 01 00 42 23 00 00
RXD: 10 10
MTH: 00 01 00 01 FE 00
NAM: gfb

Stearing (degrees left) xx.xx%
TXD: 00 2D
RXF: 01 00 42 25 00 00
RXD: 14 0C
MTH: 00 64 01 BC 00 00
NAM: sdl

Break Pedal Position xx.x %
TXD: 00 38
RXF: 01 00 22 30 00 00
RXD: 30 08
MTH: 00 64 00 80 00 00
NAM: bps

Battery AMPs (passive)
TXD: 00 33
RXF: 01 00 82 3B 00 00
RXD: 14 0C
MTH: 00 01 00 01 00 00
NAM: bta

kph xx.xx
TXD: 00 BC
RXF: 01 00 42 B4 00 00
RXD: 38 10
MTH: 03 E8 04 00 00 00
NAM: kph

mph xx.xx
TXD: 00 BC
RXF: 01 00 42 B4 00 00
RXD: 38 10
MTH: 08 D8 0E 93 00 00
NAM: mph

tps xx.x%
TXD: 02 4C
RXF: 01 02 82 44 00 00
MTD: 40 08
MTH: 00 0A 00 02 00 00
NAM: tps

Battery SOC (passive) xx.x %
TXD: 03 C3
RXF: 01 03 82 CB 00 00
RXD: 20 10
MTH: 00 0A 00 02 00 00
NAM: soc

Battery Temp degC (Max / Upper)
TXD: 03 C3
RXF: 01 03 02 3B 00 00
RXD: 38 08
MTH: 00 01 00 01 00 00
NAM: buc

Battery Temp degC (Min / Lower)
TXD: 03 C3
RXF: 01 03 02 3B 00 00
RXD: 30 08
MTH: 00 01 00 01 00 00
NAM: blc

Battery Volt (3CD pid)
TXD: 03 C5
RXF: 01 03 02 CD 00 00
RXD: 20 10
MTH: 00 01 00 01 00 00
NAM: btv

IGN (passive)
TXD: 05 2E
RXF: 01 05 02 26 00 00
RXD: 18 10
MTH: 00 01 00 01 00 00
NAM: ign

WTc (passive) xx.x
TXD: 05 25
RXF: 01 05 82 2C 00 00
RXD: 18 08
MTH: 00 0A 00 02 00 00
NAM: wtc

Gas Gauge xx.x %
TXD: 05 AC
RXF: 01 05 82 A4 00 00
RXD: 18 08
MTH: 00 64 00 04 00 00
NAM: gas

EV Button Status (on/off)
TXD: 05 21
RXF: 01 05 22 29 00 00
RXD: 31 01
MTH: 00 01 00 01 00 00
NAM: evb
​

I'm working on beautifying this into excel right now.

11011011

 

Hmmmmm That gives me an idea. I wonder if I can get one of my digital scales to work vertically? Tape a 1lbs weight to it, zero it out, do some doughnuts and compare the weight seen with the g-force data. Going in circles .5lbs would be .5g. I might give that a try sometime.

 

Allright... I've uploaded an Excel file with all the codes I hacked together. My list is by no means exhaustive, but it's better than the dinky three they originally had.

Column1	Column2	Column3	Column4	Column5	Column6	Column7
GAUGE	TXD	RXF	RXD	MTH	NAM	Notes
State of Charge	07E321CE	056186CE0000	3008	000A00020000	SOC	%
Battery Voltage	0033	0100023B0000	2010	000100010000	BTV	HV Battery Volts
Battery Current	07E321CE	056186CE0000	3810	0001000AF333	BTA	HV Battery Amps
GForce to Right	002A	010042220000	1010	00010001FE00	gfr	x.xx G's scaling a guess
GForce to Back	002B	010042230000	1010	00010001FE00	gfb	x.xx G's scaling a guess
Stearing	002D	010042250000	140C	006401BC0000	sdl	Degrees left xx.xx %
Brake Pedal Position	0038	010082300000	3008	006400800000	bps	xx.x % depression
Battery Current	0033	0100823B0000	140C	000100010000	bta	Passive CAN request
Kilometers per Hour	00BC	010042B40000	3810	03E804000000	kph	xx.xx kph
Miles per Hour	00BC	010042B40000	3810	08D80E930000	mph	xx.xx mph
Gas Pedal Position	024C	010282440000	4008	000A00020000	gps	xx.x % depression
State of Charge	03C3	010382CB0000	2010	000A00020000	soc	Passive CAN request
Battery Temp (upper read)	03C3	0103023B0000	3808	000100010000	buc	°C max for the 5 ms cycle
Battery Temp (lower read)	03C3	0103023B0000	3008	000100010000	blc	°C min for the 5 ms cycle
Battery Voltage	03C5	010302CD0000	2010	000100010000	btv	3CD pid reading
Ignition degrees before TDC	052E	010502260000	1810	000100010000	ign	Passive CAN request
EV Button Status	0521	010522290000	3101	000100010000	evb	EV button ON or OFF
Engine Coolant Temp	0525	0105822C0000	1808	000A00020000	wtc	xx.x °C
Gas Gauge	05AC	010582A40000	1808	006400040000	gas	xx.x % full

​

11011011

 

Updated the Excel file with all the codes.

Code:

CHANGELOG:

v1.0 - 06-29-08: Initial Release.
v1.1 - 06-29-08: Added Download Link.
v1.2 - 06-29-08: Corrected Brake Pedal Position
                 Added °F for all the °C readings
                 Added Engine temp gauge (not coolant temp)
                 Added Jitter Reduction (clipped last bit)
                 Added Jitter Reduction (clipped last 2 bits)

The Jitter reduction has to do with how fast the gauges refresh. The digits blur, so the x2 reduction clips the last bit, so you only get refreshes (on average) half the time. The x4 reduction clips the last two bits, so you only get a refresh (on average) 25% of the time.

11011011

 

Awsome! I just figured out how to get the TXD and MTH. I'm having a little trouble understanding RXF and RXD.