VDS calculated state vs. data available via CAN
When the VDS on a v1.5 Roadster gets reset, it loses the following information: Charge History Energy History Trip Meter Drive Time Wh/mile history graph I take this to mean that these things are calculated from info on the CAN bus and stored in the VDS. In v1.5 Roadsters, the energy history got screwed up by a firmware update more than a year ago. The regen energy numbers shown are often ludicrous. In extreme examples, the day and month regen numbers can be larger than the net energy numbers. No doubt related to that, I see the regen number going up when the car is parked. (That's a nice trick!) I take this to mean that something on the CAN bus changed but the VDS firmware hasn't been updated. Things that are preserved over a VDS reset: Charge time, mode and current limit. (Current limit is stored by location, presumably by the VMS.) Trip distance and net energy used. So, presumably all of the values in this group can be queried from the CAN bus. Trip efficiency also persists. It may be calculated by the VDS or available on the CAN bus. My most recent trip meter showed 46.7 miles, 11.32 kWh, and 243 Wh/mile. If you do the math, 11.32 kWh / 46.7 miles = 242.398 Wh/mi, which should round to 242. Since the trip distance has been found on the bus at 0.1 mi resolution, it seems like that's the best data the VDS has. If there are higher resolution numbers somewhere (VMS?), then maybe the real values are 46.65 miles and 11.324 kWh -> 242.74 Wh/mi which rounds up to 243. So, perhaps the efficiency number is calculated elsewhere and broadcast on the CAN bus. Alternatively, the VDS either may have a higher resolution energy number than what it's showing or perhaps it just does math badly. Tom
Tom, Answers, inline, below. Regards, Mark. On 24 Apr, 2012, at 1:42 AM, Tom Saxton wrote:
When the VDS on a v1.5 Roadster gets reset, it loses the following information:
Charge History Energy History Trip Meter Drive Time Wh/mile history graph
I take this to mean that these things are calculated from info on the CAN bus and stored in the VDS.
Yes, that makes sense.
In v1.5 Roadsters, the energy history got screwed up by a firmware update more than a year ago. The regen energy numbers shown are often ludicrous. In extreme examples, the day and month regen numbers can be larger than the net energy numbers. No doubt related to that, I see the regen number going up when the car is parked. (That's a nice trick!) I take this to mean that something on the CAN bus changed but the VDS firmware hasn't been updated.
Things that are preserved over a VDS reset:
Charge time, mode and current limit. (Current limit is stored by location, presumably by the VMS.) Trip distance and net energy used.
So, presumably all of the values in this group can be queried from the CAN bus.
We have discovered charge mode and current limit. It is periodically sent from the VMS. Charge kW is also on the bus. I recently worked on charge time and mechanism (on-demand/delayed) and found the command/response codes for these, but haven't found any periodic transmission for this on the bus, yet. Are you 100% certain charge time is maintained across VMS power failure? We (Michael, or me?) have also discovered trip distance and odometer. I haven't found net energy used, yet, (not really gone looking) but it is presumably there somewhere. Trip is on id#100, B1=0x9c, and is in 1/10th of a mile - that is fully decoded. There is also an odometer message with this in id#402 - that how some not-yet-decoded bytes at the start, but they don't appear to be incremental are are more likely bit states for something. If energy use is there, it is on one of the undiscovered messages. Instantaneous amps is available on the dashboard message id#400 B1=0x02 (we use it to spoof the digital speedo).
Trip efficiency also persists. It may be calculated by the VDS or available on the CAN bus.
My most recent trip meter showed 46.7 miles, 11.32 kWh, and 243 Wh/mile. If you do the math, 11.32 kWh / 46.7 miles = 242.398 Wh/mi, which should round to 242. Since the trip distance has been found on the bus at 0.1 mi resolution, it seems like that's the best data the VDS has. If there are higher resolution numbers somewhere (VMS?), then maybe the real values are 46.65 miles and 11.324 kWh -> 242.74 Wh/mi which rounds up to 243. So, perhaps the efficiency number is calculated elsewhere and broadcast on the CAN bus. Alternatively, the VDS either may have a higher resolution energy number than what it's showing or perhaps it just does math badly.
I wouldn't rule out a bad rounding in the VDS. Rounding is done in so many places in the Tesla systems, and often different. For example - reset the trip and look at it on the car speedo vs VDS. We found similar things with the miles/kilometer conversion.
Tom
_______________________________________________ OvmsDev mailing list OvmsDev@lists.teslaclub.hk http://lists.teslaclub.hk/mailman/listinfo/ovmsdev
Are you 100% certain charge time is maintained across VMS power failure?
Yes. Your question about that is why I did the experiment. As soon as I have a way to capture CAN messages, I'll do it again and let you know what I see when the VDS boots up. Also, I was reading the notes on the CAN messages. The VDS doesn't control charging. In fact there's an annoying bug with that: - plug the car in, but don't slide the locking switch forward - wait until the coolant pump turns off - slide the locking switch forward Even though the VDS is on and the car seems totally alert, it doesn't recognize the pilot signal and will not charge. I've heard several reports of this making surprised Roadster owners very unhappy to find their car not charged. To work around the bug, do anything that starts the pump going again. There's another bug if you plug in then change the timer setting. The VDS will show the new time, but will charge on the old time. That one burned me: I set the charge time earlier to get a full charge for a long drive that needed to start early in the morning. It charged on the old schedule, wasn't done when I needed it, and made for a tense day of driving. To work around the bug: unplug: close then open the charge door, plug back in. I reported both bugs to Tesla over a year ago. No action. Tom on 4/23/12 5:34 PM, Mark Webb-Johnson wrote:
Tom,
Answers, inline, below.
Regards, Mark.
On 24 Apr, 2012, at 1:42 AM, Tom Saxton wrote:
When the VDS on a v1.5 Roadster gets reset, it loses the following information:
Charge History Energy History Trip Meter Drive Time Wh/mile history graph
I take this to mean that these things are calculated from info on the CAN bus and stored in the VDS.
Yes, that makes sense.
In v1.5 Roadsters, the energy history got screwed up by a firmware update more than a year ago. The regen energy numbers shown are often ludicrous. In extreme examples, the day and month regen numbers can be larger than the net energy numbers. No doubt related to that, I see the regen number going up when the car is parked. (That's a nice trick!) I take this to mean that something on the CAN bus changed but the VDS firmware hasn't been updated.
Things that are preserved over a VDS reset:
Charge time, mode and current limit. (Current limit is stored by location, presumably by the VMS.) Trip distance and net energy used.
So, presumably all of the values in this group can be queried from the CAN bus.
We have discovered charge mode and current limit. It is periodically sent from the VMS. Charge kW is also on the bus.
I recently worked on charge time and mechanism (on-demand/delayed) and found the command/response codes for these, but haven't found any periodic transmission for this on the bus, yet. Are you 100% certain charge time is maintained across VMS power failure?
We (Michael, or me?) have also discovered trip distance and odometer. I haven't found net energy used, yet, (not really gone looking) but it is presumably there somewhere. Trip is on id#100, B1=0x9c, and is in 1/10th of a mile - that is fully decoded. There is also an odometer message with this in id#402 - that how some not-yet-decoded bytes at the start, but they don't appear to be incremental are are more likely bit states for something. If energy use is there, it is on one of the undiscovered messages.
Instantaneous amps is available on the dashboard message id#400 B1=0x02 (we use it to spoof the digital speedo).
Trip efficiency also persists. It may be calculated by the VDS or available on the CAN bus.
My most recent trip meter showed 46.7 miles, 11.32 kWh, and 243 Wh/mile. If you do the math, 11.32 kWh / 46.7 miles = 242.398 Wh/mi, which should round to 242. Since the trip distance has been found on the bus at 0.1 mi resolution, it seems like that's the best data the VDS has. If there are higher resolution numbers somewhere (VMS?), then maybe the real values are 46.65 miles and 11.324 kWh -> 242.74 Wh/mi which rounds up to 243. So, perhaps the efficiency number is calculated elsewhere and broadcast on the CAN bus. Alternatively, the VDS either may have a higher resolution energy number than what it's showing or perhaps it just does math badly.
I wouldn't rule out a bad rounding in the VDS. Rounding is done in so many places in the Tesla systems, and often different. For example - reset the trip and look at it on the car speedo vs VDS. We found similar things with the miles/kilometer conversion.
Tom
_______________________________________________ OvmsDev mailing list OvmsDev@lists.teslaclub.hk http://lists.teslaclub.hk/mailman/listinfo/ovmsdev
_______________________________________________ OvmsDev mailing list OvmsDev@lists.teslaclub.hk http://lists.teslaclub.hk/mailman/listinfo/ovmsdev
Tom,
- plug the car in, but don't slide the locking switch forward - wait until the coolant pump turns off - slide the locking switch forward
Ok. That explains the weird behavior I was having before. I would change the charge mode/timing remotely via can bus action, and nothing would happen. The next morning I would go down to the car, press the door handle, and the car would start charging! I see on the can bus that whenever you press anything on the VDS, after a sleep, it sends a id #102 B1=0x0a to wake up something in the VMS. We changed OVMS to match, and everything seems ok now. In other words, when you change the charge mode/timing/current-limit in OVMS, it wakes up the car to avoid this ;-)
There's another bug if you plug in then change the timer setting. The VDS will show the new time, but will charge on the old time.
Been there, done that ;-) Regards, Mark. On 24 Apr, 2012, at 1:35 PM, Tom Saxton wrote:
Are you 100% certain charge time is maintained across VMS power failure?
Yes. Your question about that is why I did the experiment. As soon as I have a way to capture CAN messages, I'll do it again and let you know what I see when the VDS boots up.
Also, I was reading the notes on the CAN messages. The VDS doesn't control charging. In fact there's an annoying bug with that:
- plug the car in, but don't slide the locking switch forward - wait until the coolant pump turns off - slide the locking switch forward
Even though the VDS is on and the car seems totally alert, it doesn't recognize the pilot signal and will not charge. I've heard several reports of this making surprised Roadster owners very unhappy to find their car not charged. To work around the bug, do anything that starts the pump going again.
There's another bug if you plug in then change the timer setting. The VDS will show the new time, but will charge on the old time. That one burned me: I set the charge time earlier to get a full charge for a long drive that needed to start early in the morning. It charged on the old schedule, wasn't done when I needed it, and made for a tense day of driving. To work around the bug: unplug: close then open the charge door, plug back in.
I reported both bugs to Tesla over a year ago. No action.
Tom
on 4/23/12 5:34 PM, Mark Webb-Johnson wrote:
Tom,
Answers, inline, below.
Regards, Mark.
On 24 Apr, 2012, at 1:42 AM, Tom Saxton wrote:
When the VDS on a v1.5 Roadster gets reset, it loses the following information:
Charge History Energy History Trip Meter Drive Time Wh/mile history graph
I take this to mean that these things are calculated from info on the CAN bus and stored in the VDS.
Yes, that makes sense.
In v1.5 Roadsters, the energy history got screwed up by a firmware update more than a year ago. The regen energy numbers shown are often ludicrous. In extreme examples, the day and month regen numbers can be larger than the net energy numbers. No doubt related to that, I see the regen number going up when the car is parked. (That's a nice trick!) I take this to mean that something on the CAN bus changed but the VDS firmware hasn't been updated.
Things that are preserved over a VDS reset:
Charge time, mode and current limit. (Current limit is stored by location, presumably by the VMS.) Trip distance and net energy used.
So, presumably all of the values in this group can be queried from the CAN bus.
We have discovered charge mode and current limit. It is periodically sent from the VMS. Charge kW is also on the bus.
I recently worked on charge time and mechanism (on-demand/delayed) and found the command/response codes for these, but haven't found any periodic transmission for this on the bus, yet. Are you 100% certain charge time is maintained across VMS power failure?
We (Michael, or me?) have also discovered trip distance and odometer. I haven't found net energy used, yet, (not really gone looking) but it is presumably there somewhere. Trip is on id#100, B1=0x9c, and is in 1/10th of a mile - that is fully decoded. There is also an odometer message with this in id#402 - that how some not-yet-decoded bytes at the start, but they don't appear to be incremental are are more likely bit states for something. If energy use is there, it is on one of the undiscovered messages.
Instantaneous amps is available on the dashboard message id#400 B1=0x02 (we use it to spoof the digital speedo).
Trip efficiency also persists. It may be calculated by the VDS or available on the CAN bus.
My most recent trip meter showed 46.7 miles, 11.32 kWh, and 243 Wh/mile. If you do the math, 11.32 kWh / 46.7 miles = 242.398 Wh/mi, which should round to 242. Since the trip distance has been found on the bus at 0.1 mi resolution, it seems like that's the best data the VDS has. If there are higher resolution numbers somewhere (VMS?), then maybe the real values are 46.65 miles and 11.324 kWh -> 242.74 Wh/mi which rounds up to 243. So, perhaps the efficiency number is calculated elsewhere and broadcast on the CAN bus. Alternatively, the VDS either may have a higher resolution energy number than what it's showing or perhaps it just does math badly.
I wouldn't rule out a bad rounding in the VDS. Rounding is done in so many places in the Tesla systems, and often different. For example - reset the trip and look at it on the car speedo vs VDS. We found similar things with the miles/kilometer conversion.
Tom
_______________________________________________ OvmsDev mailing list OvmsDev@lists.teslaclub.hk http://lists.teslaclub.hk/mailman/listinfo/ovmsdev
_______________________________________________ OvmsDev mailing list OvmsDev@lists.teslaclub.hk http://lists.teslaclub.hk/mailman/listinfo/ovmsdev
_______________________________________________ OvmsDev mailing list OvmsDev@lists.teslaclub.hk http://lists.teslaclub.hk/mailman/listinfo/ovmsdev
My car went in for annual service yesterday. As part of that, all power is isolated, the pem is removed, and VDS and VMS are shutdown. When I got it back, with latest 4.6.5/40 firmware, I'm glad to say everything worked as expected. OVMS resumed and I got charging status ok - I even got a SMS alert when the Ranger stopped the test charge he was doing ;-) Only issue was no pem/motor/fan temperatures in the App. Selecting the temperature screen on the VDS instantly brought them back, as expected. When I got home, on the VDS, the car had lost its memory of charge time, mode and current limit. Maybe that is just a firmware upgrade issue, rather than generic VMS/VDS reset issue, but annoying as I was hoping to lose charge method/time but not current limit and mode (which is what we see on the bus). Regards, Mark. On 24 Apr, 2012, at 1:42 AM, Tom Saxton wrote:
When the VDS on a v1.5 Roadster gets reset, it loses the following information:
Charge History Energy History Trip Meter Drive Time Wh/mile history graph
I take this to mean that these things are calculated from info on the CAN bus and stored in the VDS.
In v1.5 Roadsters, the energy history got screwed up by a firmware update more than a year ago. The regen energy numbers shown are often ludicrous. In extreme examples, the day and month regen numbers can be larger than the net energy numbers. No doubt related to that, I see the regen number going up when the car is parked. (That's a nice trick!) I take this to mean that something on the CAN bus changed but the VDS firmware hasn't been updated.
Things that are preserved over a VDS reset:
Charge time, mode and current limit. (Current limit is stored by location, presumably by the VMS.) Trip distance and net energy used.
So, presumably all of the values in this group can be queried from the CAN bus.
Trip efficiency also persists. It may be calculated by the VDS or available on the CAN bus.
My most recent trip meter showed 46.7 miles, 11.32 kWh, and 243 Wh/mile. If you do the math, 11.32 kWh / 46.7 miles = 242.398 Wh/mi, which should round to 242. Since the trip distance has been found on the bus at 0.1 mi resolution, it seems like that's the best data the VDS has. If there are higher resolution numbers somewhere (VMS?), then maybe the real values are 46.65 miles and 11.324 kWh -> 242.74 Wh/mi which rounds up to 243. So, perhaps the efficiency number is calculated elsewhere and broadcast on the CAN bus. Alternatively, the VDS either may have a higher resolution energy number than what it's showing or perhaps it just does math badly.
Tom
_______________________________________________ OvmsDev mailing list OvmsDev@lists.teslaclub.hk http://lists.teslaclub.hk/mailman/listinfo/ovmsdev
participants (2)
-
Mark Webb-Johnson -
Tom Saxton