Page 9 of 17

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Posted: Tue Jul 30, 2013 9:04 am
by GregH
Turbo3 wrote:
LEAFfan wrote:
Turbo3 wrote:Correct value is 4821
Actually, Greg said it was 8421 and we had the incorrect one.
I guess GregH and Gary Giddings need to talk ;-)

When everyone agrees on a single value then I can hardcode it. But it sounds like now is not that time.
Here are the shunts coming on after powering up the car about an hour (and 10 miles) after a "full" charge. This is with the 8421 bit ordering. Note as the shunts turn on the corresponding voltages sag a little.


Re: Active EV-CAN sampling: cell voltages, pack temperatures

Posted: Tue Jul 30, 2013 10:32 am
by TickTock
Thanks for that - having the video really helps especially if you pause it and step through it backwards (so you can see which CP to focus on when you step forward in time). It does appear to do as you say - I am no longer skeptical. :-)

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Posted: Wed Jul 31, 2013 10:17 am
by TickTock
Turbo3 wrote:Need help with ambient temperature decoding.

To get ambient temperature send:

797 8 03 22 11 5D

Get back:

79A 8 07 62 11 5D 7F 72 7F 00 (which is a reading of 73.4 degrees)

The task is to determine which of the last four data bytes are used for temp and their conversion factor.

Jim
Copying from the from the [passive] canbus decoding thread since this is an active (requested) message. I think this is an important value to decode if we want to temperature compensate the DTE computations since pack temp is usually quite different than ambient.

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Posted: Wed Jul 31, 2013 2:07 pm
by TickTock
I'd like to do some more analysis on the parameter Greg discovered in the Group1 data called "perhaps health" by Greg and Hlth in the leafApp. While it does seem to be a good indicator, I don't think it is really what we are looking for based on my observations.
Below is some information I've previously reported along with additional data from 5b3:D2 which I believe is the true SOH as defined by Nissan:
  • Just before update:
    9 capacity bars
    62.4% Hlth
    68% SOH
    45.59 Ah

    Just after the update (measured in the parking lot):
    12 capacity bars
    100.0% Hlth
    100% SOH
    65.60 Ah

    Upon arriving home:
    12 capacity bars
    50.9% Hlth
    100% SOH
    65.60 Ah

    The next day:
    12 cap bars
    50.9% Hlth
    72% SOH
    46.92 Ah

    15 days later:
    12 cap bars
    51.7% Hlth
    73% SOH
    48.3 Ah

    Today:
    12 capacity bars
    51.7% Hlth
    72% SOH
    47.42Ah

    LEAFfan's 2013:
    12 capacity bars
    90.6% Hlth
    90% SOH
    59.37Ah
Based on this, it appears:
  • 1) SOH is pretty nearly Ah divided by 65.6. Probably exactly that post-sw-update (good news if you are not able to monitor canbus)
    2) Since Hlth was able to update mid drive home but SOH (5b3:d2) and Ah didn't get updated until after a full charge, it appears hlth is not a measure capacity (maybe resistance or something else that also degrades with age).

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Posted: Wed Jul 31, 2013 4:11 pm
by JPWhite
So it looks like you gained back a few Ah after a few weeks of being on the new s/w. Hope I get that much this Friday when I take mine in for the update. I'll be able to charge to 80% routinely again :-)
TickTock wrote:I'd like to do some more analysis on the parameter Greg discovered in the Group1 data called "perhaps health" by Greg and Hlth in the leafApp. While it does seem to be a good indicator, I don't think it is really what we are looking for based on my observations.
Below is some information I've previously reported along with additional data from 5b3:D2 which I believe is the true SOH as defined by Nissan:
  • Just before update:
    9 capacity bars
    62.4% Hlth
    68% SOH
    45.59 Ah

    Just after the update (measured in the parking lot):
    12 capacity bars
    100.0% Hlth
    100% SOH
    65.60 Ah

    Upon arriving home:
    12 capacity bars
    50.9% Hlth
    100% SOH
    65.60 Ah

    The next day:
    12 cap bars
    50.9% Hlth
    72% SOH
    46.92 Ah

    15 days later:
    12 cap bars
    51.7% Hlth
    73% SOH
    48.3 Ah

    Today:
    12 capacity bars
    51.7% Hlth
    72% SOH
    47.42Ah

    LEAFfan's 2013:
    12 capacity bars
    90.6% Hlth
    90% SOH
    59.37Ah
Based on this, it appears:
  • 1) SOH is pretty nearly Ah divided by 65.6. Probably exactly that post-sw-update (good news if you are not able to monitor canbus)
    2) Since Hlth was able to update mid drive home but SOH (5b3:d2) and Ah didn't get updated until after a full charge, it appears hlth is not a measure capacity (maybe resistance or something else that also degrades with age).

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Posted: Wed Jul 31, 2013 4:50 pm
by TickTock
Yes... as reported, anyway. It is still too early to tell, though, if this is real and results in a 4% increase in range. After another month of data collection, I hope to have something more conclusive.
JPWhite wrote:So it looks like you gained back a few Ah after a few weeks of being on the new s/w. Hope I get that much this Friday when I take mine in for the update. I'll be able to charge to 80% routinely again :-)

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Posted: Wed Jul 31, 2013 8:45 pm
by GregH
TickTock wrote:
Turbo3 wrote:Need help with ambient temperature decoding.

To get ambient temperature send:

797 8 03 22 11 5D

Get back:

79A 8 07 62 11 5D 7F 72 7F 00 (which is a reading of 73.4 degrees)

The task is to determine which of the last four data bytes are used for temp and their conversion factor.

Jim
Copying from the from the [passive] canbus decoding thread since this is an active (requested) message. I think this is an important value to decode if we want to temperature compensate the DTE computations since pack temp is usually quite different than ambient.
Since the car uses Celsius you're looking for 23, not 73.4... Maybe the 0x5d?... -70? Hmmm.. Need more data points. Oops. No, that's the request Id.. Sorry

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Posted: Thu Aug 01, 2013 12:50 am
by Turbo3
Ambient temp seems to be the 6th byte. In this example that would be 0x72= 114 subtract the magic number 41 gives 73. The other two bytes seem to be interior temps.

Two other samples are 0x6f = 111 -41 = 70 which was the dash reading and
0x6D = 109 -41 = 68 again the dash reading.

Ok, one more 0x6c dash reads 67 degrees.

Just left the garage door open to cool the car down and modified the battery app to display the four bytes.

This was much easier than I thought it would be.

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Posted: Thu Aug 01, 2013 4:36 am
by TickTock
Nice work Jim! I never even got out of the gate! :-)

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Posted: Thu Aug 01, 2013 5:12 am
by JPWhite
Turbo3 wrote:Ambient temp seems to be the 6th byte. In this example that would be 0x72= 114 subtract the magic number 41 gives 73. The other two bytes seem to be interior temps.

Two other samples are 0x6f = 111 -41 = 70 which was the dash reading and
0x6D = 109 -41 = 68 again the dash reading.

Ok, one more 0x6c dash reads 67 degrees.

Just left the garage door open to cool the car down and modified the battery app to display the four bytes.

This was much easier than I thought it would be.
Awesome!! Looking forward to this being included in a future build :-)