User avatar
garygid
Gold Member
Posts: 12469
Joined: Wed Apr 21, 2010 8:10 am
Delivery Date: 29 Mar 2011
Leaf Number: 000855
Location: Laguna Hills, Orange Co, CA

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

Sat Jul 20, 2013 9:19 pm

Great work everybody.
It is wonderful to see you all investigating and helping each other.
If only more of the world worked that way.

Merry Christmas to All. :D
See SOC/GID-Meter and CAN-Do Info
2010 Prius
2011 LEAF, 2014 Tesla S85
2018 & 2019 Tesla Model 3
PU: SDG&E
Solar PV: 33 x 225W -> 7 kW max AC
Craigslist: Xm5000Li Electric Motorcycle

User avatar
TickTock
Posts: 1701
Joined: Sat Jun 04, 2011 10:30 pm
Delivery Date: 31 May 2011
Leaf Number: 3626
Location: Queen Creek, Arizona
Contact: Website

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

Tue Jul 23, 2013 6:19 am

Moving this discussion from the other thread
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.

What makes us think these are shunt bits? I've tried both and neither looks right to me - looks random.

User avatar
garygid
Gold Member
Posts: 12469
Joined: Wed Apr 21, 2010 8:10 am
Delivery Date: 29 Mar 2011
Leaf Number: 000855
Location: Laguna Hills, Orange Co, CA

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

Tue Jul 23, 2013 7:33 am

There appear to be 96x2 data bits, and 96 are occasionally on,
with the other 96 always off. So, it is a guess, but one that
seems to fit with the Battery data, and we have no other
reasonable explanation at this point for these 96 bits of data.
See SOC/GID-Meter and CAN-Do Info
2010 Prius
2011 LEAF, 2014 Tesla S85
2018 & 2019 Tesla Model 3
PU: SDG&E
Solar PV: 33 x 225W -> 7 kW max AC
Craigslist: Xm5000Li Electric Motorcycle

User avatar
TickTock
Posts: 1701
Joined: Sat Jun 04, 2011 10:30 pm
Delivery Date: 31 May 2011
Leaf Number: 3626
Location: Queen Creek, Arizona
Contact: Website

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

Tue Jul 23, 2013 8:56 am

On page EVB-26 of the April, 2011 rev of the SM, it states that the highest cellpairs are discharged until they reach the target level. I see random cellpairs getting selected - sometimes even the low ones. Under no condition should the lowest cellpair get shunted. Also, others have shown captures with lots of CPs highlighted. I really doubt more than a half dozen or so are ever shunted since this would reduce the available pack voltage.
You do not have the required permissions to view the files attached to this post.

GregH
Posts: 860
Joined: Fri Jul 01, 2011 4:16 pm
Delivery Date: 13 Jun 2011
Leaf Number: 26967
Location: Irvine, CA

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

Tue Jul 23, 2013 9:04 am

TickTock wrote:On page EVB-26 of the April, 2011 rev of the SM, it states that the highest cellpairs are discharged until they reach the target level. I see random cellpairs getting selected - sometimes even the low ones. Under no condition should the lowest cellpair get shunted. Also, others have shown captures with lots of CPs highlighted. I really doubt more than a half dozen or so are ever shunted since this would reduce the available pack voltage.
If the BMS is not very intelligent and only tries top side balancing without considering the SOH at low SOCs, then it is entirely possible at a low SOC to see a low cell have balancing prescribed hours earlier when it may have been much higher.
'17 blue Volt Premier w/ACC
'12 SL black Leaf
'11 SL blue Leaf
RAV4-EV 2002-2005
Gen1 & Gen2 EV1 1997-2003
PV 2.4Kw, 10kWh lithium battery SCE TOU-DA

User avatar
TickTock
Posts: 1701
Joined: Sat Jun 04, 2011 10:30 pm
Delivery Date: 31 May 2011
Leaf Number: 3626
Location: Queen Creek, Arizona
Contact: Website

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

Tue Jul 23, 2013 9:25 am

GregH wrote:
TickTock wrote:On page EVB-26 of the April, 2011 rev of the SM, it states that the highest cellpairs are discharged until they reach the target level. I see random cellpairs getting selected - sometimes even the low ones. Under no condition should the lowest cellpair get shunted. Also, others have shown captures with lots of CPs highlighted. I really doubt more than a half dozen or so are ever shunted since this would reduce the available pack voltage.
If the BMS is not very intelligent and only tries top side balancing without considering the SOH at low SOCs, then it is entirely possible at a low SOC to see a low cell have balancing prescribed hours earlier when it may have been much higher.
That's another thing - it makes no sense to me for these to be active while driving. It's been a while since I had the conversation, but I am pretty sure the Nissan engineer told me they only do top-side balancing (which is consistent with the description in the SM). There is no reason to shunt the cells while driving - it should only be done while connected to the charger so the remaining cells can, then, be charged more.

GregH
Posts: 860
Joined: Fri Jul 01, 2011 4:16 pm
Delivery Date: 13 Jun 2011
Leaf Number: 26967
Location: Irvine, CA

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

Tue Jul 23, 2013 10:03 am

TickTock wrote:
GregH wrote:
TickTock wrote:On page EVB-26 of the April, 2011 rev of the SM, it states that the highest cellpairs are discharged until they reach the target level. I see random cellpairs getting selected - sometimes even the low ones. Under no condition should the lowest cellpair get shunted. Also, others have shown captures with lots of CPs highlighted. I really doubt more than a half dozen or so are ever shunted since this would reduce the available pack voltage.
If the BMS is not very intelligent and only tries top side balancing without considering the SOH at low SOCs, then it is entirely possible at a low SOC to see a low cell have balancing prescribed hours earlier when it may have been much higher.
That's another thing - it makes no sense to me for these to be active while driving. It's been a while since I had the conversation, but I am pretty sure the Nissan engineer told me they only do top-side balancing (which is consistent with the description in the SM). There is no reason to shunt the cells while driving - it should only be done while connected to the charger so the remaining cells can, then, be charged more.
When we did the first lithium PHEV Prius conversions I would do exactly this. The iron phosphate batteries really pop up at the end of charge so it was really easy to identify the fullest cells and bring on balancing... shut off the (1kW.. 5A, 200V) charger and balance for 10-20 minutes (@40mA) then turn off balancing, turn on the charger.. repeat.. This of course required being plugged in for hours after a full charge.. not optimal.

If you know which cells you'd like to remove a few mAh from then it really doesn't matter much what the voltage is at the time the balancing is done. You just need to keep the shunt on for X minutes or hours.

If my Leaf is balancing, it always seems to be off when I turn on the car then I can see the shunts turn on over a few seconds from left to right (lowest cell to highest cell). I can see the same pattern come back on over a few seconds every time I turn the car back on. This indicates to me the car does NOT balance when off.. but it certainly does when on or when charging. As a matter of fact it could be balancing in the middle of an L2 charge based on an assessment made the day before at the end of a previous L2 charge. Note this would also mean the car does NO BALANCING more than 15 minutes after a full charge.. ie when the 3 blue LEDs turn off and the car goes to sleep, the balancing turns off.

I'm not sure why they don't balance with the car off, perhaps to make sure no shunts get stuck on.. more likely the LBC shuts down when the VCM goes to sleep seconds after key-off and doesn't have the ability to leave shunts stuck on while sleeping. For better or worse, I designed my battery management systems so that I could leave balancing on when the processor powered down. I had LEDs on each shunt visible from outside the pack so you could see at a glance which shunts were on.
'17 blue Volt Premier w/ACC
'12 SL black Leaf
'11 SL blue Leaf
RAV4-EV 2002-2005
Gen1 & Gen2 EV1 1997-2003
PV 2.4Kw, 10kWh lithium battery SCE TOU-DA

User avatar
DaveEV
Posts: 6245
Joined: Fri Apr 23, 2010 3:51 pm
Location: San Diego

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

Tue Jul 23, 2013 10:15 am

GregH wrote:If you know which cells you'd like to remove a few mAh from then it really doesn't matter much what the voltage is at the time the balancing is done. You just need to keep the shunt on for X minutes or hours.
Good point...
GregH wrote:Note this would also mean the car does NO BALANCING more than 15 minutes after a full charge.
It would be nice to confirm this since most have assumed (including Phil/Ingineer) that balancing can happen at any time - hence the recommendation to charge to 100% and let the car sit for up to 4 hours to help encourage pack balancing.

Should be easy enough to test - can we read CP voltages up to 15 min after charging stops? Record them then, wait a few hours and turn the car back on and check again...

User avatar
TickTock
Posts: 1701
Joined: Sat Jun 04, 2011 10:30 pm
Delivery Date: 31 May 2011
Leaf Number: 3626
Location: Queen Creek, Arizona
Contact: Website

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

Tue Jul 23, 2013 10:58 am

Also, keep track of which cell(s) were minimum at the end of the charge to see if those ever (later) get shunted.

GregH
Posts: 860
Joined: Fri Jul 01, 2011 4:16 pm
Delivery Date: 13 Jun 2011
Leaf Number: 26967
Location: Irvine, CA

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

Tue Jul 23, 2013 11:14 am

drees wrote:Should be easy enough to test - can we read CP voltages up to 15 min after charging stops? Record them then, wait a few hours and turn the car back on and check again...
Unfortunately no... When the LBC is asleep we can't ping it for cell data. :(
My only evidence for this is that when I first turn the car on I see no shunts on, then over the course of 5 seconds or so they turn on.. I'll try to capture better video of this in action. It's pretty clear when you see it.
TickTock wrote:Also, keep track of which cell(s) were minimum at the end of the charge to see if those ever (later) get shunted.
Yes. Wasn't there some post from an Android user who had a very low cell then showed a screenshot of all the shunts on except that one? I guess that would indicate it does take into account voltage at low SOC.. although it could also be that cell was really low at a full charge as well.
'17 blue Volt Premier w/ACC
'12 SL black Leaf
'11 SL blue Leaf
RAV4-EV 2002-2005
Gen1 & Gen2 EV1 1997-2003
PV 2.4Kw, 10kWh lithium battery SCE TOU-DA

Return to “LEAF CANBus”