update on cell drift of my pack - DIY Electric Car Forums
Go Back  

DIY Electric Car Forums > EV Conversions and Builds > Batteries and Charging

Register Blogs FAQ Members List Social Groups Calendar Search Today's Posts Mark Forums Read

Reply
 
Thread Tools Display Modes
  #1  
Old 09-17-2011, 10:44 AM
dtbaker's Avatar
dtbaker dtbaker is offline
Senior Member
 
Join Date: Jan 2008
Location: santa fe, nm
Posts: 3,987
dtbaker is on a distinguished road
Default update on cell drift of my pack

I just thought I would start a thread on MY pack to record results, and see if it is what other people are seeing.

First.... my pack is 38 x 100ah Thundersky cells. I am using an Elcon pfc-1500 charger with simple CA->CV curve set to finish at 139v which gives a target average of 3.658vpc at 'finish'. I have no BMS, and leave my dc-dc 'on' all the time. I do use the vehicle almost every day, and charge at night.

Longest period of non-use was 1 week, and I left dc-dc 'on' to see what drain would be.... Kill-a-watt meter measured 1.9kWhr it took to charge back up to 'full'. Confirming that if you are not going to drive for a week, you really ought to flip the main breaker to prevent ANY drain!

I did an initial top balance in parallel to 3.90v with a bench power supply, then installed and wired in series. First series charge I monitored very closely and it didn't take more than a couple minutes since cells were charged from initial parallel balance, and was glad I did because I did have a couple cells head past 3.80 while others were below the CA->CV voltage of 3.65vpc. I then 'knocked down' the high ones with a pair of 50W resistors and repeated charge cycle until I had the whole pack +/- .02v just after end of charge.

The manual final tweaking of the balance in series is amazingly sensitive to small changes, and the voltage rise from flat to starting up the knee happens fast, and doesn't last long. I found that tweaking individual cells with no more than about 30 seconds of either draining with resistors or adding with power supply was the best way to avoid overshooting the target.

The cell voltage measurement I found most practical to do a few minutes AFTER the charger 'finished', and the pack voltage settled from the finish voltage of 139 to 136.0... because it hangs at close to 136.0 long enough for me to buzz around and record voltages from each cell.

The initial balance was +/- .02 on all cell but 4 that were -.03 ... I had to button the car up and do a show and tell and local community college, so I left it at that figuring that the 'high cells' would finish at no more than 3.68vpc, which is still fairly conservative for Thunderskies.

My first balance check was after about 1000 miles.... there was virtually no change in relative balance between cells. I decided not to change balance at all to see if there would be any drift with more use.

I just did second balance check yesterday at just over 3000 miles on the pack. There was virtually no change in relative balance between cells. I decided to spend a little time to get the 4 cells that were -.03 up a little... and after a couple charge/checks, I was able to get all cells +/- .02 at end of charge.

so.... at this point I just wanted to note that I experienced NO significant relative cell drift in the first 3000 miles.
__________________
Dan
http://www.envirokarma.org/ev - '97 Suzuki Swift
http://www.envirokarma.org/ev2_mx5e - '94 Mazda MX5 Miata
Reply With Quote
Sponsored Links
Advertisement
 
  #2  
Old 09-17-2011, 01:40 PM
Salty9 Salty9 is offline
Senior Member
 
Join Date: Jul 2009
Location: 42.596N 122.688W
Posts: 747
Salty9 is on a distinguished road
Default Re: update on cell drift of my pack

Reads like you are the BMS for your pack If you had a BMS how much more time would you be spending monitoring the BMS?
Reply With Quote
  #3  
Old 09-17-2011, 08:39 PM
icec0o1 icec0o1 is offline
Senior Member
 
Join Date: Sep 2009
Posts: 197
icec0o1 is on a distinguished road
Default Re: update on cell drift of my pack

Quote:
Originally Posted by Salty9 View Post
Reads like you are the BMS for your pack If you had a BMS how much more time would you be spending monitoring the BMS?
He's doing it so we don't have to. Why would you want to reinvent the wheel every time?
Reply With Quote
  #4  
Old 09-17-2011, 09:08 PM
Salty9 Salty9 is offline
Senior Member
 
Join Date: Jul 2009
Location: 42.596N 122.688W
Posts: 747
Salty9 is on a distinguished road
Default Re: update on cell drift of my pack

If it were my pack I would be monitoring it just like dtbaker is. Not because I mistrust his results but because I don't trust the uniformity of Chinese manufacture. Given the state of BMS today I trust them even less.
Reply With Quote
  #5  
Old 09-18-2011, 04:04 PM
GizmoEV's Avatar
GizmoEV GizmoEV is offline
Senior Member
 
Join Date: Nov 2009
Location: Kelso, WA
Posts: 891
GizmoEV is on a distinguished road
Default Re: update on cell drift of my pack

It took nearly 100 full cycle equivalents for my cells to be 0.1V different at the extremes. This could have been due to several different factors and likely wasn't much energy difference. I was charging to 3.485vpc and had BlackSheep BMS boards attached which draw only about 1mA. I rebalanced my pack to 4.00vpc using the BMS boards and then removed them. I take a reading each month to see what things are doing and then record and plot the results. I do have a half-pack voltage monitor on my pack but that is it besides the CycleAnalyst.
Reply With Quote
  #6  
Old 09-20-2011, 08:46 AM
PTCruisin PTCruisin is offline
Senior Member
 
Join Date: Nov 2009
Location: Southern Illinois
Posts: 132
PTCruisin is on a distinguished road
Default Re: update on cell drift of my pack

Dan,

As you know, I have a similar setup here and have about the same number of miles on the car. I did an initial top balance and did a "touch up" after about 2,000 miles. There are small differences (0.2 to 0.3) in voltage at the end of a charge cycle, but the batteries settle down nicely in a few hours. I only use 50 to 60% DOD in my commute and never get near the bottom, so the irregularities in cell capacity are not an issue for me. I think that is the key to running without a BMS.

I am considering adding one or two more cells to the string to lower the Vpc to either 3.55V (39s) or 3.47V (40s) to alleviate the differences in terminal voltages and add a mile or two of range.

Ralph
__________________
PTCruisin
http://www.evalbum.com/2946
Reply With Quote
  #7  
Old 09-20-2011, 10:35 AM
dtbaker's Avatar
dtbaker dtbaker is offline
Senior Member
 
Join Date: Jan 2008
Location: santa fe, nm
Posts: 3,987
dtbaker is on a distinguished road
Default Re: update on cell drift of my pack

Quote:
Originally Posted by Salty9 View Post
Reads like you are the BMS for your pack If you had a BMS how much more time would you be spending monitoring the BMS?
yes, I am the BMS.... but only checked twice in 3000 miles of use, so that seems a lot less expensive than the electronic version.

I am posting my results as a matter of general interest along with my methods for those people evaluating their own needs. I know that when I was designing and installing my Li system, I found very little actual use reports from EV users to support arguments one way or the other on many points, so.... I thought I would post my situation and experiences so far.

My conclusion at this point, supported by MY data on MY pack is that a careful initial top-balance (within +/- .02vpc at end-of-charge) and charging with a simple CA-CV curve set to a fairly conservative 3.65vpc average exhibits virtually no relative cell drift after 3000 miles of use.

My goal is simply to add a data point for people like myself to establish 'reasonable' frequency to check the balance on cells, and reasonable methods to manually re-balance if required.
__________________
Dan
http://www.envirokarma.org/ev - '97 Suzuki Swift
http://www.envirokarma.org/ev2_mx5e - '94 Mazda MX5 Miata
Reply With Quote
  #8  
Old 09-20-2011, 10:48 AM
dtbaker's Avatar
dtbaker dtbaker is offline
Senior Member
 
Join Date: Jan 2008
Location: santa fe, nm
Posts: 3,987
dtbaker is on a distinguished road
Default Re: update on cell drift of my pack

Quote:
Originally Posted by PTCruisin View Post
I did an initial top balance and did a "touch up" after about 2,000 miles. There are small differences (0.2 to 0.3) in voltage at the end of a charge cycle,
I hope you mean .02-.03 vpc differences?

Quote:
Originally Posted by PTCruisin View Post
but the batteries settle down nicely in a few hours. I only use 50 to 60% DOD in my commute and never get near the bottom, so the irregularities in cell capacity are not an issue for me. I think that is the key to running without a BMS.
agreed... but I would like to say that I think .02-.03vpc irregularity at end-of-charge is about all I want to tolerate to avoid possible overcharge issues on any one cell at end of charge. I have been surprised how a tiny difference in SOC can make a rapid and significant difference in final voltage. applying a 6amp charge or drain on a single cell for 30 seconds seems to change the final charge voltage close to .05v if the rest of the pack is 'balanced'. That gives me an idea just how steep that curve is, and how important that end-of-charge balance is.


Quote:
Originally Posted by PTCruisin View Post
I am considering adding one or two more cells to the string to lower the Vpc to either 3.55V (39s) or 3.47V (40s) to alleviate the differences in terminal voltages and add a mile or two of range.
Ralph
I would be concerned that if you target a final voltage closer to the flat, you may start giving up a significant amount/accuracy of capacity in the charge cycle because the 'final' voltage may be hit when a lot of the cells are not that close to full. I am figuring that with Thunderskies and target of 3.65 +/- .03vpc is pretty safe. They seem to only be at that final voltage of a couple minutes, and settle down rapidly after the charger turns off.
__________________
Dan
http://www.envirokarma.org/ev - '97 Suzuki Swift
http://www.envirokarma.org/ev2_mx5e - '94 Mazda MX5 Miata
Reply With Quote
  #9  
Old 09-20-2011, 02:36 PM
T1 Terry T1 Terry is offline
Member
 
Join Date: Jan 2011
Location: South Coast NSW Australia
Posts: 85
T1 Terry is on a distinguished road
Default Re: update on cell drift of my pack

I have conducted a number of experiments with my 360ah 12v house power pack. 16 X Winston LYP90ah. The first pack I carefully balanced at 4v per cell till they accepted no further charge using solar and an adjustable PWM float voltage regulator. I took it up in 1v steps from 3.4v per cell till the amps in dropped below 5amps and moved to the next stage, just made it easier and less likely to have a run away cell. this pack has remained balanced with every day cycling for 60 cycles so far and shows no sign of changing. I charge to 3.45v per cell, gives a good safety margin just in case and I have built a simple charge cut off/reset if any cell goes over 3.85v, just in case. i use a simple Jusi cell logger and the alarm triggers a 30 min timer circuit that turns off the solar positive input before the solar regulator allowing the house load to discharge any over voltage from a rogue cell. Similar to DT Baker leaving his DC to DC on.

The next pack of 16 cells I charged by a method but no balancing. The difference with these cells was they are all older Thundersky TS cells, been it the packing case for 2 1/2 yrs, production date 03/07/2009. I didn't do the 4.0v conditioning first but rather let them charge to 4.45v per cell and put them to work. Initially there was over a 300mv difference and the 3.85v cut out operated a few times but over a few mini cycles during the day the cells started to creep closer together. Within a few days the 3.85v cut didn't operate at all and after 2 weeks the pack as within 17mv, sometimes within 4mv. I decided to do the 4v conditioning, the pck stay together all the way to 4v per cell, I returned it to 3.45v per cell float and let it run for a week, still in balance.
these packs haven't cycled below 70% DoD much yet but on the few times they have (rain for a few days) they have remained in balance.

Just a side note, when under a 1C or higher load they do go out of balance, up to 60mv and it's never the lowest cell before the test starts that drops low under load so the relevance of off load balancing becomes a bit questionable but as hat is the only real method we have it will have to do.

T1 Terry
Reply With Quote
  #10  
Old 09-20-2011, 03:38 PM
PTCruisin PTCruisin is offline
Senior Member
 
Join Date: Nov 2009
Location: Southern Illinois
Posts: 132
PTCruisin is on a distinguished road
Default Re: update on cell drift of my pack

Quote:
Originally Posted by dtbaker View Post
I hope you mean .02-.03 vpc differences?
I did mean 0.02-0.03V

Quote:
I would be concerned that if you target a final voltage closer to the flat, you may start giving up a significant amount/accuracy of capacity in the charge cycle because the 'final' voltage may be hit when a lot of the cells are not that close to full. I am figuring that with Thunderskies and target of 3.65 +/- .03vpc is pretty safe. They seem to only be at that final voltage of a couple minutes, and settle down rapidly after the charger turns off.
At one time I monitored all the cells in my pack with Cell logs and observed that the "knee" of the curve is near 3.5V. The charger is already reducing current at that point in the profile, so by the time the cells get close to 3.6V, there is very little current (charge) going into the pack anyway.

Before I did the "touch up" I would see the lower capacity cells get up to 3.8 - 3.9V before the charger would shut down. I don't think this is all that harmful given that the current is around an amp or less, but I rebalanced them anyway.
__________________
PTCruisin
http://www.evalbum.com/2946
Reply With Quote
Reply

Share or Bookmark this

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

 
Support DIY Electric Car
Sponsors

All times are GMT -6. The time now is 05:27 AM.


Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2014, vBulletin Solutions, Inc.

Ad Management by RedTyger