Initial Lithium pack balance?

11155 Views 78 Replies 12 Participants Last post by tomofreno, Mar 30, 2011

EVfun

Discussion Starter · Mar 1, 2011

I'm about to parallel the buggy pack up to balance it. The pack is 32 Thunder Sky cells. I'm planning a top balance as it will be easier for me to limit the discharge and charging should be a little easier to adjust when they all come up together.

I've been pouring over the threads for recommendations about the setup and didn't find as much as I'd hoped for. Here is what I have in mind:

Code:

+ ------------------------    
    |  |  |  |  |  |  |  | 
    _  _  _  _  _  _  _  _    
    -  -  -  -  -  -  -  -    
    |  |  |  |  |  |  |  |    
    ------------------------ -

Well, mine will be 32 cells long but the idea is to put a ring terminal with 2 wires in it at each positive and each negative terminal and bring the positive and negative out on opposite ends so wiring resistance and any voltage drop from the charger should be equal (as much as possible) for every cell. Right now I have the cells in series so I will charge in series while watching to get them as close as possible before paralleling them. My plan is to balance them at 3.7 volts per cell as that is a little higher than my target charging voltage of about 3.6 volts per cell.

What size wire is recommended? I know the wire should be large enough to handle the charger current, but that only requires 16 gauge wire. Is larger needed to allow the cells to self balance? Each wire (from + to + or from - to -) will be about 6 inches long and the wires coming off each end for the charger will be about 3 feet long.

Will the cells move into balance slowly if they are not being charged or discharged? I'm in that big flat middle range with all 32 cells reading 3.31 volts each.

Is there anything I'm missing here?

Thanks for your input!

See less See more

1 - 20 of 79 Posts

nimblemotors

· #2 · Mar 1, 2011

I would suggest to actually test the capacity of each cell before building a pack with them. Given the ability to do that, you can top each cell, not sure why you want to parallel all of them.

EVfun said:
I'm about to parallel the buggy pack up to balance it. The pack is 32 Thunder Sky cells. I'm planning a top balance as it will be easier for me to limit the discharge and charging should be a little easier to adjust when they all come up together.

I've been pouring over the threads for recommendations about the setup and didn't find as much as I'd hoped for. Here is what I have in mind:

Code:

+ ------------------------ | | | | | | | | _ _ _ _ _ _ _ _ - - - - - - - - | | | | | | | | ------------------------ -

Well, mine will be 32 cells long but the idea is to put a ring terminal with 2 wires in it at each positive and each negative terminal and bring the positive and negative out on opposite ends so wiring resistance and any voltage drop from the charger should be equal (as much as possible) for every cell. Right now I have the cells in series so I will charge in series while watching to get them as close as possible before paralleling them. My plan is to balance them at 3.7 volts per cell as that is a little higher than my target charging voltage of about 3.6 volts per cell.

What size wire is recommended? I know the wire should be large enough to handle the charger current, but that only requires 16 gauge wire. Is larger needed to allow the cells to self balance? Each wire (from + to + or from - to -) will be about 6 inches long and the wires coming off each end for the charger will be about 3 feet long.

Will the cells move into balance slowly if they are not being charged or discharged? I'm in that big flat middle range with all 32 cells reading 3.31 volts each.

Is there anything I'm missing here?

Thanks for your input!

EVfun

Discussion Starter · #3 · Mar 1, 2011

I actually don't have the equipment to capacity test cells. The usual goal is to simply balance them, either at the top or bottom depending on how charge termination and discharge cut-off are being handled.

This is the most information initial balance or equalizing I found in the forum.

nimblemotors

· #4 · Mar 1, 2011

that is a good point. But look at the published cycle life, 80% 2000 cycles, 70% 3000 cycles, so you really don't want to take ANY cell down to full discharge, not even the weakest ones, since it will in effect increase the cost of your batteries by 50% (if we believe their cycle life claims, which apparantly nobody has confirmed!!

)

So the best is to count the amp/hours and consider it empty once 70% rated capacity has been drained, do not depend on voltage.
And check each cell to see the smallest capacity cell, and use it as your 70% value. So if you have 100ah cells, and one is 90ah, remove only 63ah from a full charged pack. yep it will reduce your range.

gottdi said:
I'd do the bottom balance for sure. The problem lies when you are out on a drive and you are getting to the point where the pack will begin to drop but is still showing good voltage you may in fact have a cell that has gone below the safe point and you now have a full pack with plenty of poop left driving it's current through a single or maybe a couple cells which will drive it to the ground and kill it. The problem is you will not know it's happening until it's too late as your voltage meter will still show you should have plenty of power left. Now bottom balanced cells will have a ragged top but if you charge your batteries and have a decent cut off point you may have a cell that during the charge is in the 4.1 volt range but maybe not quite that high. Since it is not a static voltage reading of 4.1 volts your still going to do just fine. If a cell actually went into the danger zone with the charger your cell is only being charged by the charger and you don't have a full pack of cells trying to drive current through that cell. The cell will more than likely be just fine. You will kill a cell faster if you top balance and have a cell or two reach that danger zone too fast and you'd not be aware. Things happen real fast. I did a video about bloated cells from over discharge. You can only use the capacity of the lowest common denominator so whether you top or bottom balance your usable capacity will always be limited to those cells. The higher capacity cells will just be along for the ride. Since you WILL from time to time take your car to the limits of discharge your best bet is to have the whole pack reach that deep discharge at the same time so you don't fry a cell or two. I guarantee you will at some point take your car to the limits. It is far safer for cells and your vehicle to have your cells bottom balanced. Just because its easy to top balance does not mean it is a good thing to do. I put my pack in a middle of the road balance and lost two cells at the bottom while out on a drive. One cell bloated up like a fat pig. No venting but is sure got hot and bloated up real bad. Got a video of dissecting a cell and finding out a couple things. Don't top balance your cells. When things happen at the bottom they happen fast. It was the last couple miles.

Now that the pack is bottom balanced sort of I have not had my pack get hot nor any single cell get hot. I am going to balance them again better at about 2.6 volts static then charge them up. I will have my end voltage cut back set for 2.8 and my cutoff at 2.5 volts. My charge will be at 3.6 and terminate at 2 amps. NO BMS.

Pete

Video is now on my web page if you want to see.

EVfun

Discussion Starter · #5 · Mar 1, 2011

gottdi said:
No, your not correct. You take your cells to the bottom. Balance them and then only take them to the low portion before the knee.

I was really hoping this wouldn't devolve into another top balance vs. bottom balance debate. Top balancing is more suited to my application. I have even tighter control over discharge cut-off than charge termination due to a little feature called "low lift lockout." (and yes I will go there, gas cars do)

How 'bout the rest of the process, ignoring the chosen voltage for the moment? Will the cells self equalize if paralleled while resting at 3.31 volts each? My understanding has been that voltage is a good indicator of SOC at all levels, except that you would need to accurately measure to within a few ten-thousandths of a volt when the cell voltage was in the flat middle range. What wire gauge is recommended?

nimblemotors

· #6 · Mar 1, 2011

cells are paralleled to increase capacity, it forces them all to the same voltage. I'm not convinced this is a good thing, unless they are matched cells, but could be wrong. I really think you need to confirm how matched they are before putting them in a pack. It isn't difficult to manually do it, I'd start with bottom balance, just drain a serial pack very slowly and use a DVM to check each cell as it drains until one starts to reach empty, then quit the serial discharge, and do each cell. This can take a while, it took me three days to balance all 38 cells in a prius pack.
Once they are all discharged, then do the same charging, in serial until you are near full (in terms of amp/hours) then slow it down and check each cell as they reach full. Now you can see which cells reach full first, then charge each cell individually to see how much they differ.

EVfun said:
I was really hoping this wouldn't devolve into another top balance vs. bottom balance debate. Top balancing is more suited to my application. I have even tighter control over discharge cut-off than charge termination due to a little feature called "low lift lockout." (and yes I will go there, gas cars do)

How 'bout the rest of the process, ignoring the chosen voltage for the moment? Will the cells self equalize if paralleled while resting at 3.31 volts each? My understanding has been that voltage is a good indicator of SOC at all levels, except that you would need to accurately measure to within a few ten-thousandths of a volt when the cell voltage was in the flat middle range. What wire gauge is recommended?

See less See more

DIYguy

· #7 · Mar 1, 2011

EVfun said:
How 'bout the rest of the process, ignoring the chosen voltage for the moment? Will the cells self equalize if paralleled while resting at 3.31 volts each?

Yes...............

EVfun

Discussion Starter · #8 · Mar 1, 2011

gottdi said:
If you top then you will need to know what cells drop out first and set your cut off above that. Bet you find it's harder to control than you think. I won't debate but will state facts.

I think you misunderstand how I intend to terminate discharge. I will be counting amp hours out and stop after discharging my smallest cell to about 20% SOC. The amp hour counter can drive the battery level gauge and its low lift lockout function can short the pot box. I figure if gas cars can abruptly run out of fuel sometime around when the gauge says empty an EV can shut down the pack when the amp hour counter says exactly empty. Anyway, if safety was a primary concern in the buggy I might start by installing seat belts.

Roy Von Rogers

· #9 · Mar 1, 2011 (Edited)

In theory this is what you need to do to set up a pack for the first time.

If the pack is in parallel put a load on the pack til the pack is empty with no load, depending on which cells you have, 2.9v to 3v is a generally good figure.

Now connect the pack in series and charge it CC/CV to 3.5v x cells, if you have 32 cells thats 112v.

The problem is that finding something for a discharge device at 3 volts is difficult. Another option is to assemble them 4 at a time in series, since many times they come already packed that way. Connect a 12v load to the 4 pack and discharge to 12v no load. Do that to all, and then connect all in series and let them sit for a day or so. Check the voltage after previous mentioned time span, and if the pack is somewhere at 96v, consider the pack empty and charge pack CC/CV to112v, the pack is now full.

Get yourself a good amp hour dash counter and dont drive your pack past 75 to 80% of pack capacity. Monitor your pack via either a 3 wire half pack unbalance indicator, or if you want to check all, get an inexpensive Cellog device available at hobby shop dealers, be aware you need to mod the cellog device so it powers up on all wires.

Just make sure that any monitoring device is properly wired and fused. DONT voltage balance these cells to the top via a shunt balancer, its a waste of money and does exactly the OPPOSITE of what you want to do, and could be dangerous to boot.

If you have a 100ah pack and you fill it with 100 ah, the battery is full, any cell thats above the lowest single cell capacity you need to ignore, for its NOT USABLE energy.

Roy

See less See more

DIYguy

· #10 · Mar 1, 2011

Paralleling the pack will equalize it at what ever the avg SOC is. Light gauge wire is fine since current will be pretty low.

If you want to top balance, you need three things. A DVM that measures to 3 decimal places, a method to charge and a method to discharge. Ideally, for top balance, you want to be up around the 3.5 volt range or so, to get a decent balance. Charge the pack till you get close. Charge/discharge individual cells until you get them the same (resting), to three decimal places.

I'm sure there are lots of variations to this...but that's basically it.

EVfun

Discussion Starter · #11 · Mar 1, 2011 (Edited)

DIYguy said:
Paralleling the pack will equalize it at what ever the avg SOC is. Light gauge wire is fine since current will be pretty low.

If you want to top balance, you need three things. A DVM that measures to 3 decimal places, a method to charge and a method to discharge. Ideally, for top balance, you want to be up around the 3.5 volt range or so, to get a decent balance. Charge the pack till you get close. Charge/discharge individual cells until you get them the same (resting), to three decimal places.

Why charge/dischareg individual cells to get them the same? Couldn't I just run the cells in parallel up to 3.65 volts and hold until the current tapers to a fraction of an amp? I was hoping if I did that, then disconnected the charger and then left them in parallel for an extra 24 hours, the pack would be well balanced at the top.

EVfun

Discussion Starter · #12 · Mar 1, 2011 (Edited)

gottdi said:
I understand. You going to use a zeva? I will be using one for amp counting.

Yes, that is what I've got. The other available amp hour counters should work equally well provided they are equally accurate.

Other ones that come to mind are the new unit from Dimitri, the TBS meter, the offering from Metric Mind, and the old E-meter.

I wanted to add that I don't KNOW this will work, but if I toast a cell or two on a top balanced pack it was the smallest cell anyway, right? I have a few spares so it would mostly be the time to rebalance the pack. I'm not present all the time when charging (but am while discharging -- driving) and I really want the cells to behave when I'm away. I want to be able to charge to 114 to 115 volts (32 cells) and let the current taper off without concern that any are going high in my garage at night. I want to do that without using a BMS. If my first attempt at counting amps out doesn't work out well perhaps I'll try another one.

See less See more

DC Braveheart

· #13 · Mar 1, 2011

I don't suppose some bright spark around these parts could come up with a simple (inexpensive

) circuit that could be attached to a cell to discharge it to a set voltage (say 2.8V) and then stop to allow easy bottom balancing (I'd have thought a zener diode in conjunction with a resistive load of some sort could be use to control that in some fashion ... but it's a bit beyond me).

Also, if I understand all the above, it's probably not a good idea to bottom balance and then use a shunt-type top-balancing BMS during charging? My logic is that if some cells have a lower capacity their voltage will go higher quicker during charging, the current shunt will kick in and they will end up with less stored charge. Then when you discharge them they'll go lower then all the other cells and possible get damaged? Let me know if I've understood this correctly.

See less See more

DIYguy

· #14 · Mar 1, 2011

EVfun said:
Why charge/dischareg individual cells to get them the same? Couldn't I just run the cells in parallel up to 3.65 volts and hold until the current tapers to a fraction of an amp? I was hoping if I did that, then disconnected the charger and then left them in parallel for an extra 24 hours, the pack would be well balanced at the top.

Sure... but do you have a charger that will charge at 3.65 volts? If so, do you have a few weeks/months to wait? lol

Just charge them in series...then balance them off... or tie them in parallel then check to see how close they are. Just depends how fussy u want to be I guess..

EVfun

Discussion Starter · #15 · Mar 1, 2011

If you are going to run a BMS with shunting regulators you shouldn't bottom balance. The effect of the shunt regulators is to top balance the pack. I would recommend top balancing to help the regs along (they generally can't shunt very fast.)

For the buggy pack I am planing to top balance without running a BMS. Discharge will be limited by amp hour counting.

DIYguy

· #16 · Mar 1, 2011

EVfun said:
If you are going to run a BMS with shunting regulators you shouldn't bottom balance. The effect of the shunt regulators is to top balance the pack. I would recommend top balancing to help the regs along (they generally can't shunt very fast.)

I would agree with that.

EVfun said:
For the buggy pack I am planing to top balance without running a BMS. Discharge will be limited by amp hour counting.

If you are going BMS-less, I would have to side with Pete. Bottom balance and cut charge short. The risk is much lower.

EVfun

Discussion Starter · #17 · Mar 2, 2011

DIYguy said:
Sure... but do you have a charger that will charge at 3.65 volts? If so, do you have a few weeks/months to wait? lol

Just charge them in series...then balance them off... or tie them in parallel then check to see how close they are. Just depends how fussy u want to be I guess..

I will initially charge in series because it is easy to do. I will terminate the charge so the first cell to rise doesn't exceed 3.6 volts. Then I will then parallel the cells and hopefully it won't take to long to bring them all up.

I figured I could just take my cheap 10 amp 6 volt charger and adjust the output by running the input power through a variac. I can put a motor run capacitor on the output (to read the output with a DMM) and adjust the variac to whatever I want. I would run that in the evening when I was home and hopefully they would come up within a few days. I figure it will be pretty easy to put in 32 amp hours per evening (about 1 amp hour per cell) and the cells would be starting out nearly full.

brainzel

· #18 · Mar 2, 2011

I get aditional 7 cells to my existing 38 cell pack (SE/CALB 130Ah) within the next days and so I have to deal with the same problems.

The new cells will be expected at ~50% SOC.
So I only have to options:
Charge them to upper knee 3.5V or discharge them to 2.8V lower knee and do the same with my "old" pack.

But what does this meen practical?
top-balance:
- charge the old pack in series to 3,45V / cell
- single charge each old cell up to 3,5V / cell and let them rest
- single charge each new cell up to 3,5V / cell and let them rest
- wire them together and drive the car down to max. ~20% SOC

bottom-balance:
- dirve the car down to ~20% SOC
- discharge each old cell down to 2,8V / cell and let them rest
- discharge each new cell down to 2,8V / cell and let them rest
- charge the old pack in series to 3,5V / cell
- wire them together and drive the car down to max. ~20% SOC

optional in both procedures, but perhaps unnecessary if you did a good job above?:
- wire all cells in parallel and let them rest/equalize before serial charge

"top" seems to be much quicker and easyer to handle theoretical, but my mind wants me to do it at the bottom ;-)

Are these the ways to go? Feel free to comment it please.

See less See more

ruckus

· #19 · Mar 2, 2011

EVfun said:
I'm about to parallel the buggy pack up to balance it. The pack is 32 Thunder Sky cells. I'm planning a top balance as it will be easier for me to limit the discharge and charging should be a little easier to adjust when they all come up together.

I've been pouring over the threads for recommendations about the setup and didn't find as much as I'd hoped for. Here is what I have in mind:

Code:

+ ------------------------ | | | | | | | | _ _ _ _ _ _ _ _ - - - - - - - - | | | | | | | | ------------------------ -

Well, mine will be 32 cells long but the idea is to put a ring terminal with 2 wires in it at each positive and each negative terminal and bring the positive and negative out on opposite ends so wiring resistance and any voltage drop from the charger should be equal (as much as possible) for every cell. Right now I have the cells in series so I will charge in series while watching to get them as close as possible before paralleling them. My plan is to balance them at 3.7 volts per cell as that is a little higher than my target charging voltage of about 3.6 volts per cell.

What size wire is recommended? I know the wire should be large enough to handle the charger current, but that only requires 16 gauge wire. Is larger needed to allow the cells to self balance? Each wire (from + to + or from - to -) will be about 6 inches long and the wires coming off each end for the charger will be about 3 feet long.

Will the cells move into balance slowly if they are not being charged or discharged? I'm in that big flat middle range with all 32 cells reading 3.31 volts each.

Is there anything I'm missing here?

Thanks for your input!

I just paralleled 30 Thundersky 260ah batts. I wouldn't worry about putting the + and - out different sides of the pack as they will automatically "balance" themselves if left in parallel for a bit. Even though I had done quite a bit of 12v motor testing and 12v car circuit testing (lights, etc), there was no discernible difference in the cell voltages and no spark when connecting them in parallel.

I hooked up a 3.65v .7A charger. I know it will take a year to charge the pack full, but if I had done this when the batts first arrived they would now be full.

..and balanced.

Please be VERY suspect of those who claim top or bottom balance is better. They are both equally problematic. If you read through the forums you will find multiple stories from those with LOTS OF EXPERIENCE that bottom balance and charge to knee with no bms DOES NOT WORK -sooner or later the weakest cell will be overcharged and go "pop" which causes the charger to charge even harder and the next cell goes pop on down the line until your pack is gone. It is JUST as dangerous as top balance and driving down to the knee only to have the same weak cell turn to dust.

This is why I started the bms thread to see if anybody had a good way to balance across the ENTIRE spectrum to avoid the Lilliputian arguments.
http://www.diyelectriccar.com/forums/showthread.php/future-bmsi-55352.html?p=230156#post230156

How many amps can you charge at 3.65v? my .7A is beyond weak...

I guess if I got 2 or 3 and put them in parallel it would only take 3-6 months.

See less See more

Roy Von Rogers

· #20 · Mar 2, 2011

ruckus said:
Please be VERY suspect of those who claim top or bottom balance is better. They are both equally problematic. If you read through the forums you will find multiple stories from those with LOTS OF EXPERIENCE that bottom balance and charge to knee with no bms DOES NOT WORK -sooner or later the weakest cell will be overcharged and go "pop" which causes the charger to charge even harder and the next cell goes pop on down the line until your pack is gone. It is JUST as dangerous as top balance and driving down to the knee only to have the same weak cell turn to dust.

If you start out with an empty 100ah battery pack and you fill it with 100ah, the battery pack if full, after all thats what you purchased. If any cells are above the lowest capacity cell due to variance in production, its unusable capacity. And if you voltage charge each cell to its max capacitance, you have an unbalanced pack, exactly what you DONT want.

Yes you can fill each cell to its full capacity and then use a downcounter to not exeed the max capacity of your pack, the qestion is why would you want to do so, since it puts the cell in to destruct mode if you ever get too close to the bottom, and there is no realiable indication while driving that a cell is going in to destruct mode. Remember these vehicles may be driven by your wife/daughter/girlfriend, so KISS needs to be applied.

Anyone that has watched JR at EV-TV charge a set of redwood sticks, should by now understand why these cells need to be charged that way.

No one that I know recommends that these batteries do not need some monitoring, thats what needs to be discussed. What is the best and simplest way to monitor a pack so anyone will have an indication that there is a problem, just like the check engine light on an ICE vehicle, and a fail save charging system in case there is a failure while the vehicle is being charged.

Roy

See less See more

1 - 20 of 79 Posts

This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.