[PTCruisin]

I finally got the TS100Ah cells installed and wired this weekend and I am wondering if I should just "go for it" and connect the PFC series charger or use a single-cell charger to bring them all up to 3.7V?

The 38 cells in the string are all at 3.309V (plus or minus 1mV) right now, so I assume they are fairly well balanced and at approximately half capacity (as shipped). Also the total voltage measured is what is expected from the string.

The series charger is setup for a CV of 3.65V per cell so I think the possibility of overcharging is slight but I certainly don't want to find out the hard way that I've overcharged a cell (or two) I have a BMS in development based on cheap celllog8's so I could monitor the voltages during charge and shut off the charger when any cell reaches 3.7V

Charging each cell individually (CC/CV) to 3.7V is possible, but very time-consuming.

What to do?

 

[illuminateddan]

LiFePo chemistry really needs the CC/CV charge. If you just try to pump 3.7v into them the charger will struggle as it will not be able to supply the 3.7v as the cells draw a wopping amount of current from it. Bare in mind that 100ah cells can easily take 100a + charge without even heating up!

Personally I would be performing a 'by the book' CC/CV charge on the cells as a string (or a couple of strings) to get them all to 3.7v.

A really great (if long) video is Jack Rickard's friday shows- the December 18, 2009 SPECIAL - LiFePo4 Battery Charging Basics Constant Current/Constant Voltage Charge Curves for Sky Energy and Thundersky LiFePo4 Cells

He basically spends 105 minutes explaing battery chemistry and the charging, and then charges a cell. The video shows exactly what happens and why it does. Lots of people bag Jack for his views on balancing, but I have found no one else who spends as much time producing such informative videos with this level of detail.

Also worth watching is his video on balancing. you may agree or disagree, but at least you'll have a broader view of the subject and may save yourself some dead (murdered!) cells!!

Hope this helps, good luck!

Dan

 

[dexion]

I really should have taken pics of when I got my cells. I laid them out on the kitchen floor. Hooked them all in parallel (using 10G wire) pumped 3.65V at 15 amps from my adjustable power supply for about a week until all the cells came up. That way saves you the hassle of watching them and switching cells.

 

[mhud]

I really should have taken pics of when I got my cells. I laid them out on the kitchen floor. Hooked them all in parallel (using 10G wire) pumped 3.65V at 15 amps from my adjustable power supply for about a week until all the cells came up. That way saves you the hassle of watching them and switching cells.

I did a similar thing for my lead acids. I just used this coated angle (I think zinc-coated, but I'm not sure) since the resistance on it was very low. Something like that would work well for the LI cells since they're so close together.

Just be REALLY careful not to create a short. Cover up all the opposite-charge leads until you have all the first set wired up and secure.

 

[PTCruisin]

LiFePo chemistry really needs the CC/CV charge. If you just try to pump 3.7v into them the charger will struggle as it will not be able to supply the 3.7v as the cells draw a wopping amount of current from it. Bare in mind that 100ah cells can easily take 100a + charge without even heating up!

Personally I would be performing a 'by the book' CC/CV charge on the cells as a string (or a couple of strings) to get them all to 3.7v.


Hope this helps, good luck!

Dan

Thanks Dan, I've watched all of Jack's videos on LiFePO4 batteries and I agree with him mostly. My plan is to charge them with the Elcon PFC charger using the CC/CV charge profile with the CV set to 3.65V/cell (138.7V). I will monitor Ah/Wh drawn from the pack under load and set the empty point at 80% DOD.

I intend to babysit the initial charging and monitor with cell logs on each cell (the cheap ones). Eventually, if needed, I'll connect opto-relays to the alarm ports on the cell logs and use the HVC to shut down the charger if any cell reaches the limit.

Early on I bought a homebrew single-cell charger from Frodus that is a 48v switching supply driving a 3.7v vicor dc-to-dc converter supplying 20 amps. At that time I believed that top balancing was the way to go, now I'm not so sure.

 

[illuminateddan]

Thats pretty cunning!
Just be careful though, LiFePo4 is a COMPLETELY different system to FLA or AGM (lead acid) cells. Each has it's own unique charge curves, methods and floats (lifepo has no float though!).

Charging lifepos with a FLA curve or method would be like trying to make a ham toastie using a chocolate cake recipie. Both are cooking, but both are completely different.

 

[PTCruisin]

I really should have taken pics of when I got my cells. I laid them out on the kitchen floor. Hooked them all in parallel (using 10G wire) pumped 3.65V at 15 amps from my adjustable power supply for about a week until all the cells came up. That way saves you the hassle of watching them and switching cells.

Hey Dex, I remember reading your post on that and thought that was the way to go, but I got anxious the get the cells in the car so I wired them in series.

How long have you had the cells in service and have you noticed any large difference in cell voltage/capacity? Have the cells remained balanced over multiple charge cycles? Thanks!

 

[dexion]

The cells have about 250 cycles on them now to about 40% dod. So, that may be 125 80% dod cycles but who knows at this point. I really havent noticed any spread in the cells. They are bang on unless I go down below about 70%dod which I have done about 10 times. There is one cell that gets to 3.7V before the others by about 30 seconds at 6amps so I just hit that one with a resistor before charging for about .2amps. Other than that they are all spot on. There doesnt appear to be any loss of capacity as well (I just went to 65% dod today and all the cells read about 3.294V so, a ways to go yet.) But I am pretty easy on the cells. Really its 30% going to work and 40% going home (more up hill) with charging at work.

 

[Sun Motors]

I brought each cell up to the recommended 4.2v for the first charge after I had the cells all in place. First brought them up to 3.7v or so with a series charger and then did 4 cells at a time with a small multi charger to reach 4.2v. Took a couple of days, and watching the cells when the voltage was rising fast was critical. But I didn't find much authoritative advice on how to do this either.

Now I just charge them up until the first ones hit 3.6v and then drive them down 50-80aH (less that 50%) and they seem to be holding together within 20mv so far without any top or bottom balancing. One cell is starting to act a little strangely though and I'm about to try adding a couple of aH's to see what gives.

 

[JRP3]

No need to charge them to 4.2, in fact TS now lists 4.0 as max voltage. I know some claim you have to charge to 4.0 for the initial charge but as I've pointed out repeatedly the cells have already been fully cycled at the factory to check resistance and capacity.
People really make charging LiFePO4 cells more complicated than need be. You don't need to fully charge them every time and will probably get longer life from them if you don't. Since you have Cell logs just set them to beep at 3.7 or so and charge in series till the first one starts to beep and then watch them closely or stop charging. Remember that the higher amperage used for charging the voltage will climb higher sooner before being full, so if you hit 3.7V at 30 amps there will be more capacity left in the cell than if you hit 3.7V at 10 amps. Also, if you balance the cells at the top they will be unbalanced at the bottom, and top balancing will not give you any more capacity than your smallest cell. Top balancing will make unattended charging safer but then makes a low SOC event more dangerous.

 

[JRP3]

Another advantage of bottom balancing is you know your cells capacity and which ones are the smallest because they fill up first. With top balancing you don't know until you pull the smallest cell down near empty while you're driving.

 

[hipowerkevin]

I would suggest first time, charge each cells individual, them assemble for battery pack, but not assemble for pack then charge in series.

To build a nice battery pack, all cells should be fully charged to CV mode, and keep cell's voltage difference as small as possible,you can see CV volt volume in datasheet.

then, you could discharge battery packs, like to 80% DOD, then fully charged back by charger in series, to see that all cells in closest volt? it is important.

Two things need to make sure: cells can not be over charged (voltage higher than charge cut-off voltage), cell can not be total discharged (voltage lower than discharge cut-off voltage). Hipower battery is 3.85V and 2.0V, but I am not so clear with thundersky.

By the way, which type of BMS now you using? is it well negotiated with your charger? the charging curve and charging status is quite important and should be well communicated between BMS and charger by CAN BUS 2.0b protocol or so.

Also, BMS could monitor the cells status, you could view all detail info well on screen, or download to your computer.