Results 1 to 8 of 8
  1. #1
    mtbr member
    Reputation: BrianMc's Avatar
    Join Date
    Jan 2010
    Posts
    3,684

    Free Li-ions: wish to build a pack

    The testing 18650 thread has prompted me to test 6 Samsung 18650 cells I have.

    I have looked at commercial solutions, and with batteries in hand, a DIY saves a fair piece of change, and gets these batteries used until they die and not recycled without being used (a shame).

    I would like to make a 3S2P pack with them if all 6 pass after testing is compete. These are unprotected cells so a PCM's is essential as I like my house, bike, and hate burns. The balancing PCMs and chargers I have found are for 3 series packs.

    Questions:
    1. Can I use two 11.1 volt PCMs, one per 3S set, with no interference?

    Looks like I could, but then I would have two unbalanced sets of three balanced cells. If I pair the cells in parallel they will look like a larger single cell, the pairs are balanced though the cells in each pair may not be. A commercial pack I could examine used a single PCM with six cells paired like this.

    2. Will a level meter on one report whole pack's condition close enough?

    With one PCM to feed a meter it reports pack level: solved.

    3. Can an 11.1 V charger charge both 3S sets in parallel but at half current to each?

    If using three 2P sets, the charge rate to each cell is halved and charge time doubled but all current goes through a single PCM so there's no possiblity of different charge rates on each set.

    Status:
    The cells took a full charge but I have to do the resistive discharge tests to confirm they are worth the trouble.

    Battery U stated that PC battery makers isolate the cells of their packs to reduce the chances for a runaway cell affecting its neighbors. Anyone build a pack with some isolation, if so, what did you use?

    Comments?
    Last edited by BrianMc; 08-03-2010 at 07:09 AM.

  2. #2
    mtbr member
    Reputation: BrianMc's Avatar
    Join Date
    Jan 2010
    Posts
    3,684
    The 6 cells came to me in pairs connected in parallel.

    They are marked as 24E or 2.4 Ahr each, so nominally 4.8 Ahr for the pair. At least 2 X the capacity to dissipate as in OldMTBfreak's method from the testing thread. I had three 0.47 Ohm 5 watt resistors which in series give a nominal discharge rate about 10% higher than the 4.7 Ohm on the lowest capacity cells that would not hit 3.2 volt for 2 hours.

    Tested battery voltage every 15 minutes, until voltage dropped to 3.61 at 105 minutes then every 5 minutes. The pair produced 102% of nominal Watt-hr rating at 105 minutes, and hit 3.27 Volts at 115 minutes.

    Following 0.5 C discharge rate chart (second one down on this web page:

    http://www.ibt-power.com/Battery_pac..._ion_tech.html

    If the other two pair test as well, I'll order the PCM, gauge and charger. The first two I have decided on and will get the trailtech connectors I need to share the shipping cost.

    I am not as sure about chargers. My NiMH smart charger is stopping at 13.6 V and not topping the old pack off to its maximum 14.1 V. It doesn't seem to be trickle charging anymore. It's gotten dumb. The separate overnight trickle charger I have for it has no temperature cut-off but will bring it to 14.1 V. Reading in older threads about batteries I saw comments about chargers failing after some time. So I can buy two chargers one for each chemistry with a voltage range for the NiMH one, or for a bit more, a universal charger covering different chemistries including ones I do not yet have and a wide range of voltages for lights I haven't thought of yet, so it may have a longer useful life, for only a bit more than two chargers cost. Anybody have one of these?

    http://www.all-battery.com/tenergytb...i-fesla-1.aspx
    Last edited by BrianMc; 08-05-2010 at 06:41 AM.

  3. #3
    mtbr member
    Reputation:
    Join Date
    Apr 2006
    Posts
    594
    Hi Brian, The thing to remember about about using a resistor to discharge the battery is; as the voltage drops, the current through the resistor drops. I=E/R So the wattage requirement for the resistor is for a FULLY charged battery. Resistors can generally be be overdriven by 2X without much complant. Me, I have a handfull of 10 ohm 100 watt adjustable wirewound resistors that I use for these tests. Your methodology (for your test) looks good. I write the capacity on each cell, after testing. To save money, if I'm gonna assemble the battery in 2p or 3p configuration, I parallel 2 or 3 cells. Then hook these in series. This lets me use only one of the protection PCB's. BTW, I feel the protection PCB to be very important. I have had a few little "oopeses" in my shop with the battery cells. Usually resulting in burns, smoke, curseing, danceing around, screaming, and spilling the bong; you get the picture. The protection PCB prevents this with the completed battery. Before this operation however, I ALWAYS match cells. I usually buy a few extras. I have never salvaged used li-ions though. I feel the short life of lithium batterys requires that I buy new ones for my projects. Fer instance, my laptop battery only lasts about 3 years. By 2 years, run time is pretty short. My laptop has eight 18650 cells in the battery, I have rebuilt it twice. It's starting to wimp out again, I guess it'll require another rebuild this winter. I guess I'll bring up one more point. I started putting balancer circuitry in the batteries this spring. You can get protection PCB's that also balance. I just made an external balancer that plugs into the battery during charging. The battery will have a longer life and can deliver more power when all cells are charged to the same potential. James

  4. #4
    mtbr member
    Reputation: BrianMc's Avatar
    Join Date
    Jan 2010
    Posts
    3,684
    Quote Originally Posted by OldMTBfreak
    Hi Brian, The thing to remember about about using a resistor to discharge the battery is; as the voltage drops, the current through the resistor drops. I=E/R So the wattage requirement for the resistor is for a FULLY charged battery.
    Thanks for the leg up here, James. I picked up on the 4.2 and 3.2 Volt start and end voltage currents of 0.9 and 0.68 from your example and recognized this would be fairly straight line decrease in current flow until about 3.6 volts where it dives.

    The cell pairs were 4.18-4.19 V to start. I decided to test every 15 minutes and record voltage in a spreadsheet, calculate the mean current that 15 minutes the watt-hrs consumed. I am starting at 2.97 A and dropping to about 2.3 A. I summed the Ahrs and Watts used and they match the nominal ratings. .

    Quote Originally Posted by OldMTBfreak
    Resistors can generally be be overdriven by 2X without much complant. Me, I have a handfull of 10 ohm 100 watt adjustable wirewound resistors that I use for these tests.
    Each reisistor is .47 Ohms 5 watt wire wound so at max current of 3 A, only 1.5 watts so underdriven by a factor of more than 3. They do get toasty, though. If I am going to do more I will improve my toolset.

    Quote Originally Posted by OldMTBfreak
    Your methodology (for your test) looks good. I write the capacity on each cell, after testing. To save money, if I'm gonna assemble the battery in 2p or 3p configuration, I parallel 2 or 3 cells. Then hook these in series. This lets me use only one of the protection PCB's. BTW, I feel the protection PCB to be very important. I have had a few little "oopeses" in my shop with the battery cells. Usually resulting in burns, smoke, curseing, danceing around, screaming, and spilling the bong; you get the picture. The protection PCB prevents this with the completed battery. Before this operation however, I ALWAYS match cells.
    By Ahr and how wide a gap will you accept?

    Quote Originally Posted by OldMTBfreak
    I usually buy a few extras. I have never salvaged used li-ions though. I feel the short life of lithium batterys requires that I buy new ones for my projects.
    These were charged three times, so are essentially new, or I'd not bother with them either. In the pack, they would not hold a charge, and it looks like the quiescent current of the circuitry was a bit high.

    Quote Originally Posted by OldMTBfreak
    I started putting balancer circuitry in the batteries this spring. You can get protection PCB's that also balance. I just made an external balancer that plugs into the battery during charging. The battery will have a longer life and can deliver more power when all cells are charged to the same potential. James
    I am opting for the PCM with balance as it will monitor discharge as well and the extra $6 on one pack isn't going to break the bank. These pairs have nice connections I can't really improve on, so I will hope the cells in the pairs are as well mated as the pairs are.

    From another thread here, one person said they did packs with balance charging and without and saw little differnce in lifespan. Since depth of discharge and rate of discharge play a large role in battery life a 5-10% gain would not be noticeble but would more than pay in delayed replacement. I suspect this is a bit overkill, but peace of mind for the DIYer is worth a couple more bucks. Your testimony shows the reality of the need to take all precautions within reason.

    All test OK, so now the decision on how to charge them.

  5. #5
    mtbr member
    Reputation:
    Join Date
    Apr 2006
    Posts
    594
    Lord Brian, lots of good points. I have found that the batteries charged on the balancer have a noticeably longer run time (no numerical data yet). It will be very nice if balancing extends the batteries life. After testing the new cells, I strive for > 10% difference. I usually buy a few spare cells with each purchase. I "cycle" each new cell 4 or 5 times before matching. I feel that you should treat the Li-ion batteries like eggs. I don't charge at a very high rate (.5 C Max), or discharge at a high rate (my highest draw light is 2.2 amps). I use Taskled controllers and set the cutoff at 3 volts per cell. I charge my batteries using a lab quality CV/CC power supply. I charge in this manner; 1st I set the current for around .5 C. Then I set the PS voltage .1 volt per cell low. If it's a 5-cell battery, then instead of setting PS to 21 volts, it's set to 20.5 volts. I then wait until the charge current falls to 0. Then I plug the balancer in and reset the PS voltage to 20.9 volts. I then wait until all the balancer leds are on; battery charged and balanced. It looks to me like this the more cells, the more the balancer is required. James

  6. #6
    mtbr member
    Reputation: BrianMc's Avatar
    Join Date
    Jan 2010
    Posts
    3,684
    Quote Originally Posted by OldMTBfreak
    Lord Brian, lots of good points.
    Battery University is good but this real-world experience is better. Thanks, James.

    Quote Originally Posted by OldMTBfreak
    I have found that the batteries charged on the balancer have a noticeably longer run time (no numerical data yet). It will be very nice if balancing extends the batteries life.
    Since the lowest capacity cell hits max voltage first, it limits use of the other cell's capacities for storage or discharge and will be the first to go when the pack fails. One goes, they all go. Makes a lot of sense that balancing increases capacity noticeably, and may hold off failure for a bit, if not a not a remarkable amount (need very good records to pick up a 5-10% gain given unequal treatment packs get over different rides).

    Quote Originally Posted by OldMTBfreak
    After testing the new cells, I strive for > 10% difference. .... I "cycle" each new cell 4 or 5 times before matching. I feel that you should treat the Li-ion batteries like eggs. ....It looks to me like this the more cells, the more the balancer is required. James
    I assume you meant <10%.

    Nice tip on the 4-5 cycles before final selection. So I have a couple to go, yet. Since these are almost new I could order a few if there is a problem cell or two and cycle them as much to make them equal.

    The Li-ion discharge curves show not much life difference in longevity between 0.1C and 0.5 C discharge rates, but it gets shorter fast at higher charging and discharging rates. Your thought that higher voltage packs have a greater need for balancing is borne out in the Battery Management Systems required to safely operate really large packs, and I read that somewhere recently.

    Battery U suggested that the military increased Li-ion pack longevity with larger capacity packs, charged them to say, only 3.9 volt or about 80% capacity and discharge it to about 3.5 volts (about 10% capacity). That would still be lighter than NiMH with no memory and many more cycles with a higher capacity approach to end of life. We'd curse minutes from home or the trailhead and car if our lights went out if we knew we could have loaded another 20% capacity. An extra pack costs us more in weight penalty and outlay.

    Sounds like good well managed chargers = long lived batteries, too. Not to mention fewer Ooops moments. So, unless I hear ill of that charger, I will order one. The version without the AC supply had four stars from 3 reviews, so it looks decent.

  7. #7
    mtbr member
    Reputation:
    Join Date
    Apr 2006
    Posts
    594
    So uh which charger is that? lol I worked for the USAF for many years as a systems engineer. I had a couple of battery operated projects that could not be allowed to fail. So I bought fancy battery chargers/testers/analyzers and I bought a lot of batteries. The silver-zinc batteries were cool; thousands of dollars, one use. I spent your money like water. Li-ion batteries hadn't been developed at that time. I make my own packs and I feel matched cells are the best way to get both a longer run time and a longer life. I'm using a separate balancer now that plugs into the battery during charge. I'm going to buy protection PCB's with the balancer function built in next time. James

  8. #8
    mtbr member
    Reputation: BrianMc's Avatar
    Join Date
    Jan 2010
    Posts
    3,684
    The first PCB did not like me checking voltages across the pairs. Shorted out the last set. Just got the second PCB (non-fried) installed. It won't allow output if the cells are too far out of balance and the frying was on one set so I had to charge again to balance the sets and get output.

    The first use after charging at 0.5 A, was for two separate 30 minute runs at 0.5 Amp draw about 14 hours apart. I unplugged at the destination, and on setting out again, found that the PCB had stopped output. Since then, I have run it through two discharge/recharge cycles going deeper each time. The last one to 90%. The cells were through two very deep discharges already, so this will be the last before going into service if the discharge without incident. Capacity is good.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •