Author Topic: balancing a string of LiIon cells  (Read 700 times)

0 Members and 1 Guest are viewing this topic.

dnix71

  • SuperHero Member
  • ******
  • Posts: 2280
balancing a string of LiIon cells
« on: December 19, 2017, 07:18:05 PM »
A coworker gave me a battery powered leaf blower to work on after both batteries quit. A quick Amazon review search revealed that fully 1/2 of the people who bought this POS would not recommend buying it to a friend. Many users reported that the batteries would work <10 times and then refuse to charge. A single 20v battery is $70. My friend got 2 batteries, a leaf blower and weed whacker as a set from Sam's Club for the same price.

I can charge the cells one at a time using jumpers and a plugin 18650 charger. All of the cells are just fine. The pack is utterly unserviceable. The cells are the type without internal protection and are double spot welded in series. The pack cannot even be disassembled without breaking half of the welds to slide out the cells from the molded bottom. The circuit board is soldered to the cell stack in 6 places. The whole pack uses tamper proof torx screws. The cells must have been expensive, as they are rated 2 amp-hours each and can withstand basically a dead short discharge into the leaf blower.

Is there some way to buy protection discs and charge the stack in series with the discs draining off the excess voltage? I can make room under the cover the hard way with a Dremel if need be. The charger is also junk, but I saved the case to use as a cradle when it's rebuilt.

https://www.amazon.com/S12020A-Lithium-Ion-Replacement-Accessory-130321002-2pk/dp/B015QJ1ZA8

The link goes to Amazon for a 2-pack replacement showing the assembled battery. Note the not-rational price for something that won't even last a season.

https://www.amazon.com/product-reviews/B00LVT19T0 Product reviews. Note how poorly made this was.

Simen

  • Sr. Member
  • ****
  • Posts: 441
  • Country: no
  • Grimstad, Norway
Re: balancing a string of LiIon cells
« Reply #1 on: December 19, 2017, 11:40:01 PM »
Have you tried to charge the pack by connecting 20 volt to the battery pack? The circuit board connected to the cells should be doing the same job, but better than protection discs - if it's not that circuit board that is faulty, that is... (Best to use a lab-psu type where you can limit the current to 1.8 A (0.8C).)

Do note that protection discs would only provide over-charge/-discharge, and not provide balancing, as the circuit board does.

With 6 wires from the circuit board would make it a 5-cell pack, and 20 volt would mean 4.00 volt/cell. I'm guessing that actual top-charge voltage lies at 4.10 volt/cell (20.5V). It is normal for power tool battery packs to have a top-charge of 4.00 to 4.10 volt instead of the normal 4.20 volt, to make it live longer...

(The little pin in the center of those 'tamper-proof' torx-screws, are often easily broken off with a small flat screwdriver... ;) )

But...
Since it was the battery packs that 'quit' and the cells seems ok, there is more than likely the CMS/BMS circuit board inside the packs that is faulty. (and may be the reason for the charger to quit.) It could also be one or two cells that can't handle the heavy load anymore (but seems to charge ok).
I will accept the rules that you feel necessary to your freedom. I am free, no matter what rules surround me. If I find them tolerable, I tolerate them; if I find them too obnoxious, I break them. I am free because I know that I alone am morally responsible for everything I do. - (R. A. Heinlein)

OperaHouse

  • Hero Member
  • *****
  • Posts: 1084
  • Country: us
Re: balancing a string of LiIon cells
« Reply #2 on: December 20, 2017, 03:26:06 AM »
In a world thatd oesn't want you to work on anything, it is time to buy a $4 security bit set.  Youtube is full of spot welders made from two turns of a microwave transformer.  A must if you continue to play with Li cells.  Factory likely went low bid on the BMS board and got just what they paid for. Or just one cell went bad.  I just wish I could fit 18650 cells in NiCad packs.

dnix71

  • SuperHero Member
  • ******
  • Posts: 2280
Re: balancing a string of LiIon cells
« Reply #3 on: December 20, 2017, 05:08:20 PM »
Simen I tried charging by bypassing the board, and yes, two cells seem a little weaker than the others. The last cell is isolated from the negative output tang by a thermal circuit. I tried bypassing the thermal itself with a straight wire and the pack would flash bad. If I added some resistance it would go into test mode and never leave. I assume from that that there is some resistance needed to bias a transistor to switch in the last cell for charging and device use and that circuit is fried.

This design is fail-safe but not at all sane. The current draw needed to operate a weed whacker or leaf blower is beyond anything reasonable for anything but nicads, which are basically immune to heat and short circuiting. I built a 20v pack using 18650's with built-in protection circuits for a different weed whacker and they work for about 5 seconds and then the drain kills them. If you let them recover for a minute or so you can repeat the process until they are discharged.

To make the battery pack work in the leaf blower I soldered a wire to the final negative terminal on the board and then wrapped the other end around the negative metal prong that goes in the device. I balanced the cells by charging each one one at a time with the little wall charger I have for loose 18650's and a pair of alligator jumper clips. That got all the cells between 4 and 4.1. The charger cuts off between 4.1 and 4.2.

Operahouse I have tamper proof torx bits, but two on the charger were recessed so even a bit on a holder wouldn't reach until I drilled out the shell hole a bit. There are also hidden screws on each battery which require destroying the cap that hides them to remove the cap. Somebody absolutely positively did not want anyone to touch the stuff inside (without them knowing it and voiding the warranty.

Simen

  • Sr. Member
  • ****
  • Posts: 441
  • Country: no
  • Grimstad, Norway
Re: balancing a string of LiIon cells
« Reply #4 on: December 21, 2017, 01:01:43 AM »
Those cells one can buy with built-in protection circuit are usually meant for single-cell devices, such as led lights and e-cigarettes, and are seldom/never used in series. Also, those circuit usually cuts out at 5-7A drain, usually too low for power tools (unless it's in a high voltage packs (48-80 Volt).

Have you measured the Amp usage of those whackers?

Li-cells can be designed for different usage, and those used in power tools are usually high-drain types. I have a Bosch lawn-mower with a 36 Volt, 2.6Ah battery with 20pc. Sony US18650VT cells @ 1,3Ah each. (2P10S) Each cell can handle 10A continuous discharge, and 1.2A charge, so the pack are good for 720 Watt continuous power drain.

But if one tries to use cells from, let's say Sanyo UR18650A (2,25Ah, max. drain 2.15A) from a laptop, on a weed whacker, you'll ruin the cells fast.

But back to your packs; :)
You're sure the last cell are electrically isolated by that thermal sensor? Usually, those sensors are only wedged between two cells for thermal contact... I'm guessing that that 'isolation' are the current sense-circuit, and it usually consist of a piece of thick wire on the circuit-board.

I assume your packs are 1P5S - for a weed whacker, i'd really prefer 2P5S, so that the load on each cell would be half.

If you really want to test your cells on the whacker, i'd remove the bms-circuit board entirely; charge each cell properly, and see how it goes on the whacker. But you should really monitor each cell as you test-drain the pack, and stop the drain as soon as one of the cells fall below 3.5V (no load; probably around 2.5V@max load).

As soon as you fint out the max. Amp usage of the whacker, you can start searching for a replacement bms-board. (which probably are the culprit here...) :)
« Last Edit: December 21, 2017, 01:14:25 AM by Simen »
I will accept the rules that you feel necessary to your freedom. I am free, no matter what rules surround me. If I find them tolerable, I tolerate them; if I find them too obnoxious, I break them. I am free because I know that I alone am morally responsible for everything I do. - (R. A. Heinlein)

dnix71

  • SuperHero Member
  • ******
  • Posts: 2280
Re: balancing a string of LiIon cells
« Reply #5 on: December 21, 2017, 07:51:35 PM »
I cut the circuit board off at the solder tangs, leaving the battery pack intact. This confirmed what I suspected but couldn't see because the bottom of the board was hidden.

This pack takes the 5 cells in series and runs them through a heavily heat-sinked pair of power mosfets. MXP 4002AT.

Those mosfets are dc-dc converter types. The cells are Samsung INR18650-20R which retail for $10 each. That explains the stupidly high battery price.

https://www.amazon.com/Samsung-INR18650-20R-2000mAh-Rechargeable-Battery/dp/B010EZGE6E

The person who designed this pack was 'tetched.' There is no way those tiny mosfets could push the power needed to run a leaf blower for long. There is a diode which appears to be set to prevent reversing the last cell and that diode is visibly fried. It looks like a 1N4007. 30 amps instantaneous will pop that diode in under 10 milliseconds.

This leaf blower does not have a soft start switch. It needed a pack twice the size it has. Same voltage but double the cells to even be close to being sane, plus a soft start of some kind to prevent the dead-short rush when the blower is turned on.

Or run the 20v motor at 15v. It wouldn't blow near as hard but it would run longer and wouldn't tear up the pack.

Now I'm still not sure what to do. Those are expensive, high quality cells. It would be a shame to ruin them in a leaf blower. My co-worker bought both packs and the tools for about $70. The cells alone cost $100 retail. That's kind of like printer ink used to be. I could buy certain inkjet printers, take the ink out and throw away the printer and still come out ahead.
« Last Edit: December 21, 2017, 07:59:16 PM by dnix71 »

Simen

  • Sr. Member
  • ****
  • Posts: 441
  • Country: no
  • Grimstad, Norway
Re: balancing a string of LiIon cells
« Reply #6 on: December 21, 2017, 11:00:41 PM »
If you really like the weed whacker/leaf blower those battery packs belongs to; you could make one 2P5S pack out of both packs, get a proper BMS/PCM, find a snug-fitting box to put the cells in, and leave the lid off the box but use the top/connector part of one of the old packs as a lid...

Finding BMS/PCM boards that can continuously handle the full 40A+ of 2P cells might not be easy, but most proper boards handle 4x inrush currents, so a 20A continuous board probably does the trick well... But there's a price to pay; example: http://www.batteryspace.com/pcmwithequilibriumfunctionfor185vli-ionbatterypack20amplimit.aspx

Edit;
Those Samsung cells are really nice (22A max. cont. discharge, 1A/4A charge) - you should really find a proper use for them. :)
/Edit
« Last Edit: December 21, 2017, 11:10:44 PM by Simen »
I will accept the rules that you feel necessary to your freedom. I am free, no matter what rules surround me. If I find them tolerable, I tolerate them; if I find them too obnoxious, I break them. I am free because I know that I alone am morally responsible for everything I do. - (R. A. Heinlein)

dnix71

  • SuperHero Member
  • ******
  • Posts: 2280
Re: balancing a string of LiIon cells
« Reply #7 on: December 22, 2017, 06:02:48 PM »
I tested the pack on the leaf blower using a Turnigy meter. 320 watts was the peak power used at 25 amps peak. The running watts was about 260 at 15 amps. The pack dropped to about 15v under load and held there. After running the blower for about 1 minute total I left the pack to cool down and the cell voltages were 3.75, 3.88, 3.90, 3.91 and 3.94.

Two of these packs together could start most small cars. I have a window a/c that draw just under 500 watts running, and a rice cooker that pulls about 400 watts.

Two of these packs would run the a/c or the rice cooker. The cells from both would fit in my shirt pocket. A Tesla has 7000 of these 18650 cells. I have no trouble believing the performance and range of that car now.

Simen

  • Sr. Member
  • ****
  • Posts: 441
  • Country: no
  • Grimstad, Norway
Re: balancing a string of LiIon cells
« Reply #8 on: December 22, 2017, 11:47:14 PM »
Two of these packs would run the a/c or the rice cooker.

That they would -  for 10-12 minutes... ;) (20V x 4Ah = 80Wh.)

3.75 Volt to 3.94 Volt are a bit too unbalanced for my comfort. i prefer to have LiIon cells within 20 mV to call them 'balanced'; more than 40 mV within the 3.55-4.20 Volt range, and i start to worry.

To get an idea of each cells health and ability to perform under heavy load, you should measure the internal impedance. your Samsung cells should have an internal impedance at <= 18 milliOhm when new. As the internal impedance rises with age, the heavy load capability falls and thus, the capacity under heavy load.

There are several  LiIon cell chargers that can measure internal impedance - i use an 'iCharger 1010B' :)
I will accept the rules that you feel necessary to your freedom. I am free, no matter what rules surround me. If I find them tolerable, I tolerate them; if I find them too obnoxious, I break them. I am free because I know that I alone am morally responsible for everything I do. - (R. A. Heinlein)

dnix71

  • SuperHero Member
  • ******
  • Posts: 2280
Re: balancing a string of LiIon cells
« Reply #9 on: December 23, 2017, 02:00:25 PM »
I just bought a new camera to replace the one that died when the batteries leaked in it. >:(

Here is the top and bottom of the circuit board. The top is the side with the tangs that go in the charger and tools. The bottom can only be seen if you remove the board, as there is zero clearance between the cells and circuit.




Mary B

  • Hero Member
  • *****
  • Posts: 812
Re: balancing a string of LiIon cells
« Reply #10 on: December 23, 2017, 04:19:58 PM »
I would test all the transistors and diodes, a curve tracer would be best and if you have an oscilloscope you can build one pretty cheap. The only other suspect is the big charge controller IC. The diode you say is fried is located where on the board? If it is where I think it is I suspect it is used to bias a transistor on and off and is not in series with the full current draw...