View Full Version : Can anybody explain why Thundersky EMS BMS pack at 17V charger only 15V capable?

26 February 2011, 1009
Those with data loggers on their GPR-S motorcycles might be able to solve this battery murder mystery. (However, the battery might not be dead - has not been put on the road after charging yet.)

Has anybody ever seen this before? Another Electric Motorsport BMS and/or Thundersky (now Winston) LIFEPO4 mystery problem that should not have happened.

There are 4 packs (16 cells) in this Electric Motorsport bike. One of these prismatic LIFEPO4 packs was charged with a .5A (low power) "wall wart" trickle charger that only puts out 14.8VDC, confirmed by three good quality multimeters. (one meter is a Hioki)

The Thundersky pack was 13.2V before charging. It has an Electric Motorsport BMS which may, or may not, have High Voltage Cutoff (HVC).

After a 7 hour charge, this Thundersky LIFEPO4 pack is over 17V. Each cell has over 4.2 volts, with one cell nearly 4.4V. The cells have not swelled or ballooned, and are cool to the touch. (This may be a surface charge, though.) The pack voltages were confirmed by a second multimeter.

Can anybody try to explain this mysterious behaviour?


The Electric Motorsport BMS has LED lights. Two LEDs are in steady state (the BMS units are not longer connected to the master BMS board.)

The wall wart puts out 1.22V AC as measured on a Wavetek multimeter.

26 February 2011, 1541
I don't have an answer, all I can do is relate how my Modalis Rev 2 BMS worked on my GPR-S. Each board had (note past tense) six small green LED lights, one for each cell in the 4-cell pack under the board and two others that would turn red, but I can't recall why. The little green LED lights would turn on as each cell would hit about 3.6 volts, then that cell would stop charging.

At some point (when it was working properly) every cell would show a green light and the charger would stop charging. At that time the "72V" pack would read about 84 volts, which would drop to 80 volts as soon as the throttle was turned on (which would drop to 78 volts within 100 yards).

After a while one or more lights would go off and the charger would start charging again for a while, until all the lights were green again. This all worked just fine for about 70 charge-discharge cycles and then everything started going weird and the last pack in the line, with the "master board" on top, started to eat batteries.

I thought the batteries were going bad, but all replacing them did was to cause the replacement batteries to expand and their voltage would drop rapidly when being discharged to around 1/2 volt. After eating up several batteries, I gave up and tossed the Modalis for a mini-BMS, but the last two batteries in my pack continue to have problems. So I don't have a clue just what changed from the original set-up with no real problems, to the current situation where the last two batteries in the 24 cell daisy chain keep going bad.

Attached is a photo of the Modalis master board, showing the LED indicator lights. If I recall correctly, the center lights turn green and the two outer lights turn red.

27 February 2011, 0811
I just remembered what the two outside red lights indicate. They indicate that the batteries are connected to the BMS and that the BMS is doing something.

27 February 2011, 0848
The voltage reading on this pack is down to 13.93V.

Incidentally, the other TS LFP packs have been charged by the same wall-wart and voltage increases (same number of hours) have only been .1V or .15V higher, from the charging.

27 February 2011, 0924
Are your "last two batteries" the ones connected to the PLUS or the MINUS of the charger?
What (replacement) brand charger did you use with the mini-BMS setup?

27 February 2011, 1245
The Modalis uses shunts to perform balancing and allows an individual cell to reach 4.2v. When a cell reaches 4.2v, the shunt bypasses extra current and allows the other cells to continue charging. This occurs until all the cells reach 4.2v. After removing the charger, the surface charge dissipates and the cells settle down to ~3.5-3.6v (eventually they'll drop to 3.3Xv but that might take a while).

The Modalis is designed to interface to an SSR to turn the charger on/off. When a cell hits 4.2v, it shuts off the charger and bleeds the high cell for a while. Then it resumes charging until all cells match a certain level. Sounds like the wall wart is on continuously and forcing charge into the cells when the BMS should have turned off the charger. In any case, even with charger control, 4.2v is way too high for the cells.

27 February 2011, 1252
The Thundersky pack was 13.2V before charging. It has an Electric Motorsport BMS which may, or may not, have High Voltage Cutoff (HVC).

Unless you have built a circuit connecting the BMS to a circuit that controls the wall wart, whether or not it has HVC won't matter, as the HVC won't be able to protect the cells from overcharging by the wall wart.

27 February 2011, 1628
Are your "last two batteries" the ones connected to the PLUS or the MINUS of the charger?
What (replacement) brand charger did you use with the mini-BMS setup?

I am still using the same charger that came with my GPR-S. Photo attached.

Attached is a photo of the last pack of batteries, showing how they are hooked up using the Mini-BMS. The two batteries at the left side of the pack are the ones that keep going bad. From the attached photo, it looks like the take-off for the power is coming from the plus terminal. That is the way my pack was set up by EMS.

When my Modalis BMS was working, I checked the voltages on my batteries several times and none of them charged more than 3.6 volts. After charging to this voltage, they would shunt back and forth and settle at around 3.35 volts. I have never seen more than 3.6 volts on any of my Hi Power battery cells, using the Modalis system or the mini-BMS system.

But that never kept the two batteries on the left from expanding and shorting out, but this only has occurred when they were supplying power to the motor, never when being charged.

08 March 2011, 1315
A technie explained the cause, which I confirmed with a Google search: The AC ripple must be added to the DC voltage for the "total voltage" that the battery will receive.

08 March 2011, 1317
Some BMS have large resistors to burn off excess voltage.

08 March 2011, 1438
Yup NonPolluter, some Battery Management Systems use resistors.... but I'd add that MOST battery "balancers" that I've seen use a resistor (or a FET in dissipation mode).... Some BMC are LVC and HVC only, just the cheap solutions for those that want to balance themselves.

I think there were a couple capacitor based ones that shuttled charge from the high batteries to a low battery (Powercheq's). I haven't seen any commercially available power conversion BMS's (they use an isolated DC-DC converter on each battery)..... but I've built one and it worked great (essentially a uC controlled Isolated DC-DC on each battery). I think those are basically all of the types.... resistive, capacitive and power conversion. Most of the commercial ones are resistor based, which can be problematic if they fail and are failed ON, they discharge the battery.