Hey Eric. Good to have you back. I missed your big type.
Much easier on older eyes.
There's no owner's manual, per se, but the building and testing instructions are
here. I do, however, think a Help function would be great. If nothing else, it might keep Richard from answering similar questions so much.
The order page is here:
http://www.tppacks.com/products.asp?cat=26. I'm finally getting around to adding a few more items. I've been working on some things that are a bit more LiPo-oriented, seeing how many here are going that direction, myself included. With LiPos, the tendency is to make the packs compact and lightly wrapped. They don't really lend themselves to having a heat-generating full BMS board co-located with the cells. For that reason, I've split the LVC and charge functions into separate boards. The LVC portion can and should be co-located with the cells. For the best performance, it is better to parallel cells, and then put the parallel blocks in series. The LiPo packs many of us have been using usually come as six 5Ah cells in series. Each of these 6s packs have 7-pin balancer pigtails and 10-gauge main power leads. To do the parallel first, then series technique, you really need to connect the balancer plugs together, so I've included onboard connectors for this on the LVC board.
For charging, I'm doing a new balancer that is based on independent 6-channel sections. It is similar to the existing BMS cell circuits, in that shunts are involved to hold the high cells at a cutoff voltage, but there are some significant differences. First of all, LiPos don't normally have the longevity of LiFePO4's supposed "thousands of cycles". Most RC LiPo packs last for hundreds of cycles, not thousands. What many have found, however, is that pack life can be significantly improved if they don't get charged to the full 4.2V per cell, but to 4.10-4.12V per cell. The difference is only 2%, but it seems to make a big difference. Anyway, what I found is that if a lower cutoff voltage is used, there's a lot more voltage headroom available for high cells to rise a bit above the cutoff and keep the current up higher for a longer period of time. With the current BMS circuits, the cutoff is set to whatever the max voltage that the cell is supposed to be charged to, which is 3.7V for LiFePO4 and 4.2V for LiPos. The 4.2V value is picked in order to make sure the cells don't get to the danger point. Cobalt-based LiPo cells will start exploding if the voltage gets too much over 4.3V.
In any case, what I discovered is that by using an even lower shunt voltage turn on point of 4.10V, and by increasing the shunt current to about 650mA, I can still keep the per cell max voltage well under 4.2V, even with 10-20A charge currents. I did some worst-case tests where I purposely had one block of cells (in a 6s3p 15Ah pack configuration...) at a voltage that was 200mV higher than the other five blocks. This is a very unrealistic difference for healthy LiPo packs. All RC LiPo chargers won't even let you charge the pack if there is this much difference. They tell you the pack is done, and needs to be discarded. Anyway, I then charged the whole pack to 4.10V x 6, or 24.6V, at a 10A rate. The shunt came on for the high cell way before the rest, but since there is no "throttling" of the overall current involved, the rest of the cells had the full 10A to keep charging. The high cell's voltage kept rising as well, but because of the shunt, it was at a much slower rate. The high cell's voltage peaked at about 4.18V. It took awhile, probably a couple hours but eventually the high cell came down, and the pack ended up perfectly balanced. If throttling was involved, it would have been a
lot longer, probably overnight.
The other difference in this new balancer is that I added a bit more logic so that a bi-color red-green LED could be used on every channel. What happens is that it starts out red. When the cell's voltage hits about 4.07V, it starts to transition to green. At about 4.80-4.85V, the LED is orange, and when it hits the point of when the shunt comes on, at 4.10V, it is fully green. Depending on the charge current, this transition takes about 2-3 minutes. With reasonably balanced cells, they usually all start to transition within seconds of each other. This gives a good indication of the relative balance of the cells in a pack.
Finally, by eliminating the need for current-limiting "throttling", each 6s balancer section is fully independent. That means you don't need to worry about connecting 6s "sub-packs" to the balancer in any particular order. The new LVC boards have a single 7-pin output pigtail that can be plugged into any 6s balancer section. You also don't have to worry about overlapping a cell, like you do with RC balancers, which balance to the lowest cell. This balancer balances to a set voltage point (4.10V...), so they will all end up at the same place. The initial board layout I did is for two 6s sections. Here's what it looks like:
View attachment 12-Cell Charge Balancer-v3.3 - PCB.jpg
It is sized to fit between the rails of a small extruded aluminum box. The shunt resistors mount from the bottom of the board, and make thermal contact with the bottom of the case. This turns the whole box into a heatsink. In order to handle the higher shunt currents, the higher power BD136 shunt transistor is used, and is mounted horizontally with a small heatsink pad underneath it. I'm working on the first "production" units right now, and will hopefully get these up on my site later today. Because these are a ton simpler than the existing BMS, in terms of parts count, I will offer these in various kit options, and I may also offer a fully assembled/tested version.
-- Gary
