Even Newer 4 to 24-cell Battery Management System (BMS)

[ejonesss]

will the new design allow the use of a standard car battery charger? or even the 5 or 12 volt output from a computer power supply?

 

[toyguru]

I'm happy with just the board and parts list, I think it would be a great learning experience to build my own. I'll be using it for 12 TS cells, so I'll have to experiment to find the right components to get the 4.25V cutoff. I'll probably order a few boards, in case I screw up the first one.
Thanks again for all the hard work on this!

 

[methods]

I am getting excited. . .

 

[ahambone]

Hello Everybody,

Long-time lurker, first time poster. Thank you to everyone who has posted over the last few months and makes this thread really enjoyable. Gary, when the new 24s BMS boards become available you can count on me ordering a couple for tinkering with them as full blown EV-car BMS.

On a side note for everyone in this forum, I have an interesting lead on LiFePO4 cells. Like everyone else here, I keep looking for a more legitimate way of finding A123 caliber LiFePO4 cells (beyond their developer kits and gutting DeWalt packs). I found a vendor (http://www.voltphreaks.com/) that sells loose 18650 LiFePO4 cells online at reasonable discounts. Their cells are smaller capacity than A123's 26650s that I've been testing against but they look like they are cheaper Wh/$. I wouldn't mind soldering 3 or 4 of these cells together to replace and A123 unit. I have ordered 80 of their cells to play with (you have to order 50 to get a volume discount). If anyone is interested, I will post my testing results to the appropriate forum.

Cheers,
--Adam

 

[fechter]

will the new design allow the use of a standard car battery charger? or even the 5 or 12 volt output from a computer power supply?

Yes, just about any source that is current limited and has enough voltage will work.
A car battery charger might be a bit low for 4 cells and certainly a 12v output from a PSU would be too low. You can tweak the voltage up on these in many cases.

 

[Tiberius]

Gary,

A board and parts list would be fine. I'd like two please.

Just having the PC board is ideal for my circumstances - its easier to ship, and I will have some of the parts already.

I was going to offer to be a focal point for one shipment to the UK and then redistribution. But on reflection, I think that could work out more expensive overall. Its highly likely to attract customs and tax, whereas multiple small envelopes may not.

Nick

 

[GGoodrum]

Hello Everybody,

Long-time lurker, first time poster. Thank you to everyone who has posted over the last few months and makes this thread really enjoyable. Gary, when the new 24s BMS boards become available you can count on me ordering a couple for tinkering with them as full blown EV-car BMS.

On a side note for everyone in this forum, I have an interesting lead on LiFePO4 cells. Like everyone else here, I keep looking for a more legitimate way of finding A123 caliber LiFePO4 cells (beyond their developer kits and gutting DeWalt packs). I found a vendor (http://www.voltphreaks.com/) that sells loose 18650 LiFePO4 cells online at reasonable discounts. Their cells are smaller capacity than A123's 26650s that I've been testing against but they look like they are cheaper Wh/$. I wouldn't mind soldering 3 or 4 of these cells together to replace and A123 unit. I have ordered 80 of their cells to play with (you have to order 50 to get a volume discount). If anyone is interested, I will post my testing results to the appropriate forum.

Cheers,
--Adam

Hi Adam, and welcome.

Many of us have used their 2A single cell chargers for awhile now (I have 16 of them...), but I wasn't aware they were now selling new, higher power LiFePO4 cells. At 10C, they would appear to similar in performance to the PSI/BMI/LiFeBatt cells, but about 1/10th the capacity. I think you will find that the a123 cells are still stronger. 10C is plenty, for most applications, though.

Anyway, do post a new thread when you are ready to do some tests.

-- Gary

 

[Malcolm]

Being a Brit I just wanted to stake my place in the queue

... and say a big thank you to Gary and Richard and Bob and everyone else for all the work you've put into this

... and of course ask a dumb question

I'm planning to build a 108V 60Ah LifeBatt/PSI pack for a small car. The pack will be split into three 36V 60 Ah sub-packs, each charged by its own 36V 30A lead acid charger. I'll be fitting a heat sink to the control FET to keep it cool during charging. What I'd like to know is whether it would be better to use two boards daisy-chained together with just one control circuit activated, or to use three boards with an active control circuit for each sub-pack? Hope this makes sense.

 

[ejonesss]

are the cells charged independently and not in series? or is groups of 4 at a time in series? because if each cell has it's own charging circuit then the charger supply can be the 5 or 12 output from a computer supply or even from a car battery.

i thought the new design includes a voltage and current regulator so the raw power from a 12 volt transformer converted to dc (good enough to run a motor) would work.

do you have some pictures of the final unit that ships?

will the new design allow the use of a standard car battery charger? or even the 5 or 12 volt output from a computer power supply?

Yes, just about any source that is current limited and has enough voltage will work.
A car battery charger might be a bit low for 4 cells and certainly a 12v output from a PSU would be too low. You can tweak the voltage up on these in many cases.

 

[dnmun]

mr jonesing, if you will go back and read through the thread from the start, and even over on to the thread that xyster started last year where the idea originally gelled, you would have a complete view of what you are asking about. i doubt if there are enuff answers people will be able to deliver to get you to the point where you can do it yourself, but that is what i would recommend. it will keep you from asking embarrasing questions, and reduce the anger of those who feel your recent behavior along with others demanding a rip off, has been very disrespectful of gary specifically and without any appreciation of the value of this open source design. i now understand why bob spends little time here, and i worry about who will be next to disappear like xyster did.

i am sure there are others who feel the same way, but i hate to hijack the board with what really are personal comments, but i would like to see the thread lead away from gary having to answer every little question, he is taking the orders at thunderpower too. so go there for ordering the boards and parts list, and maybe a new thread or 2 could be started for group buys of the parts and splitting the costs that way, and also a thread for people who build the boards up.

again i want to thank gary, richard, and bob for doing this, and i wonder if we could perhaps talk gary into selling off the first pass boards with the leads reversed for the transistor as momentos. if you guys would initial them, we could do a dutch auction and let people who want, bid on them and save them as knick knacks for the fireplace mantle. thanks so much you guys, dm

 

[commanda]

Gary, Richard,

Please don't forget to post your final circuit after the current round of testing is complete.

Amanda

 

[methods]

. . . mr jonesing, if you will go back and read through the thread from the start . . .

Good advice.
If that is too much reading then just go back to page 17 and you should get what you need.

I start the bidding on the fubar board at $50.
I can solder bent leads, or even mount it on the other side of the board :wink:

-methods

 

[ejonesss]

thanks to everyone who worked hard to develop the new bms board.

i will just wait until the new board design comes out .

and hopefully it will work for me too.

 

[PJD]

Toyguru wrote:

I'll be using it for 12 TS cells, so I'll have to experiment to find the right components to get the 4.25V cutoff.

Just so you know, there seems to be a consensus that TS cells should not be regularly charged to 4.25 volts per cell. Optimum cycle life is obtained when charging is limited to about 3.7 volts per cell. The capacity difference between these two voltages is only a few percent.

 

[methods]

I know that I am way late in this thread. . .

Is there any way to lay the pads down on the for the TC54 that would make it easier to fit the more common SMT parts in addition to the dip?
I am betting that I will be able to solder the smt's anyway but a small pad would make it easier. I have a microscope and infinite patience. . .

After long hard thought I decided that I must go with 3.0V LVC to protect me from the scenario where I leave the bike on and forget about it for a weekend.
If I were to go with a 2.7V LVC Murphy would without doubt come to tell me about his law.

-methods

 

[ejonesss]

correct me if i am wrong but what you are describing is.

from the http://www.tppacks.com/proddetail.asp?prod=EBKE-A123-4s4p-KIT page

The kit includes the following: 4-channel LVC board, 5-pin male balancer/charger plug, three inter-connection G10 plates with nickel-plated battry springs and copper straps which are used to connect the four cells in each block in parallel, one G10 bottom plate with attached threaded rod, and a 14" length of black 5"-wide PVC heat shrink tubing. Also included is all required hardware (nuts, lock washers, etc.).
_______________________________________________________________________________
Carhartt Work Pants budapest property

 

[GGoodrum]

I know that I am way late in this thread. . .

Is there any way to lay the pads down on the for the TC54 that would make it easier to fit the more common SMT parts in addition to the dip?
I am betting that I will be able to solder the smt's anyway but a small pad would make it easier. I have a microscope and infinite patience. . .

After long hard thought I decided that I must go with 3.0V LVC to protect me from the scenario where I leave the bike on and forget about it for a weekend.
If I were to go with a 2.7V LVC Murphy would without doubt come to tell me about his law.

-methods

Even using the 3.0V TC54 parts won't help if you simply leave the controller on. By using the ebrake line, what happens is if a TC54 trips, it is basically just killing the throttle. In order to prevent true bonehead moves, like leaving a controller on, which I myself may, or may not have done ( ), we'd have to add Randomly's active cutoff feature, and add a couple more 4110s to control the negative pack lead. Actually, maybe what I will do is a small board, just for this function, unless it is something that everyone can't live without (true active cutoff...).

 

[ejonesss]

the active cut off is a good idea.

maybe possibly a latching relay so that when you turn on the bms you have to push a button (like the on off setup used in the http://neodymics.com/Images/V24ProtoSwitch070818E.pdf does).

unfortunately 2 things i can think of.

1. i dont know if it is good for a controller to turn it on via the power wires instead of the power switch (the surge).

2. voids compatibility with regen braking ( of course i am not sure if that would work through a bms)

I know that I am way late in this thread. . .

Is there any way to lay the pads down on the for the TC54 that would make it easier to fit the more common SMT parts in addition to the dip?
I am betting that I will be able to solder the smt's anyway but a small pad would make it easier. I have a microscope and infinite patience. . .

After long hard thought I decided that I must go with 3.0V LVC to protect me from the scenario where I leave the bike on and forget about it for a weekend.
If I were to go with a 2.7V LVC Murphy would without doubt come to tell me about his law.

-methods

Even using the 3.0V TC54 parts won't help if you simply leave the controller on. By using the ebrake line, what happens is if a TC54 trips, it is basically just killing the throttle. In order to prevent true bonehead moves, like leaving a controller on, which I myself may, or may not have done ( ), we'd have to add Randomly's active cutoff feature, and add a couple more 4110s to control the negative pack lead. Actually, maybe what I will do is a small board, just for this function, unless it is something that everyone can't live without (true active cutoff...).

 

[GGoodrum]

correct me if i am wrong but what you are describing is.

Huh?? That kit is something I did for building a123 packs. It has nothing to do with the BMS. I now have a new construction method I'm trying, which will use the same G10 plates and nickel-plated springs, but instead of stacking them end-to-end, I'm going to place the 4-cell blocks side-by-side. Each block basically adds "mini-studs" for each cell, which makes wiring and configuring various pack setups much easier. Here's what they look like:

I'm using this building block approach to build an 8s6p "booster" pack for my Townie.

 

[methods]

... In order to prevent true bonehead moves, like leaving a controller on. . .

Speaking of bonehead, i was being a bonehead not thinking of that.
It would be somewhat easy to leverage your active low output to trigger another circuit.
Currently I use a TP2200 extreme 4S pack to power my controller (can anyone say overkill?). This same voltage can hold my contactor closed. All I have to do is put together a simple circuit to open the low power relay given an active low input. This would open the main contactor and save my batts from Murphy.

Sounds like it would be worth my while.
Fairly unique to me though as not many use a Kelly

-methods


EDIT: I would steer clear of trying to implement something like this on the BMS. Too many people run too many different power levels. Trying to design something that is good for the guy running 36V20A and the guy running 120V190A is too much work. Not to mention that you would have to deal with the bridge resistor (to save the contacts). Best left to the end user to develop.

 

[velias]

I think the BMS circuit should be left as is for now, its simple and provides all the basic functions everyone needs. If people need active low voltage cut off or surface mount devices then that should be something in the long term. I think having this board will be great and I can't wait to buy it as it is now. I've had to go back to using lead acid after my cyclone made-in-taiwan overly-complicated BMS on my lifepo pack blew up for no reason a couple of weeks ago.

 

[Patriot]

I have a really dumb question.

If I build a battery pack using 32 cells that are 3.2v/10ah each, in order to make a 48v/20ah pack, would using 2 BMS's be the best way to go?

I intend to build the pack like this. 16 cells in series to create 2-48v/10ah packs. Then, wire the two in parallel to create one 48v/20ah pack, right?

So, wouldn't having 2 BMS's be the best way to wire up, and balance each individual cell?

I've never quite understood how they build a 32 cell pack using only a single 16-cell BMS. Unless the wires are spliced to serve two cells in parallel for each 10ah pack somehow?

Someone please enlighten me out of my ignorance.

 

[rf]

I have a really dumb question.

If I build a battery pack using 32 cells that are 3.2v/10ah each, in order to make a 48v/20ah pack, would using 2 BMS's be the best way to go?

I intend to build the pack like this. 16 cells in series to create 2-48v/10ah packs. Then, wire the two in parallel to create one 48v/20ah pack, right?

So, wouldn't having 2 BMS's be the best way to rie up, and balance each individual cell?

I've never quite understood how they build a 32 cell pack using only a single 16-cell BMS. Unless the wires are spliced to serve two cells in parallel for each 10ah pack somehow?

Someone please enlighten me out of my ignorance.

There are so many competing considerations.

Tying cells together in parallel pairs or small groups makes them into larger more robust cells -- that require less hardware to control and charge. Conversely, separate whole packs gives you some redundancy. If one pack fails you can get home with the other. Two separate packs could possibly be easier/cheaper to charge too, depending on what's available.

Richard

 

[Patriot]

^^^ OK,

You've thoroughly confused me with that one.

Not only that, but I'm still wondering what the heck I should do.

I figure, if I have two large 48v/10ah packs in parallel, then they will basically discharge pretty evenly to give me the current I need. Also, like you said, if one big pack fails, the other pack in parallel continues on.

However, how do I wire in BMS's for cell balancing for charging/discharging?

 

[GGoodrum]

Good question. With a123 cells, I always parallel them first, and then put the parallel blocks in series. The reason is that I've had several years experience with these, going back to beating the crap out of packs in RC helicopters, and I've never ever had one cell fail on its own. In fact, until I started using them on ebikes, and learned how to over discharge them, I had never managed to even damage a cell, much less kill one. In an RC helicopter, it won't stay in the air without the voltage staying up over about 3.0V per cell. When a pack would dump, however, you had about 10-15 seconds to decide where you were going to land.

Anyway, because of my experiences with these, I have no worries that a single cell will fail, and will pull down others that are paralleled with it. As long as each block has LVC protection, I've never had a problem. With my LiFeBatt packs (2 x 12s1p 36V/10Ah...), they are in a "single-p" configuration, so this doesn't apply, but I've never had a problem with a cell going bad on its own. I've also done two 16-cell 48V/10Ah packs on bikes for a friend, and for my brother-in-law, and both of those are still going strong. If I was going to make a 20Ah/2p configuration using these PSI-based cells, I would not hesitate to do them in parallel first, and then use a single BMS board.

With the 10Ah Headway cells, I think the jury is still out. I've heard of some early failures, which would cause me to do separate packs, isolated in parallel with a Schottky, but I've also heard they might have a newer version that seems closer to the PSI/BMI/LiFeBatt in terms of performance.

For any cells that are based on the use of bags, I would definitely do separate packs. Too many failures with these.

-- Gary