Group Projects
Members' Projects
Battery management module
davedave — Wed, 03/02/2010 - 10:41am
The idea is this: create a battery managent board for devices with many expensive batteries. Eg: electric vehicles.
A typical electric vehicle will have between 8 and 50 batteries worth as much as the rest of the vehicle.
So the concept is a microcontroller to run each board (one per battery) and solder bridge jumpers to select battery chemistry, running off the battery it is attatched to. These batteries are 3.4 to 12v nominally so a low power chip should do as long as it is protected from higher volages and noise.
Communication on one bus with no controller and an interface board to display results.
I'm open to suggestions on this.
- Login or register to post comments
- Printer-friendly version
Search
Member Tweets
evildeece
1 hour 49 min ago
—
Spray on clothing! http://t.co/djQzAie7 #mhv
|
MakeHackVoid
1 hour 56 min ago
—
Space is closed (was open 7 hours)
|
evildeece
2 hours 13 min ago
—
3D printed RC plane http://t.co/8LCfuR3C #mhv
|
evildeece
2 hours 16 min ago
—
21 more MHVBoard #arduino clones packed up and ready to go to local school & uni #mhv
|
MakeHackVoid
9 hours 2 min ago
—
Space is open until 22:00 (estimate)
|
projectgus
12 hours 22 min ago
—
Club Mate induced madness MT @dev_dsp @MakeHackVoid It tastes like ground! and like tea! and like ground! Pretty much the best thing ever.
|
|
dev_dsp
13 hours 5 min ago
—
@MakeHackVoid It tastes like ground! and like tea! and like ground! It's pretty much the best thing ever.
|
MakeHackVoid
13 hours 6 min ago
—
Also there are rumours of @dev_dsp's "weird European hacker juice"
|
MakeHackVoid
13 hours 8 min ago
—
Rumour has it that tonight's Maker Meetup will include an electric motorcycle, RepRap 3d printer, and SMD oven! Come! http://t.co/vmahKECz
|
evildeece
17 hours 38 min ago
—
OK Go's new music vid, using a car to play over 1000 instruments: http://t.co/Qca3G1Eu #mhv
|
- 1 of 12
- ››




Some links
deece — Thu, 04/02/2010 - 8:44amTheres a list of some open source BMS designs here: http://liionbms.com/php/bms_options.php
I was considering the same a couple of years ago when I got into e-biking, but I never had time to do it and instead just bought a commercial board from HK.
I would ditch the jumpers and use a serial port for configurations, less IO pins consumed so you can use a cheaper controller. Controllers generally have multiple ADCs, so you should be able to read multiple cells with some resistor divider networks (and some way to turn the whole thing off to reduce current draw).
reader comments
peeknpoke — Tue, 14/09/2010 - 11:35amJust some comments:
It is not clear what sort of battery voltage or capacity you are aiming for. Considering the task the battery monitor has to carry out, I cannot understand why you are shying away from the use of a microcontroller. Most microcontroller A/D read to a resolution of 10 bits, the last bit being a bit rubbery, so lets say 9 bits. You need to read your individual cells to 0.1 volts, so using one controller for up to 50 volts worth of batteries is about it.
A road drivable car is a whole different ballgame to an electric bicycle. There are so many economies to be made by sticking to already commercially available equipment that it becomes difficult to design something that is perfect for all applications. When you get to the design and testing phase, considering a broad range of applications becomes too expensive and time consuming to make it worth while.
For bicycle / recumbant / scooter arrangements, there is so much commercially available in the 48 volt range that it would be hard to look elsewhere. This market is pretty much dominated by sealed lead acid batteries, but other technologies are also quite viable. The various battery technologies are so different, that it is not practical to design for all. For 48 volts, a single monitor/controller is viable, and a communication bus is hardly necessary.
When you get up into the road drivable rhelm, the industrial motors, gearboxes, and speed controllers available for 600 volt (don't quote me on this, but I think its 600) are so cheap and ready made that it would be ridiculous to go it alone. You get locked in pretty quickly to 150 off lithium phospahate cells. If you were designing in clusters of around 48 volts, you would be looking at ten or more monitor/controllers that would be working at voltages too far apart to avoid an opto-isolated common bus. Once you have a number of controllers, and a common bus, it is pretty hard to get around the need for a separate overall controller. You could do it with alot of programming and persistence, but why? You really need one overall system control/display anyway.
For charging Lithium batteries at a high current, charge balancing is required. To charge all cells to a set voltage, then bypass any excess charging current means that when some cells reach full charge before others, the bypass load and transistor will have to handle the full current and dissipate alot of heat. This makes for a complex and difficult build. If you use a common controller with a bi-directional communication bus, you can monitor each battery voltage as it charges, and compare it to the voltage on all other batteries in the string. By keeping the cells balanced as they charge, the bypass transistor and resistor have to handle nowhere near the same current and subsequent power dissipation.
Peeknpoke.