The cells are charged in series and viewed to the charger as one large cell. The charger has no way of equalizing the SOC (state of charge) for each individual cell. I've designed a simple cell balancer that takes advantage of the nature of lithium during charging. LiFePo4 cells are 3.2v nominal and like other lithium technology stay right around this nominal voltage until the end of the charging cycle when they climb in voltage very quickly. The cell balancer is designed to start shunting energy at 3.6 volts that would otherwise be going into the cell. The balancer shunts 1.5 amps at 3.6v and goes up from there as the voltage increases. Let's say you are charging your pack at 10 amps, then each cell is receiving 10 amps. If a cell reaches 3.6v 1.5 amps of this energy is shunted and the cell is now only receiving 8.5 amps. The remaining cells which are below 3.6v still receive the full 10 amps which allows them to catch up over time and perhaps multiple charges. The lower the overall charging rate then the higher percentage of the energy can be shunted allowing for a complete cell balance in one charge. For example if you can limit your charge to 1.5 - 2.0 amps you can simply leave the charger on and wait for all balancers to indicate they are shunting and therefore the pack is now balanced.
Here is a picture of a completed cell balancer. They simply need to be connected to each cell only. There is no master unit or mess of wires to tie them all together. I named them Volt Blocher. My my last name Blocher (pronounced Blocker) and the fact they will keep the voltages down by shunting away the energy.
Here is a picture of them installed in the car. I needed to make 45 units for my setup. My wife helped out on the assembly line and really saved me some time.
Here is a picture of the units in action. In the closer four units you can see that three of the four units are shunting energy indicated by the red LED. At this point the picture was taken only about 6 or so of the 45 units were in this state meaning the remaining cells needed to catch up.
This looks much cooler in person but thought I'd add a shot with most of the cells now shunting. Lots of red and green lights. After about three hours of charging I had just four cells remaining that weren't balanced. I'm going to continue to charge at a slow rate and see how long it takes to bring them level with the others.
There will soon be two others using the Volt Blocher units. Dimitri and Jim have ordered their DIY kits and will hopefully have them up and running soon.
For more information and ordering please visit VoltBlocher.com
6 comments:
Looks like a nice piece of work, as usual, Brian. The way you describe your system it seems to protect against over voltage, but you might want to take note of this recent discussion:
http://www.batteryvehiclesociety.org.uk/forums/viewtopic.php?t=1825&postdays=0&postorder=asc&start=0
Where LiFePo4s are suspected of overheating and burning (!), possibly as a result of a BMS/charge management problem not detecting an individual failure in a cell. If I understand your system it could still deliver full charge to a cell/pack when there is a failure? Anyway you might want to take a look.
Brian, two questions: 1) do you have any low voltage protection at cell level in your system (if I remember you have a Link 10 for the pack) and 2) if you use regen, do you have any over voltage protection when driving? I just ask because at present on my project I have charge balancing, but no low/high voltage protection for the LiFePo cells when driving and wondered if you think that's an issue. Simon
Wow thanks for the link that is amazing that they caught fire, if indeed they did.
I'm not using a link 10. I'm using the BMS system from TS which is not a full manaagement system but partial. The BMS and charger are fully itegrated.
The BMS will:
* Monitors for high voltage and reduce the amp output on the charger until the voltage drops below a safe level.
* Monitors for high temperature and take the same action as high temp when charging.
* Monitors for low voltage and will not charge at all until the voltage is raised.
During non charging conditions the BMS will beep rather loudly if you reach low/high voltage, high temperature and high current but it doesn't have an active role (i.e. cut back on the motor controller). You just have to listen and make the corrections yourself based on the problem which is shown nicely on a 7" LCD touchscreen.
The new cell balancers that I've designed just work hand-in-hand with this BMS. Since I already have high voltage protection these balancers don't protect completely from high voltage, they just delay it to give the lower cells time to catch up in charge. The higher rates you charge at, the less time you'll be giving the balancers to work. However, once balanced for the first time I doubt I'll ever need to charge it slowly again.
I don't have regen and if I did I wouldn't have full protection from high voltage while driving. What would happen is the cell balancers would kick in and start burning off energy as needed. They would not be able to keep with regen but would help.
The real question is would you ever be in a scenario where you are fully charged (just pulled the car off the charger so at 3.6v or so) and then start heading down a hill for awhile? If not I seriously doubt you'll ever see enough power regenerated to cause a problem.
Nice work!
You should also check out http://www.lithiumbalance.com/technology.html
>>>The cell balancer is designed to start shunting energy at 3.6 volts that would otherwise be going into the cell. The balancer shunts 1.5 amps at 3.6v and goes up from there as the voltage increases<<<
Goes up from there? What is the max amps it can shunt? Are they basically shunting everything at 3.8 Volts? (A good place to stop charging)
Does it do anything when you are not charging?
As the voltage increases the amount of energy shunted also increases to try and keep the cell voltage down.
The following are approximate numbers:
3.6v - 1.5 amps
3.8v - 1.8 amps
4.0v - 2.0+ (Only tested this once just to see the amps continue to rise, I like to keep mine at 3.8 max also)
So these numbers are what the balancers will shunt. How many amps are still going into the cells is based on your charger.
For example if your cells are at 3.6 volts but you are charging at 5 amps then 3.5 amps is still making it into the cells. If the cells are at 3.8 volts then only 3.2 amps are making it into the cells.
Most chargers have options for setting voltage and/or amperage to some degree. You can decrease your final charge amps to just slightly above the balancers shunting ability. This allows for a very slow final charge stage in which the cells will be able to balance nicely. Using higher amperage will give less time for balancing but not waste as much energy in shunting. You can adjust this as needed to meant your requirements for balancing. Typically you'll do a nice slow balance up front and then a much shorter balance cycle since these cells stay fairly well balanced. You can then always come back later when you want and change it back to do a slow and more complete balance.
Post a Comment