Rooting around in the fascinating stuff at the bottom of a draw labelled 'Engineering - Junk Miscellaneous'. Delving amongst the delightful...
Active cell balancing, an automotive conundrum
Cell balancing in lithium ion car batteries seems to tough nut to crack without excessive complexity.
The issue is that Li-ion cells are terribly intolerant of over-charging.
0.1V can be the difference between 600 cycle lifetime and 15 cycle lifetime. Another 0.2V and the thing can burst into flames. This said, I expect cars will use the Fe-type Liion batteries which are a bit tougher than the Co or Mn types.
In a car battery, up to 100 of them need to be charged in series.
Even monitoring them is tricky and several companies including Maxim and Linear Tech have come up with 12 cell monitors that have clever ways of passing voltage measurement data down the 300V-long chain to ground so that the electronic control unit can make use of it all.
Anyway, once the battery is nearly charged, some cells will be full and some will still need some more energy.
The ones that are full must not be provided with any more energy, so the minimal solution is to stop charging as soon as the first cell in the battery is full.
The minimal solution that will result in a full and balanced battery is to bypass each cell with a transistor and a resistor once it is full. The resistor has to pass enough current to pull the cell down to a voltage where it effectively stops charging.
Not a bad approach, but lots of dissipation from by-passing all the full cells when the battery nearly full. And heat seriously shortens the life of Liion cells.
A neater solution would be to active balancing – passing the bypassed current to a cell that still needs charge – but how to do it?
Accuracy warning: From now on ‘Alice’ is close to guessing…
A flying capacitor switched by a pair of transistors for each cell could pass current up and down – but the switch voltage limits how far the charge can be sent up the chain.
Also, a capacitor charged by a full cell is at almost exactly the voltage needed by the undercharged cell – it could be higher or lower depending on the cells – so there is no guarantee charge will not just be stolen from the nearly charged cell instead of being delivered to it.
This last problem will also occur if the charge is passed up a ladder of capacitors, although this would solve the voltage limit issue.
Inductive active schemes, it seems to me, can handle higher and lower voltages, but what a lot of switches would be required – I think 4/cell.
Or maybe spare energy be fly-backed (flown back?) to one capacitor per every 12 (say) cells, and then redirected to a needy cell by more switches and inductors.
And then how would it be passed between lots of 12.
All in all, a tough nut indeed.
I have seen one solution, but for the life of me I cannot find it again.
Linear Technology, a company whose design skills I have the utmost respect for, is to reveal details of its LTC3300 battery charge balancer in the next two or three months.
I can’t wait to find out how it is done.
I just found this: A Review of Cell Equalization Methods for Lithium Ion and Lithium Polymer Battery Systems.
These techniques do seem to need wall matched cells as they assume a charged cell has a higher terminal voltage than a discharged cell, and there are indeed a lot of switches.
Let’s hope Linear Tech has something really novel.
If you respond below, our spam-blocking system may throw away your comment.
Better email email@example.com with ‘Balancing’ in the title.
No email addresses are collected for marketing (or any other) purposes from responses to this blog. I will keep it that way for as long as possible.