Maxing out the USB lithium charger

In its previous incarnation a TL431 shunt regulator was used to ensure that the lithium cell could not be charged to more than 4.2V. That is to say that when the cell has reached its maximum voltage, the TL431 must shunt all of the current flowing through the limiting resistor. On the previous example it was calculated that an 8R2 resistor would limit the current to within what the TL431 can safely shunt.
The new approach adds a PNP output transistor to the TL431 in much the same manner as a Sziklai pair. Now the maximum shunt current is the 100mA that the TL431 can handle multiplied by the gain of the PNP transistor. As the TL431 maximum current rating is no longer the limiting factor, the only constaint is the maximum current that a correctly specified USB charger can supply (500mA) when the battery is at its lowest voltage, the value came out as 4R5 - with 4R7 being the most suitable preferred value. The transistor used in the prototype is a 2SB857, this was the junk-box part that came readily to hand while rounding up the parts - there's nothing particularly special about this device, its rated -50V, -4A and a hfe min of 60, just about any TO220 PNP with similar ratings will do the job - the one in the prototype was not fitted with any heatsink - with no load (maximum shunt current) it did get moderately warm, but not in the least uncomfortable.
As the PNP transistor now does most of the heavy lifting and its collector is returned to the negative rail, a current sensing resistor can be inserted in the collector lead to switch a second transistor which operates the cell full indicator LED. This would have been more complicated with only the TL431 as inserting the current sensing resistor in the anode lead would create a variable offset voltage working against its reference point. The current sensing resistor in the collector lead is included in the feedback loop and has very little effect on regulation.
Discussion (1 comment)
Detlef 11 years ago
Ian, I do like the idea of a simple charger for LiIon's ripped from Laptop Battery Packs. They can turn out to be quite useful if it comes to supply a small portable selfmade device. If you include the cell into the device's box you sure don't want the charger to be more complex, thus expensive, than the device itself. I use them a lot for small µC operated devices where battery lifetime shall be maximised.
Actually, when reading your description,I got the impression that this is nothing else than a so called Balancer circuit used to charge multi cell packages of LiIon's to avoid over charge one cell while it's neighbour still needs to be charged. While beeing necessary in multi cell arragements, for a single cell it's kind of waste of energy in my opinion.
May I suggest to cut off the charger from the cell by a transistor (possibly a MOSFET ?) when it has reached it's final charging voltage, rather than heating environment instead?
I'll be glad if you take my suggestion into account.
Any help I can provide?
Best regards.
Ian.Field 11 years ago