tangent
Top Mall-Fi poster. The T in META42.
Formerly with Tangentsoft Parts Store
- Joined
- Sep 27, 2001
- Posts
- 5,969
- Likes
- 58
I've been looking into making a circuit you can add to a PPA battery board type NiMH fast-charger circuit to prevent the battery from being over-discharged. Here's what I've come up with: Schematic, PDF, 20KB
The first comparator circuit compares the divided-down battery voltage to a 2.5V reference. (The battery voltage is applied between B+ and GND.) Since the bottom leg of the sense divider is fixed at 100K for practical reasons, you pick R10 to give the trip point you want. It's a little more complicated than that, because of the hysteresis provided by R12, but we can ignore that for now. If you want the battery to be cut off from the load at 8V, you'd set R10 to about 220K. When the battery voltage is higher than that, the comparator's output is pulled high by R13, which turns on the Q1+Q2 Darlington, allowing current to pass from the battery.
The second comparator circuit is just a way to force the pass transistor on when the wall supply is connected to the fast-charger. (REG symbolizes the input to the regulator in the PPA battery board circuit, a convenient place to test for the presence of wall voltage.)
Anyone see a problem with this circuit?
The only doubt I have is how efficient it is. I can see two possibilities: either all of the current used to drive the Darlington makes its way to SB, or some of it is also wasted down the comparator outputs. What governs how much current takes each path?
Assuming that all of the drive current makes its way to SB, then the biggest current waster is the comparator, which can be a low-power part; we don't need speed here. All the other currents are down in the tens of microamps. High efficiency is good, because the circuit will still be draining battery power after it cuts off the load. You don't want the monitor circuit to itself be the cause of over-discharge! If you have a 750mAh battery pack, and you cut off the power when 99% of the safely-usable charge is taken, an 0.5mA drain means you have 15 hours to get around to turning off the power before the last 1% is used up. This assumes you're using a low-power comparator, in order to meet that 0.5mA target.
Getting back to the hysteresis: it needs to be there because the sense voltage will fall very slowly while the battery discharges. It will even bounce a tiny bit as the load current varies. (Maybe I should put a stabilizing cap in parallel with R10.) I haven't rigorously run the numbers, but I believe there's will be something on the order of a volt of hysteresis with typical R10 values and the other given resistor values. That should be enough to keep the power forced off once the first comparator circuit's output falls.
The first comparator circuit compares the divided-down battery voltage to a 2.5V reference. (The battery voltage is applied between B+ and GND.) Since the bottom leg of the sense divider is fixed at 100K for practical reasons, you pick R10 to give the trip point you want. It's a little more complicated than that, because of the hysteresis provided by R12, but we can ignore that for now. If you want the battery to be cut off from the load at 8V, you'd set R10 to about 220K. When the battery voltage is higher than that, the comparator's output is pulled high by R13, which turns on the Q1+Q2 Darlington, allowing current to pass from the battery.
The second comparator circuit is just a way to force the pass transistor on when the wall supply is connected to the fast-charger. (REG symbolizes the input to the regulator in the PPA battery board circuit, a convenient place to test for the presence of wall voltage.)
Anyone see a problem with this circuit?
The only doubt I have is how efficient it is. I can see two possibilities: either all of the current used to drive the Darlington makes its way to SB, or some of it is also wasted down the comparator outputs. What governs how much current takes each path?
Assuming that all of the drive current makes its way to SB, then the biggest current waster is the comparator, which can be a low-power part; we don't need speed here. All the other currents are down in the tens of microamps. High efficiency is good, because the circuit will still be draining battery power after it cuts off the load. You don't want the monitor circuit to itself be the cause of over-discharge! If you have a 750mAh battery pack, and you cut off the power when 99% of the safely-usable charge is taken, an 0.5mA drain means you have 15 hours to get around to turning off the power before the last 1% is used up. This assumes you're using a low-power comparator, in order to meet that 0.5mA target.
Getting back to the hysteresis: it needs to be there because the sense voltage will fall very slowly while the battery discharges. It will even bounce a tiny bit as the load current varies. (Maybe I should put a stabilizing cap in parallel with R10.) I haven't rigorously run the numbers, but I believe there's will be something on the order of a volt of hysteresis with typical R10 values and the other given resistor values. That should be enough to keep the power forced off once the first comparator circuit's output falls.
