EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: paulca on December 06, 2017, 12:05:08 pm
-
I want to design a low voltage cut off for my DC Load. I want it to accept a LiPo balance port and switch the gate voltage to the Mosfet off if a single cell drops below a configured voltage.
I will hopefully get a schematic up and running this evening, I'm having issues getting LTSpice running but I have a solution for that now.
The Load is based on Peter Oakes design, so my plan is to switch the control signal off (somehow) when my new circuit detects an low voltage on a single cell.
I just want to check with you guys the sanity of what I am planning on doing.
1. If I use a voltage source or reference and a divider with a potentiometer, then a voltmeter to show the voltage that circuit is producing, I can set the low voltage reference voltage via the pot with indication.
2. That voltage then feeds into 3 comparators (I happen to have a few around) which compare the (in my case) 3 cell voltages. If a single voltage drops below the reference voltage that comparator will switch on and energise the cut off.
3. The cut off could be a relay on the high side load conductor, or just a transistor to ground the gate wire or even the 500mV control signal. The gate wire currently comes out of the current sense op amp through a 10K resistor. If I pull that to ground before the 10K with a 1K resistor that should shut the gate off, shouldn't it? The downside is the current sense opamp will run to it's upper rail trying to open the gate, but I'm not sure if I care.
4. The cut off would be latched with 2 NOR gates so once triggered a manual reset button will need to be pressed to release the cut off. This would prevent the load bouncing as the cells recover above the cut off when the load comes off.
An idea I am toying around in my head is to use differential amps to output the difference between the reference voltage and the cell voltage. Then somehow use the minimum difference output to proportionally lower the load control signal. This would be ideal. If I drain a big LiPo at 5A a cell might register 3.0V and the cut off trigger. This cell could then recover to 3.2V or higher. If I gradually reduce the discharge load as I approach 3.0V it would give more accurate results when finished. Probably a bit more complicated to do though.
All of this of course would prevent any cell in the pack going below 3V, but it would not prevent the pack unbalancing which can cause damage under load. A "unbalanced" disconnect could be done by comparing the outputs of these differential amps and if there is more than X voltage difference, engage the cut off or wind back the load.
Currently I acheive all of this just by connecting a LiPo monitor to the balance port while discharging. But while it will start beeping loudly at me, it has no way to shut the load off and at 5A a cell can drop pretty suddenly once it gets to 3.0V. If I am out of the room, it could drain it way to far.
-
Googling around suggests that most people resort to a micro controller for this kind of thing. The cut off for a given voltage is easy enough, but controlling the discharge rate based on how balanced the cells is or how close a cell is to cut off seems to be more of an arduino project than complex discrete components and opamps.