Low voltage cut-off

[Brown]

I am using a 12v 5000mAh NiMH battery to provide back-up power to a circuit. The circuit draws about 500mA. To prevent damage to the battery, I need to disconnect the circuit when the voltage gets to the 10-11 volt range. I have seen a device that would work but it is relatively expensive when compared to the circuit I am powering. The Sun Saver Solar controller from MorningStar would work because it disconnects the load at 11.5v. Is there a simple, low power solution that I can build myself?

 

[justDIY]

nimh have a rather flat discharge curve, making it hard to detect their actual state of charge.

A simple solution could involve a current sensing resistor (something small like 100mOhm or 10mOhm to make the math easy), an op-amp and a microcontroller.

The op-amp takes the voltage (which will be positive or negative depending on charge/discharge) and converts to a positive only voltage the microcontroller can understand.

The uC is programmed to count how much time the battery spends charging versus how long it spends discharging (and at what rate). or simpler yet, the uC expects you'll take care of charging and will assume the battery starts each day at 5000mAh. if you worry about discharging only, you can dispense with the external op-amp and just use a diode to block reverse voltage from reaching the pic (be sure to account for the Vf of the diode).

When the uC counts that 4500mAh (or whatever) has been discharged, it can disconnect the load, regardless of the terminal voltage on the battery.

Maxim has some battery supervisor chips that do all this for you - they're fairly cheap, but not easy to work with.

 

[stevez]

justDIY's suggestion will likely provide the proper protection. An alternative that might provide some benefit would be to use a comparator (LM339, LM311, among many others). The inputs to the comparator would be your battery voltage and a reference voltage. Google on voltage comparator - lots of explanations. Note also that op-amps can be configured as comparators.

I've no experience with this but I'd consider configuring the comparator so that you have some hysteresis - unless you can test your circuit to be sure that you don't need to add more. A brief explanation - the output of a comparator is either on or off. A comparator has two inputs, A and B. If the voltage at A is higher than B the output is high. If A is reduced so that it is less than B the output of the comparator will go low the instant that A is less than B. The problem - when A is very close to B a little noise can confuse the comparator. Hysteresis essentially adds some difference so that the output won't switch back and forth so easily. Google on comparator hysteresis for more. I think the National Semiconductor datasheets do show how to add hysteresis.

 

[Hero999]

You mean a column counter?
Can't you get ICs for that?

I've noticed that NiMH cells seem to drop their voltage very quickly when discharge or is that only after the damage has already been done? I would have thought that a quick roll-off would be easy to sense using a traditional comparator.

 

[stevez]

If my memory is correct, the drop off occurs quickly at the end of charge. Adding to that is that temperature affects battery voltages so finding a good trip point could be challenging if the circuit is to be kept simple.

 

[justDIY]

a dedicated coulomb counting chip would be the ideal solution but most of them are tiny little buggers designed for inclusion inside a laptop battery pack

looking at these graphs, a reoccurring theme is clearly evident
https://images.google.com/images?q=nimh+discharge+curve

the 'damage' is done by the time the cell voltage drops significantly.

another thought that might work, provided you leave enough margin for error, and you start with a full charge every time... a simple timer. if you have a more or less constant load on the batteries, you can time how long to discharge, add some fudge factor and then use that time for future discharge cycles.

 

[ericgibbs]

hi brown

Have a look at the MAX8211 ic.

Designed for battery low volt detection.

 

[audioguru]

The moment the load is turned off by the circuit then the battery voltage jumps up, so have the circuit latch off, and have a reset button or have a power-up reset circuit for the circuit to work the next time.

 

[Brown]

Thanks for all the ideas. Not as simple as I thought! I 'll let you know what I try. Please forward any other ideas you think of. Regards, Brown

 

[justDIY]

how simple do you need it? a single chip solution (like those from Maxim) is pretty simple, as long as your soldering skills are up to par.

your choice of battery is making things "difficult." if you can switch to lead-acid, it has a nice predictable discharge curve since measuring cell voltage is a very accurate way to determine SOC (state of charge).

 

[pubudusolar]

Dear friend I need solution like this problem
Thank
Pubudu
I am using a 12v 5000mAh NiMH battery to provide back-up power to a circuit. The circuit draws about 500mA. To prevent damage to the battery, I need to disconnect the circuit when the voltage gets to the 10-11 volt range. I have seen a device that would work but it is relatively expensive when compared to the circuit I am powering. The Sun Saver Solar controller from MorningStar would work because it disconnects the load at 11.5v. Is there a simple, low power solution that I can build myself?

 

[rjvh]

Dear friend I need solution like this problem
Thank
Pubudu
I am using a 12v 5000mAh NiMH battery to provide back-up power to a circuit. The circuit draws about 500mA. To prevent damage to the battery, I need to disconnect the circuit when the voltage gets to the 10-11 volt range. I have seen a device that would work but it is relatively expensive when compared to the circuit I am powering. The Sun Saver Solar controller from MorningStar would work because it disconnects the load at 11.5v. Is there a simple, low power solution that I can build myself?

Then I would sugest to read the thread where the answer is in it already

Robert-Jan