Hi all,

I have a number of 12V lead acid batteries which need to be charged and discharged repeatedly. They are discharged in pairs (in series) and in order to discharge them, I have a resistive load of 2.2 Ohms to draw a bit over 10A on average.

This all works fine but I have to continually monitor them to ensure that the total voltage does not drop below 22V to avoid damage to the cells. Therefore, I thought that it must be possible to devise a basic controller for this, with a relay to disconnect the load when done.

My original thought was to use some zener diodes to control the relay, but it would also be beneficial to log the voltage over time to see what effect the charge/discharge cycles are having. So I thought that maybe a basic microcontroller would be the best bet to do it all - my question is would the basic schematic attached be sufficient, or would I need more interfacing circuitry that would make this more complicated that it's worth?

Note - the potential divider in parallel with the load is purely to scale the voltage for the microcontroller ADC. It is this voltage which would be logged. The button in parallel with the relay is to turn it on initially - provide power to the microcontroller so that it can turn on the relay for sustained operation. Once the end voltage is reached, it can simply turns off the relay and it all shuts off to prevent further drain.

Thanks!

 

...12V lead acid batteries which need to be charged and discharged repeatedly.

What battery "needs" this?

 

What battery "needs" this?

I want to get the maximum capacity from the batteries. I have been advised by the manufacturer that always keeping the same pairs together and charging/discharging new batteries 5-6 times will help to train all the cells.

I did this with a few pairs and it did seem to improve capacity, so I would like to do the same for future pairs.

 

A simple approach would be to just use a timer. I mean, you "know" (by looking at a data sheet) how long it should take for a fresh battery to reach the discharge level you need. Maybe set it for a few minutes less and you're done?

In your circuit, I'd use a power n-channel MOSFET to switch the load and just drive that with the microcontroller directly, eliminating the BJT and the relay. You will need a "logic level" MOSFET if the µC voltage output is less than 10V. That just means a MOSFET that turns fully on at whatever voltage signal you have. A normal MOSFET needs 10V, a logic-level MOSFET as little as 3V or so.

 

That's true, it would be an easier solution. However, I still would like to log the voltages and to have it all built in would be nice so I would most likely still use a microcontroller for that. Besides, a timer would not be much easier than the original zener idea, would it? I assume it would still just drive a relay?

Do you think that in principle the design above would work? Do you think that I would need some protection for the relay contacts, since on make/break, there could be hefty currents?

 

If you already have a µC watching the voltage, there's no need for any other voltage detection such as a zener, and again, I'd drop the relay and use a MOSFET. Switching a resistive load at ~10A at 24 volts is "trivial". If you already have the relay, it's fine to use.

 

Thank you for your advice. Not sure why I thought a relay would be best, probably just thought it easier.

Would the fact that the µC 0V supply is higher than the FET source (assuming N channel) cause a problem with driving it?

 

I'm not sure what you mean by that. Aren't the battery and micro sharing a common ground? You need the gate voltage (depending on the MOSFET) to be in the range of 0-15V relative to the source pin of the MOSFET, which is normally at circuit ground.

 

In my schematic above, I use the FET/relay to remove power to the µC by sitting in line with the power rails. This is in order to ensure no further battery drain. So the µC 0V line is actually connected to the FET rather than battery 0V.

Or is there a better way?

(As an aside, so that I can learn for next time - what is the reason that a FET is better than relay? Is a relay just overkill unless for more power, or does it use more power to keep it turned on?)

Thanks for all your time.

 

Oh I get it. Yeah, don't do that. The output state of the controller cannot be defined when it is not powered, and it could be bad for it to have inputs connected with no connection to both power rails.

Leave the controller powered. Compared to the currents involved, that little micro will probably draw less than the self-discharge rate of the battery.

A relay would be fine for this but is larger, has moving parts, and does require power to energize the coil. Relays are more costly than a transistor and become increasingly challenged as switching frequency increases. That concern is not relevant to you application. Relays (including solid state relays) are fantastic when you need to completely isolate one power supply from another for instance when you are using DC to switch an AC line. Again, not relevant here.

 

If you just want to cut off discharging at 22 volts, you can use a voltage comparator (search for 741). You can also use LM3914 to monitor the voltage condition of your batteries on 10 step LED display. Just configure it properly and you can use one of those LEDs to operate the cut off relay.

 

If you just want to cut off discharging at 22 volts, you can use a voltage comparator (search for 741).

The OP has said he wants to log the voltage decay data during the process, so the micro is already there for that purpose.

And the 741 is an op-amp - the oldest and worst available - not a comparator. LM339 quad comparator is a popular choice.

 

Thanks to both of you for the feedback.

Ok, I will leave the µC connected and just put it into a sleep state to minimise current until I can disconnect it. Glad I checked that and didn't end up damaging anything!

Thanks for clarifying the pros/cons..

I think I have all the info I need to try and design something... thank you very much!