minimalist SLA battery float charger

[Ian.M]

I've been futzing around with a design for a nominal 13.8V current limited and temperature compensated float charger for maintaining SLA batteries.   

The requirements are to run from an old laptop power brick providing between 18V and 20V DC, and to closely follow the battery manufacturers charging specs, and not to drain the battery if the supply is off.

Its short circuit protected, but not reverse polarity protected.  Multiple batteries could be maintained by fitting more Schottky output diodes, or for higher value batteries, it would probably be preferable to duplicate the whole circuit but share a PSU.



Here's the simulation results.  The X axis is the terminal voltage of the battery being charged.

The temperature sensing diode next to the battery has been swept from 0 deg C to 50 deg C (highest/rightmost to lowest/leftmost curves respectively)

At the moment its only a LTSPICE simulation.  You'll need the TL431 model from http://www.audio-perfection.com/wp-content/uploads/TL431.zip.

Comments and suggestions are appreciated.

[Circlotron]

Seeing you are going minimalist, I'd just go for a bare bones adjustable LM317 setup but with a high wattage resistor on the input side. Size it so that the '317 input voltage is just enough to stay in regulation at the maximum required current. Have used this setup on about 3000 chargers for 12V 7Ah SLA charger for automatic sliding doors with zero failures. Set it for 13.65V @ 20 deg C and Bob's your uncle. With a variable temperature environment of course, you might want to stick with temp compensation.

[Ian.M]

Good point and thanks for the feedback.

I should probably knock up a few of your circuit using a car bulb as the ballast resistor to get the batteries on float while I work on the final design, or there is a taper charger circuit that combines current and voltage limiting using a combo of the LM317 constant current and constant voltage circuits, Fig. 1 in the original National Semiconductor Linear Brief 35 (republished http://www.ti.com/lit/an/snva581/snva581.pdf).
I suppose I could add a resistor/thermistor network to that to get temperature compensation.

However the added TL431 does give me a clean switch over between CC and CV modes, which should allow faster recovery from partial discharge than the simpler taper charging circuit as well as allowing simple temperature compensation.  The extra diode in parallel (D2) should be at the coldest point in the charger enclosure and is there so it doesn't loose CV regulation if the sensing diode (D3) gets disconnected.

I have considered dropping minimalism and complicating it with a PIC MCU to provide status monitoring, switching  resistors into the feedback network to add a fast charge mode that then drops back to float mode, so I am still interested if anyone can spot any flaws in the basic circuit.

[mikerj]

How about replacing D1 with a relay and have the coil energised by the charger supply voltage?  That way you get rid of the temperature and current varying Vf drop and have zero reverse leakage for minimal complexity and cost penalty.

[Ian.M]

I probably will use whatever 2A 40V Schottky I can get cheaply rather than a genuine SS24, with a preference for leaded components, and I don't expect to pay more than £0.50 per diode, which certainly wont get me a decent relay. 

Also, as a LM317 can pass reverse current (up to 15A pulse), I'd need a blocking diode on the input or a voltage threshold circuit to prevent the battery holding the relay in.

I'd also loose the ability to gang multiple batteries off the same float charger simply by adding output diodes for each.

The output diode Vf drop and its tempco may even be in my favour as I was thinking of using two sense diodes in series, to increase the compensation and was concerned that that would be overcompensated.

However, your suggestion has been very useful as if I add the relay with a MOSFET on the 0V side of the coil with the gate to the battery (via an ESD protection network) I can also use it for battery reverse polarity protection.  I suspect a low part count solid state reverse polarity protection circuit idea will come to me in due course so thanks for jogging my thoughts in a useful direction.

[7aman]

Hi.
Sorry for bringing up this old post.

I am designing an intruder alarm system. System needs a backup battery. Battery type is 12V sealed lead acid battery. Battery capacity is limited to 7Ah and typically it is a 4.5Ah one or 7Ah. It is recommended to limit charging current up to 10% of capacity. So I decide to use 450mA as current limit.

I tested some battery charger that are using lm317 and I am not satisfied with their performance.
I test your circuit. It is working good enough. But I wonder it could be more minimal?

For now I don't care about temperature compensation so I removed D2 and D3 and R6. Also I removed other elements such as R5, D2, C2 and overall performance is almost the same.

I have some questions about it.
1- Did you actually make your circuit? And what was the long period performance?
2- What would I miss by removing D4?
3- What would I miss by removing R5?
4- What are the advantages of adding capacitors in lm317 output like C2?

 Final circuit is this: