Matson make a series of lead acid battery chargers. I picked up the MEC060 model designed for flooded batteries for AU$129 from Battery World. They are switch mode, automatic three stage chargers. Mine charges at up to 6A until the battery reaches 14.7 volts (bulk stage), hold it there for over 2 hours (absorption stage), then switches to float stage (holds at 13.8v). They make a Gel model that does the same thing except it charges to and absorbs at 14.1v.
I happen to have both flooded and gel batteries that I'd like to charge with the same charger, and avoid paying for a second charger (AU$160, and a special order from Battery World, at least from my local franchise). One possibility would be to change the output voltage to a compromise like 14.4v, which is perhaps a little slow for chargine floodies, and a little hard on gels, especially at low temperatures. When I removed the top cover, I found three trimpots, as shown below:
VR1 sets the maximum charge and absorb voltage, and is located at the bottom of the above picture, just to the right of the yellow transformer. I found that mine allowed a variation from 14.45 to 15.3 volts (yours may vary of course, depending on the exact characteristics of various components). Unfortunately, 14.45 is just a little high for my liking when charging gel batteries, and also changing the absorb voltage also changes the float voltage. The difference between absorb and float doesn't change much, stating at about 0.9v. That would make the float voltage about 13.55, which is still quite reasonable (13.5 to 13.8v are considered good float voltages, at 25Â°C). You can make this modification without even removing the printed circuit board. (Removing the PCB is not a big drama, just 5 screws and two awkward nuts near the fan). I found by trial and error that the other two trimpots seem to set the maximum current in the bulk stage, and the temperature at which the internal fan cuts in. There would be no good reason to change either of these (however I did, since I guessed wrongly which trimpot would set the absorb voltage, and I thought one of them might set the float voltage independently of the absorb voltage).
Wanting to set the absorb voltage to 14.1 for gel batteries, I traced the circuit as follows:
That way, I can adjust the output voltage to still be 14.7v, but reduce it to just over 14.1v by shorting out VR1. All these changes can be done without removing the PCB.
Although the above picture shows all through-hole components, a good part of the circuitry is surface mounted underneath the PCB:
The thick black wire is original.
The problem with the above is that the float voltage is now about 13.2 with VR1 shorted, and about 14.0 without it shorted. 13.8 is the very maximum I'll tolerate for the float voltage, so when the 16K resistor is added, R18 needs to be increased to compensate. I found 2.7K to be suitable. I could live with the reduced float voltage when in the gel position, or switch to the flooded position when the battery is charged. The trouble with the latter is because the absorb phase is timed rather than current based, if the power fails while a gel battery is connected and the switch is in the flooded position, the gel battery will get over 2 hours at 14.7 volts, which is too much. So to do the job properly, R18 needs to be decreased in value when the switch is in the gel position. I found paralleling it with a 3K3 resistor is about right. Since R18 is under the board and needs to be replaced, the PCB has to be removed. A connection has to be made to the collector of Q2. This is circled above, along with an unused hole conveniently nearby that will fit a wire-wrap wire.
The final circuit is as follows:
Internal view with switch installed:
You can see a new TO92 transistor at the bottom right. This is because I managed to blow up the original Q2; I had to keep shorting its base to ground to turn it off, so I didn't have to wait 2 hours all the time for the absorption stage to time out. Somehow, I managed to fry it such that there was a short from base to emitter. Amazingly, I couldn't find a replacement surface mount transistor in the garage, so I carefully drilled three holes and mounted some TO92 NPN transistor through three new holes. There was no chance of reliably soldering it under the board; those surface mount transistors are really tiny. Obviously the characteristics of this transistor are not important.
View from the front:
I'm getting 14.70 (adjusted by trimpot) for flooded absorption stage, and ~13.7v for float (the float voltage jumps around a bit). When switched to gel, I get around 14.15v absorption, and about 13.6 volts on float (also jumps around a bit). I've adjusted the current limit to very close to 6.00 amps, and the fan seems to come on at a reasonable heatsink temperature. I've set the trimpots at these levels using white-out (liquid paper stuff). I'm quite happy with these numbers, and the final result.
These modifications worked for me, and would probably work for you too, but it depends on the tolerance of individual components. Be prepared to vary values a little for best results.