Automatic 4 battery swapper

[Gary Hammond]

I just got my automatic four battery swapper unit built and hooked up to a small roller skate wheel sized SSG for testing and it works great. I designed and had printed four PC boards to hold the relay circuits and then hand wired a small universal board for the trigger coil interrupter relay circuit. There is one board with 6 latching relays for each battery, so 4 relay boards with a battery connected to each board. I have it wired in a "split the negative" arrangement and it is time based switching every 6 minutes. The device uses an Arduino Mega 2560 Microcontroller to supply the logic for all the boards. An auxiliary 12 volt battery powers the Arduino and the momentary 100 ma combined total for the relay coils each time it switches.

I test ran it for 9 hours yesterday and the overall voltage of the 4 run batteries only fell from about 13.12 volts at rest to about 12.95 volts at rest. No over unity with this small machine, but it still recycled a lot of electricity!

Here's some photos of the device.







Gary Hammond,

[Gary Hammond]

And here's a video of it running.

Gary Hammond,

[Gary Hammond]

Hi All,

Just a little update on the swapper. This afternoon I hooked it up to my best bicycle wheel SSG and paralleled each of the 5.5 AH batteries to get 11 AH's each. It was taking the first two charge cycles way over 17 volts in the six minutes I had it programmed at. So I reprogrammed it for a 2 minute cycle time and ran it through about 8 cycles. And it was still charging each parallel pair of batteries to about 16.5 volts. It's not good for the batteries to charge this high!

I either need to load the output or go to a voltage based sensor instead of the time based way I currently have it programmed. Or maybe both. This is the same SSG I had hooked to the manual swapping circuit with 105 AH batteries.

Gary Hammond,

[Aaron Murakami]

Hi Gary, did you design the whole circuit yourself or did you use some ready made schematic?

[Gary Hammond]

Hi Aaron,

Made it myself. Figured out a schematic, designed the boards and had them made. Bought all the components and then soldered it all together. I did previously wire up a "poor mans split the positive battery swapper" following R.S.'s schematic using circuit breakers. Then I changed it to a split the negative so I could get away from using an isolated winding on the power coil. The relays are all connected the same way as the circuit breakers were.

Here are pictures of the control schematic and relay boards.





Gary Hammond,

[Gary Hammond]

Further update

I hooked the battery that drives the Arduino and switching up parallel with the charge battery, and after a couple of hours running this way all the batteries were beginning to drop in voltage. Then I changed the switching cycle to every 4 minutes and reduced the switching on times from 500 ms to 300 ms. It ran this way for about 5 hours and the battery voltages all continued to gradually drop.

When I saw the initial 17 volts of the charging battery on the first couple of cycles, I got excited thinking this was going to go overunity. Working with fully charged batteries gives better results than after the batteries all get a little lower in charge. As the battery under charge gets near 15 volts, the differential voltage running the machine can drop as low as 10 volts. This is too low to get the desired results from the coil collapse.

This machine apparently isn't going to self run at 12 volts input. Haven't tried it at 24 volts yet. The batteries may be too small for that. Instead, I will probably try swapping 24 volts on my attraction motor first to see how that one performs. It has a lower current draw and pulls a flux gate generator which can recharge the Arduino battery.

Gary Hammond,