On Tuesday, June 19, 2018 at 5:47:17 PM UTC-7, wrote:
If you drop 0.7V on a 13V battery, you're going to be drawing 5% more power and have 5% less endurance.

I have slightly different numbers. I use low drop (0.4 v) diodes and per battery switches with K2's. Operating rules are when bus gets below 11 volts, it's time to turn on another battery. (about 4-5 hours with glide computers and transponder, and not too talky on the vhf.)

Most of the electronics have switching power supplies, so as the volts go down they adapt and the amps go up. The K2's have a fairly flat discharge curve up until they fall off the cliff. All the diodes do is slightly adjust where on the cliff edge we stop and go to another battery.

So at the worst case switchover point the batteries are at 11.4 and I'm loosing 3.5% in the diodes. Costing perhaps a few minutes out of 4 hours. Seems simple and cheap to prevent bus glitches and prevent circulating currents between the batteries.

The old plan was pairs of SLA's as friends. They got bought each year at the same time, charged in parallel, and discharged in parallel. This also worked, but the K2's seem to last for years and provide more run time.

But you are also losing and additional 3% in the higher current draw, 6.5% total, 15 minutes out of the 4 hours. Those of you running two batteries should consider simply running them in parallel. Just make sure you charge them in parallel too. This is actually better for battery life (particularly true for SLAs) because the average depth of discharge is less. Might want to check with the battery manufacturer and see of the BMS will like that - some do, some don't.