Shunts

[GTBecker]

I've been diverted to another unintended project, providing clean 12VDC to the projects on the boat. I've built a battery-monitor/charge-manager using a Netmedia LCD-X that I had on hand. That processor/keypad/LCD/relay driver controls a sealed box in the engine compartment that relay-switches batteries and a charger/power-supply, and samples battery and load voltages. http://www.ustream.tv/channel/hydralight

The device alternates two Group 27 lead-acid semi-deep-cycle batteries between project load and charge so that the load never sees the charger and the battery on load never discharges below 50% (12.2v). [There are also two Group 27 batteries on their own small charger that I do not manage which start and run the boat.]

Naturally, this thing has become much more capable than necessary for the task, but that's part of the fun - taking an idea to its practical completion.

I've gone through several commercial multimode chargers and returned each after finding that none can can do what I need, so I've now settled on a Todd PC45 from ebay, essentially a 13.6v 45A power supply that provides a simple grounded control input to boost the output to 14.6v. That allows two easily-controllable charge modes, 13.6v for bulk charging and floating the batteries, and 14.6v for "finishing" (the last 10% of a complete charge, when the electrolyte gasses). [BTW, I've learned that a $5 carbon monoxide sensor at Sparkfun ( http://www.sparkfun.com/commerce/product_info.php?products_id=9403 ) is actually better at sensing hydrogen than it is CO, so I might add gassing detection, too.]

I want to add three shunts to measure the charger output current and each battery current, and I think I have the apparent choice of putting a shunt in the positive leg or the ground leg of each. The advantage of the ground leg is that the shunt voltage (50mv for 30A) is referenced to ground so the measurement can be single-ended, needing only a simple amplifier and one ADC input each. The disadvantage of that path is that there are many boat load returns that would need to be combined on a busbar that would then connect through the shunt to ground (the common boat ground is the engine block which connects to the primary start battery negative post, where most load grounds congregate). Alternatively, a shunt in the positive path requires a differential measurement but it is logistically easier to arrange, and I have already run one of the two required lines - to measure voltage at the battery posts.

Any experience or suggestions?

Uh, sorry for the verbosity. I didn't intend write a short tome.

Tom