Why do battery needs a float current?

Yeah, your installation is certainly deserved of better wiring and I'm pretty sure you'll like how solar handles charging.
I never did any measuring or correlation of charging specifications to as built numbers. I just threw heavier wire at everything and called it a day. Most of my connections use #4AWG welding cable as I needed some and availability through a local distributor mandated a minimum of 100' purchase. Of course, I didn't use the entire roll but will have other uses for the remainder. I'm already thinking ahead to installing more battery capacity. A guy needs power to fry all of those fish, lol.

Good luck with the rest of the install.

No way! Greater than 95% of that power is going into the batteries.

Perhaps your batteries aren't as fit as used to be. The spec is rated for a new battery.

Sal

Almot wrote:


Heat loss in poor contacts and booster cables must be the reason.

Mex, give me a break... Manganin shunt and precision DMM in Baja, sure... Besides, +/- 0.1A accuracy of my crude measuring device, namely the controller display, was enough to notice 0.3A current where in theory there should be 0.14 A (total of controller internal consumption and self-discharge). Not that Trimetric is more accurate, same +/- 0.1A.

Heat loss in poor contacts and booster cables must be the reason. Controller is maintaining the battery voltage at the right level due to thin voltage sense wires, but the power loss in power cables and contacts (or voltage drop, whatever) causes controller to increase the output so that wouldn't be a voltage drop on the battery end. Which in turn creates those extra currents. Or the currents cause controller to increase the output. Whether it was egg or chicken first, heat means current.

westend wrote:
Is that #8 battery booster cable copper coated aluminum or stranded copper? I just bought some 4 AWG booster cables, got them home, and found out they are copper coated aluminum. Also, those jaws are about the worst connector to a bare wire when you look at the amount of surface area in contact.

Don't know if it's a pure copper, but the booster is a cheap Walmart junk, jaws are falling out already. Yes, jaws have little contact with bare wire, and the copper-ish metal on them is more green than yellow :)

Yesterday put up the batteries to real service - connected to trailer DC wires. With circuits of LP fridge, bulbs not yet converted to LED, pumps and all. This morning batts went into Absorption at 10 am, into Float at 11 am (3A Float setpoint), and at sunset were still floating at 0.7A. Worse than 0.3A previously observed after a week of pure floating, but this was to be expected. Current with trailer loads was 1.3-1.4A, but fridge circuits and other parasites draw 0.6-0.7A on the average, so this leaves 0.7A for battery. Hopefully things will improve with proper wiring. No time even to go fishing, dammit...

You are trying to analyze chemistry with hexadecimal logic. Sort of like using an adding machine to display a recipe for chocolate chip cookies.

The only measurement that makes sense in battery logic is watt hours.

And the only reliable way to determine low scale amperage is with an appropriate manganin shunt and precision DMM.

A 97% storage efficiency for 30 days would not be unusual.

Measure one battery against the other if you have a doubt about your finding.

Almot wrote:
My only guess is that battery becomes a space heater while on the float…

Correct! Your batteries, as well as the cabling & each of the interconnections, are resistors in the charge circuit. So, while a portion of the current is used to replenish losses due to self-discharge, the balance of the energy is being dissipated, in the form of heat... from every component in the circuit.

Using your example of 0.2A of current that you have not yet been able to account for, and assuming there are no other parasitic loads; It appears that the 2.7W (13.6V * 0.2A) is being "used" for heating (and electrolysis). That would indicate a total circuit resistance of ~68 Ohms (R=V/I, or 68=13.6/0.2) . Although this resistance value is much higher than I'd expect to see from healthy AGM batteries (I'd guess that'd be just a couple of milli-ohms each), it's within the realm believability for the entire circuit, especially when considering your use of booster-cables/clamps.

A closer review of your wiring (especially the crimped/soldered/bolted connections) may reveal some opportunities to reduce resistance and improve overall efficiency. But at the end of the day, since you are charging with solar, I wouldn't worry too much about the losses you've described.

FWIW, with my system (also 300A/h AGM's), with batteries at 100% SOC I typically see a float current of ~0.1A @ 13.6V (charger display indicates "0A"). Of course the resolution of my TriMetric is limited to +/- 0.1A, which hinders my ability (and interest) to monitor with greater accuracy.

Cheers,
-Mark

I would guess your current measuring anomalies are due to the connections. Is that #8 battery booster cable copper coated aluminum or stranded copper? I just bought some 4 AWG booster cables, got them home, and found out they are copper coated aluminum. Also, those jaws are about the worst connector to a bare wire when you look at the amount of surface area in contact.

PS: I shouldn't have made fun of BFL13. It's just that I envy his Trimetric monitor :)...

He was right that "monitor" needs power, and my controller as a monitor needs power too. There is 5% loss in MPPT conversion, yes, but there is also a fixed amount of power that controller draws all the time: 0.9W typical operating power, 1.5W maximum (with backlight and other things On, though they were Off in my case). Let's agree on 1.2W - easier to divide by 12V - so 0.1A is accounted for, 0.2A still remains.

3 other possibilities:

1) 10ft #8 booster cables (20ft total 2 leads) from controller to battery. (There is also a voltage sense wire so let's not go into voltage drop). This resistance shouldn't result in 0.2A current, no?

2) As I couldn't clip booster cables to controller posts, I clipped them to a pair of scrap wires #12, each piece 1ft long, and screwed those wires to controller posts. So there is additional resistance in these #12, and also on all the jaws of the booster.

3) I've badly damaged my AGM during 10-11 months storage when they self-discharged to 60% SOC, as described in other post. Though this SOC is not terribly low, and other than mysterious runaway Float current the batts behave well. Load-test of one battery with 100W 12V bulb (repeated twice) showed that it has retained its full capacity - and then cr-appy Chinese bulb burned out so 2 other batts were not tested. But they all show the same daily AH intake if floated one by one.

Water in AGM doesn't - normally - boil out.

Float voltage =13.6, battery manual suggests 13.6-13.8 for "float use". Temperature-compensated with sensor on battery, right now 13.8V at 18C, just where it should be. Outside thermometer few feet away reads 19C in the shade but batteries are in the shade AND in the wind.

Controller display accuracy +/- 0.1A exceeds the discrepancy of the observed 0.3A VS expected 0.0A (display won't read anything below 0.1A). Separate multimeter reads quickly fluctuating current 0.21-0.40 A, controller display is slower to respond and is averaging it to 0.3A.

red31 wrote:
float voltage too high

This.

you cant over charge a battery if you dont boil off the water.

The float voltage is to simply insure the battery's self discharge is compensated for. So once it's bulk charged....