"Larry Jaques" wrote in message
...
On Sun, 12 Mar 2017 10:30:01 -0400, "Jim Wilkins"
wrote:

David R. Birch wrote:
I've been think about installing this in my garage which has no
electricity running to it. I've googled "Solar Powered Garage
Door
Opener" and only found general talk, no one who's actually done
it
and no company that makes one.
...........

I went with circuit breakers for this, both in to the
controller/battery and out to the water heater.

Be careful which Type of low voltage circuit breaker you use.
http://www.wiringproducts.com/atc-circuit-breakers
I looked into them recently for my homebrew 50A battery charger's
output and the advice I found was to use manual reset breakers for the
fastest opening.

The two chargers I assembled from my collection of transformers and
Powerstats have ~0.1 Farad output caps that let them serve reasonably
well as variable power supplies. The caps mainly stabilize digital
meter readings. A series diode prevents big sparks from a surge back
into the cap when connecting to a battery and a shunt diode plus the
fuse or breaker (should) protect the cap from connecting the battery
backwards (dirty battery, dim light), since reversed polarity would
pass through the series diode. Electrolytics are supposed to
olerate -1.5V without damage.

The 50A charger was made from a 50A buzz box arc welder transformer
which has enough self-inductance that the output ripple is only around
1V p-p at 20A. As expected, as current increases the voltage droops
considerably until it hits the normal arc voltage range around 20-30V.
It needs to be watched or run through a solar controller when charging
batteries because the current doesn't taper off as voltage rises.
However it brings them rapidly to the point where I can change to a
smaller charger that does taper automatically. I bought this for when
I need good voltage regulation and current limiting instead of a
bulletproof output.
http://www.ebay.com/itm/like/1422203...&ul_noapp=true

Temperature rise limits its continuous (1/2 hour) output to 25A at
28V without the fan. With the fan the outside is cooler but I can't
measure internal winding temperature. The 30A NTE panel-mount breaker
passes 50A for about a minute, long enough to measure the voltage drop
across whatever I'm testing.

I use 1A or 10A from it to measure wire resistance to the milliOhm.
Unlike an expensive lab supply the output tolerates abuse such as
inductive kickbacks from motor and transformer windings.

...An ammeter in the load
cable will affect the low voltage cutoff.

By what, 0.3v (a diode drop), Jim? Will a shunted ammeter also
cause
troubles like this? What's your depth of discharge, anyway? Sounds
too deep.

The "burden" of the shunts is 75mV at full rated current.

Since my batteries are only for backup there's no normal depth of
discharge. If I knew how I'd try to get maybe 5 - 10 cycles from a
backup battery, but I wouldn't discharge it that far or leave it there
unless I had no choice. IIRC my laptops are set to warn me at 30%
remaining capacity.

When I test lead battery capacity I draw them down to around 10V -
10.5V at the expected current draw. As the electrolyte depletes their
internal resistance rises and decreases the loaded terminal voltage;
they recover to above 11.5V. The drop is so rapid below 10.5V that
there's little to gain by going lower. You can pull more current from
them later by letting them rest while the electrolyte diffuses.

My system uses non-gendered Anderson Powerpoles to connect the
components and a meter like this can be temporarily inserted to
measure voltage and current.

Wish I had the good luck you guys seem to with those, but the lone
pair I've used has been trouble 2 out of 3 times I've used it.

I use them because my digital meters read current in one direction
only and batteries both charge and discharge. They also disassemble
with a small screwdriver unlike other options. Generally I solder the
inner end, leaving the exiting strands free to flex, then pull-test
the assembly and sometimes check the voltage drop. I have the cheap
imitation crimper and that's enough to make it work. It came from a
flea market and wasn't marked to do Andersons although it had the
right pin numbers so I don't have an on-line reference.

https://www.amazon.com/Digital-Batte.../dp/B01HM24MVG

Those are going for $7.54 delivered via eBay sellers.
http://tinyurl.com/hqk98uy

I prefer the better Bayite (plastic box) ones you referred me to
that
go for $18 w/ 100A shunt, but have pairs of each for the new system.
I bought 50A shunts for the Bayites @ $5 a pop because my initial
1kW
system will be 37.5A max.

Metering is a user preference. I suggested my opinion of the low-cost
minimum to detect problems but I have many more on my systems.

The Amp-Hours total is useful on solar systems. It will preserve the
reading if you add a 9V battery to keep it powered through the JST
connector (from an RC hobby store.) Watt-Hours is less useful
because
the battery discharges at a lower voltage than it charges.

Huh? 'Taint a 9v connector.

I think it's a JST connectors used on model airplane servos. The
pinout on my blue 60V 100A Watt Meter is Neg, Pos, Reset starting at
the top. The pin spacing measures 2.5mm.

-jsw