"ads" wrote in message news:jdniqgt3jvgnrnavu...@4ax.com...
If you're knowledgeable enough to build your own from available parts,
LiFePO4 cells can make batteries which are good for 5,000
charge/discharge cycles (about 13 years at one charge/discharge cycle
per day) at prices comparable to AGM sealed lead-acid batteries
(what's typically in the older jumpstart packs).
I've done that on a very small scale when the manufacturer wanted $50
for a replacement AGM battery for a small UPS - more than twice what
the UPS cost me. The LiFePO4 cells plus a BMS plus tax and shipping
were under $20 and I spent less than an hour building the battery
pack. The new battery lasts longer than the original at the same load
and, at my age, it might be a "lifetime" battery ;-)
If you have some hours of sun available, then solar might be your
fallback power source. No one who has built even a small solar system
would say it's cheap - even the solar-charged LED lighting in the shed
out back was about $200 for parts - but MUCH cheaper than having an
electrician out to run power under a concrete driveway to get lights
out there for maybe 20-30 minutes a day.
I don't say isolated DIY solar is cost-effective compared to grid power, or
grid-tied solar, but it's about the only alternate source that's easy to
experiment with, so I claim what's installed and working on my Form 5695 and
write off the rest as a hobby or education. I've also built windmills and a
water wheel and briefly looked into bio and wood gas generation.
Solar certainly does work well enough to be part of a backup plan if your
area suffers from extended power outages, such as New England's ice storms.
With full winter sun my batteries recharge to float voltage by ~10AM after
running the fridge all night. Cold panels are more efficient, the other day
I saw 440W from the four 100W panels. At float voltage lead acids don't gas
but revert to a slow taper charge. I've read and tentatively confirmed that
the battery is at 70~80% SOC when it reaches float voltage. When the grid
returns I can fully charge and equalize them outdoors.
Wishfully dividing the lower purchase price of flooded marine batteries by
the longer claimed cycle life of AGMs, the cost of battery depreciation per
KWH is still higher than my $0.18/KWH grid power. Depth of discharge doesn't
seem to matter if you keep total capacity constant. I've seen a graph of DOD
vs cycle life that revealed the same lifetime total delivered power for all
The obvious alternative is a generator, if you don't mind running it all
night for the fridge. I found a small used Honda that I can completely drain
to store indoors and carry out to a small shoveled patch to quickly recharge
the batteries if it's overcast, while I struggle to clear the path half way
around the house to a larger generator in the shed. I already had a
feedthrough to pass other wiring through the wall. At my age I have to plan
for injury from falling on ice and thus have an easy to carry backup. I also
rigged the wood stove chimney so that I can flip open the rain cap and run a
cleaning brush through it (weekly) while standing on the ground. The unusual
chimney, handrail and antenna structures up there are painted to disappear
against the surrounding trees.
I've been lucky enough to find fairly reasonably priced AGMs for enough of
my second-hand UPSs, from Amazon or replaced-on-schedule ones from a flea
market. The dealer told me some had to be swapped out from critical devices
after 3 months. He has a conductance tester and lets me check the promising
ones with my HF carbon pile tester. As a favor I tested all his stock and
found that one of the brands he bought for nearly his selling price wasn't
worth salvaging while another was excellent.
If you can run a low-cost UPS at greatly reduced load it may be OK with
larger external batteries, unless it has shutdown timer. A common
cost-cutting trick is to match the heatsink's thermal mass to battery
capacity to avoid fan cooling. A larger battery can overheat them. Even my
originally expensive APC1400 has quirks to avoid or reprogram.
I was a Lithium battery tech and know how to build a BMS, and I keep an eye
on Lithium prices, but for liability concerns I assembled my solar system
from purchased power modules and packaged batteries and wired it according
to published standards. The solar>grid>battery switching was accomplished
with diodes and careful voltage adjustment, not active control. If there's a
fire here anything I built is automatically suspect. Although I've been on
the build teams for prototype electric vehicles I don't have the experience
or equipment to make a safe plastic enclosure for bare cells, I can only
fabricate from sheet or "billet" metal.
At Segway we used only the manufactured battery modules to build
experiments. Mounting them on other than the normal chassis casting was my
problem. The dimensioned CAD print says "see the pattern" for the battery
studs so I had to reverse engineer them and cut-and-try the locating taper
on the old manual metal lathe in my basement, which was easier than getting
time to fiddle with the CNC machines or learn how to use the CMM.
My wet batteries are in boat boxes and were wired according to boating
specs, mainly waterproof inline fuseholders on 7" leads, just outside the
box where they won't ignite hydrogen if they blow. Some of my inline
fuseholders have needed their Fastons tightened with needlenose pliers. The
AGMs are replacements in jumpstarter and power pack housings, with added
I use my homebrew controllers only as temporary battery chargers and
testers. Much can go wrong in a battery powered system, such as the battery
backfeeding into an unpowered supply, overheating when recharging a drained
battery at high current, or failure of blocking diodes, so I overspecify and
bought the equipment to test components and systems thoroughly.
For example a blocking diode on the input of my HF "45W" controller shorted
and sent battery voltage to the panels, which fortunately did no harm. I
have meters everywhere and saw it, and found they had used two undersized
Schottkys in parallel instead of one with an adequate rating and larger
package, which is what I installed. Afterwards I added blocking diodes to my
panel wiring and curve-traced the leakage at night.
The returned Li battery packs (not Segway) I investigated for warranty
repairs showed a wide range of performance. The BMS recorded their history
of charges, discharges, temperature, etc. One from a vehicle in hot Arizona
had deteriorated in less than a year. Many showed significant loss of
capacity in 3 years, with a wide spread of values between packs and cells. I
have obtained up to 15 years of useful life from flooded lead batteries at
half the price of AGMs and 1/4 to 1/10 the price of Lithiums, so my
new-old-stock of $1 tabbed 18650s is reserved for rebuilds.
I've seen similar results from AGMs, including complete failure in 3 years
though I have one that's still strong after 10. A common fault has been a
bad cell that caused the other 5 cells to overcharge at 120% of the float
voltage. I suspect but can't prove that as the reason they swell. I got the
APC1400 UPS for my solar system for free from a dealer because he couldn't
figure out how to removed the badly swollen AGMs. Its fairly high idle power
drain appears to be typical of true sine inverters, so I'm exploring DC
appliances such as the T60.