Jon Parker wrote:
On Wednesday, April 22, 2020 at 2:26:48 PM UTC+1, wrote:
There is no obvious need to switch the power between charge and the
sockets?
I know very little about solar PV. I considered that an appliance drawing more power than the panel could supply might harm the panel, so I thought it best to switch the panel out of the circuit to ensure power was only ever drawn form the battery.
If I were considering doing this, I would be installing a mains powered
12v (13.8v) SMPSU - do you not have any mains power in the garage?
I do have mains power in the garage. I will look into this solution, seems to be neather than my proposed solution!
Thanks
You need to study the "load" carefully, as the SMPS pricing
will rapidly shoot skyward, if you attempt to get too much capacity.
The thing is, ATX power supplies come perilously close to
sufficiency and ampere-capacity. There are supplies with
ratings up around a kilowatt or 1200W or so. (Think 12V @ 80A
and the PCI Express cables)
ATX computer supplies put out 12.0V, while a
battery eliminator might put out 13.8V. When you look at this,
you might say "well, surely a small adjustment will give the 13.8V",
but it isn't always that easy. (Some equipment has overvoltage
protection, which is integrated somewhere and shuts off the
equipment if an "abnormal" value of voltage appears. The 13.8V
adjustment might cause that protection to trip. A control in
one part of the circuit, does not necessarily control every
subsystem in the box.)
Supplies come with "expectations" about the kind of load expected.
capacitive ["5000uf max" or a similar statement, phase margin]
resistive (ham radio???)
inductive (motors... 10X loading from motor when rotor stalled)
Motors are particularly tricky, because of their load current
behavior. When the rotor is stalled, some motors draw 10X the
nameplate current. On an SMPS with electronic overcurrent
protection, the time constant on overcurrent can be
quite short. Which means the SMPS turns off, as soon as you flick
the switch on the air pump.
On capacitive loads, the message here is about the "hobbyist"
approach to electronics. To make projects constructed at
home more robust, a hobbyist adds capacitors. A capacitor is
a short term energy store. And you would think, "well, perfect,
I slap those into my circuit, the motor 10X current comes from
the capacitor, honor is satisfied". But the SMPS has an internal
feedback circuit path, and if around 5000uF is added to
the output of the circuit (a small amount compared to what
clever hobbyists buy), this drives the control loop crazy.
As a general rule, easy public specs for SMPS do not provide
a phase margin statement, and what the limit is for adding
capacitors to the output. You must contact the supplier
and beg for the information (like we had to do at work
when buying these).
We can't "fix" the ATX supply particularly, by adding capacitors.
Ham radio operators, look for items similar to the following. Radios
intended for in-car operation, need a bench supply when the
radio is run in the shack at home. Devices like this attempt
to solve that problem. This one happens to be
adjustable, and you could, say, adjust it when a longish
cable runs to your linear. Key up the linear, adjust
the supply so that the linear sees the desired value
on its end (13.8V or whatever).
https://www.amazon.ca/MegaWatt-S-350.../dp/B00JZBE97U
OK, so that would probably work with a 30 amp resistive
load (say, four incandescent car headlights on hi beam).
It's harder to say, what would make a good DC power source
for an inductive load.
I tried looking in the transformer catalog, and couldn't
find just the right 18V transformer, to build a classic
supply. This would be the "linear supply approach", different
than SMPS, wasteful and inefficient if used all day long,
as some parts of the regulation circuitry kick off
tons of heat. Even the diode bridge rectifier can get
scalding hot when used at high load, such that the
diodes have to be mounted on a heatsink, as well as
the regulating circuitry. The heatsinks are measured in
"square inches of active area", and you might end up
needing a "cubic foot of fins for cooling" if the load
is high enough.
Like the SMPS, the price of implementation goes up
with the ratings. The linear might withstand more
abuses (such as the motor start transient), as the plus
for that approach. But it will be inefficient and
kick off heat, whereas the SMPS can be "80+ efficient"
and "active PFC", properties the power supplying company
likes. At these kinds of loads, some SMPS are
efficient enough, they barely need a cooling fan
to keep the smaller heatsink inside, cool.
Each approach has issues:
SMPS - unknown tolerance to motor startup current
- not typically rated for inductive loads, won't
say "sure, go ahead, put an X ampere motor on here".
- a solution that (nominally) lasts forever, perhaps
an inductive spike could damage it (I did that once,
that's how I know).
- a specialized SMPS could likely deal with any load
type, but at a guess, the designer would charge a fortune
for this ("because they could"). It's the average SMPS
which is not ready for all load types.
Linear - potentially more tolerant of loading types
- might support an inductive load
- something you build yourself, pretty simple hookup.
- a solution that (nominally) lasts forever
- inefficient, at these power levels "many chunks of iron".
Maybe 80 pounds weight.
Solar/battery - battery as reservoir, is very tolerant of load types.
Wonderful. Already I've gone to heaven. Don't
forget the fuses! Batteries take no prisoners.
A friend of mine got burned once, by not following
the rules ("do not wear jewelery near auto batteries").
His metal watch strap, the link pattern was burned
into his wrist! Cover the battery terminals with
insulators over top.
- can hook together using purchasable subassemblies.
I was even able to buy clamp style terminals to fit
around the round battery terminals, to cable up a
battery.
- can be charged by a wall powered charger if needed,
and this could even be used as a substitute for the
solar panel. You could hook up a "smart charger", if
you could get one that starts the charge cycle again
at 11.8V. (So the battery cannot be run flat and damaged
by neglect over long periods of non-usage.)
- big minus, replacement of battery at regular intervals,
as a function of how well the charging method makes
the best use of battery chemistry (sulphation).
- for automotive batteries, try not to use more than
25% of the amp-hour rating - the batteries last longer
if the discharge is shallow. To pump up a tire, you
might not have too much trouble with this. To pump up
all the tires at an auto rally, maybe not. Do the
maths, once you've measured the pump and fully
understand the loading it presents. More expensive deep
discharge batteries are also available.
For automotive batteries, you can find info on this site,
to understand some of the requirements, and then compare
what it says here, to what a smart charger is doing.
https://batteryuniversity.com/learn/...d_acid_battery
The battery plus smart charger looks like the winner to me,
mainly because (barring accidents involving car batteries),
it's a pretty simple setup. It can't provide power forever,
like the other two solutions could. If you keep a multimeter
handy to it, you can use that for determining current state
of charge (assumes maintenance-free battery where you can't
access the fluid with a hydrometer).
At work, each company parking lot had a battery "on wheels",
used for jump-starting cars with dead batteries. The device
was kept at the security desk, and could be signed out. And
that seemed to work pretty well. I didn't hear too many stories
about "hmmph, went to use it and someone forgot to charge
the battery". That way, we didn't need towing trucks
prowling the lot, looking for customers :-)
Paul