Posted by AEM on September 25, 2004, 12:49 pm
recently set up a very small solar power installation at home (until then
we where absolutely without electricity) consisting of a 50W solar panel and
a 12V 125Ah battery. When I was browsing prices before buying this I noticed
(not for the first time) that wherever I turn they tell me that there is
NO-Way I can buy some batteries noe and then add some more later on when I
got more money. So I started thinking , and came up with the simplest
solution one can imagine. But the people I've asked so far could not see
anything wrong with my theory witch is this:
(diagram at:
http://www.new-haven.org/~sido/sidoblog/2004/09/using-old-batteries-with-new-not.html
)
I can't see any way these two battery-banks could affect each-others
performance with this setup, if somebody can prove otherwise please let me
know.
--
Do What Thou Wilt Shall Be The Whole Of The Law.
Posted by Vaughn on September 25, 2004, 1:08 pm
> (diagram at:
http://www.new-haven.org/~sido/sidoblog/2004/09/using-old-batteries-with-new-not.html
)
> I can't see any way these two battery-banks could affect each-others
> performance with this setup, if somebody can prove otherwise please let me
> know.
Yes, I think that will work, and I think you only need the two diodes on
the positive side of the batteries, and you have to use special Schottky Diodes
or you will lose too much voltage.
There are some problems with your setup:
1) There is no way to distribute the current from your two charging sources to
keep both batteries charged. If there is wind and no light you charge one
battery and if there is sunlight and no wind you charge the other battery.
2) There will always be some voltage and power loss from the diode drop and the
resistance of the extra equipment.
3) You have added complexity to something that should be simple.
To find out more, do a Google search on something like "battery isolation
diodes". Diode battery isolation is nothing new or uncommon, so there is a
world of information out there.
Vaughn
Posted by DJ on September 25, 2004, 10:35 pm
>>http://www.new-haven.org/~sido/sidoblog/2004/09/using-old-batteries-with-new-not.html
> Yes, I think that will work, and I think you only need the two diodes on
> the positive side of the batteries
I agree there. Ground is ground.
> 1) There is no way to distribute the current from your two charging sources to
> keep both batteries charged. If there is wind and no light you charge one
> battery and if there is sunlight and no wind you charge the other battery.
How about putting both of the charging sources into *one* buss, and
from there, through diodes, to the battery banks? That would work.
There would usually be voltage, and the batteries wouldn't care from
what. True, one might end up overpowering the other, the windmill
probably overpowering the PV, but, then, if your biggest problem is
that you have too much electricity coming in... ;-)
Tricky thing might be needing to put load diversion on *BOTH* battery
banks just in case, though, as some windmill types need it...
DJ
Posted by ptaylor on September 26, 2004, 6:09 am
AEM wrote:
> recently set up a very small solar power installation at home (until then
> we where absolutely without electricity) consisting of a 50W solar panel and
> a 12V 125Ah battery. When I was browsing prices before buying this I noticed
> (not for the first time) that wherever I turn they tell me that there is
> NO-Way I can buy some batteries noe and then add some more later on when I
> got more money. So I started thinking , and came up with the simplest
> solution one can imagine. But the people I've asked so far could not see
> anything wrong with my theory witch is this:
>
> (diagram at:
>
http://www.new-haven.org/~sido/sidoblog/2004/09/using-old-batteries-with-new-not.html
)
>
> I can't see any way these two battery-banks could affect each-others
> performance with this setup, if somebody can prove otherwise please let me
> know.
>
>
I did something similar with a bank of random smaller batteries of
different sizes,and ages(largest is 17AH.) and a few diodes.
I found that a bridge rectifier has everything you need for 2
batteries,just hook it up.
connect all gnds together.(common gnd)
The two ~ (ac) terminals go to the the + of each batt.
The + of the rectifer is your output(to loads),and the - is your charge
input (solar,wind). For something large like car batteries,you'd need a
large bridge rectifier,But you could make one out of 4 large diodes. (1
pair per batt)
One small drawback is that the diodes will drop a bit of voltage,so your
output will be a bit lower than the batteries (about 0.7V for most
diodes) I actually used a bunch of high current schottky/fast
diodes,which have about half the voltage drop..(~0.3V)
Posted by Bill Darden on September 27, 2004, 3:01 am
You will definately require a diode isolator; otherwise, you will have
charging problems with different batteries, especially if they are
different plate formulations and capcities. The key will be to apply
the correct temperature compensated charging voltages, as recommended
by the battery manufacturers, directly across the battery posts. For
more information on increasing battery capacity, please see Section
7.3 in the Car and Deep Cycle Battery FAQ on www.batteryfaq.org.
Kindest regards,
BiLL......
wrote:
>recently set up a very small solar power installation at home (until then
>we where absolutely without electricity) consisting of a 50W solar panel and
>a 12V 125Ah battery. When I was browsing prices before buying this I noticed
>(not for the first time) that wherever I turn they tell me that there is
>NO-Way I can buy some batteries noe and then add some more later on when I
>got more money. So I started thinking , and came up with the simplest
>solution one can imagine. But the people I've asked so far could not see
>anything wrong with my theory witch is this:
>(diagram at:
>http://www.new-haven.org/~sido/sidoblog/2004/09/using-old-batteries-with-new-not.html
)
>I can't see any way these two battery-banks could affect each-others
>performance with this setup, if somebody can prove otherwise please let me
>know.