Chris Harper wrote in a message to Roy J. Tellason:
 CH> On 06 Dec 96 at 21:32, Roy J. Tellason of 1:270/615
 CH> wrote to Chris Harper about :
 RJT> What will you consider "discharged"?  I've seen
 RJT> specs for that going to,  say,  10.5 volts for a 12 volt battery, 
 RJT> though I wouldn't want to make it a habit to run them down that 
 RJT> far.  Voltage can also be misleading if you're dealing with 
 RJT> surface charge at all.
 CH> The point I will call discharged, is the point in which my 
 CH> inverter will no longer produce at least 110VAC, given the 
 CH> batteries' input. 
Hmm,  I would test things to maybe see if they would operate at a little 
lower voltage.  Say,  105,  or maybe even lower.  Some switching power 
supplies have pretty wide input voltage operating limits.
 CH> I would think that would be determined by the voltage of the 
 CH> batteries. Am I mistaken?
Sounds reasonable to me.  One thing to keep in mind,  though,  is that a 
switching power supply is going to regulate its outputs properly,  and if the 
line voltage input to it is a little on the low side,  it'll draw current for 
a longer period of time to make up for it.  So your average loading will go 
*up* as the voltage drops,  drawing more current,  rather than down as a 
linear resistive load would do.
 CH> I was considering trying to find a less than 3 volt battery I 
 CH> could switch into the battery (series connect) when voltages 
 CH> get low in order to drain the other batteries further, if it's 
 CH> beneficial to. If it's not, then I can save the trouble and 
 CH> cost of the low voltage battery setup.
I wouldn't think it'd be that beneficial to take lead-acid cells all the way 
down,  if you can avoid it.  Also,  you're typically looking at 2v/cell,  so 
it'd be kind of hard to find a 3v lead-acid "cell".  Finally,  you'd have a 
different state of charge between that "extra" cell and the rest of them.  I 
always figured it'd be better to have all cells more or less tracking so that 
they're in about the same state.
 CH> BTW - do you know the average "usable capacity" (of an average 
 CH> cycle) of an average 12 volt deep cycle battery I can obtain 
 CH> easily? I need to estimate how many batteries I would need to 
 CH> have 200-300 KW available (total, not per hour;-) to be used at 
 CH> around 1.5-3KW/Hour. This should give me around five days 
 CH> storage capacity, I believe. That is, If I could pull 10KW from 
 CH> a battery before it no longer will run the inverter, I would 
 CH> need 20-30 batteries. But I have no idea how much wattage I can 
 CH> drain from an average deep cycle battery.
I think that in large part this is going to depend on your inverter,  and on 
a number of other variables,  such as what kind of loads you're running 
there.  Linear loads,  such as lighting,  will draw fairly constant current 
with a slight decline in draw as the voltage goes down.  Switching regulated 
loads,  which includes not only computer gear but also most tv sets these 
days,  will probably draw more current as the voltage goes down,  at least 
within the set's normal operating range.
You'll probably have to get at least some of the stuff you plan to use and 
try it out in various configurations to see what sort of results you get.  
Keep in mind that there are also a lot of factors that are going to affect 
the outcome,  such as temperature,  age of the battery,  design of the 
inverter,  and so forth.
 CH> I doubt any single average deep cycle battery could deliver 
 CH> 3KWH for any length of time without suffering increased wear, 
 CH> and I intend to hook several in parallel to reduce the load on 
 CH> any single battery. 
The one thing that bothers me about connecting batteries in parallel like 
that is what happens if one of them has a shorted cell?  Typically it'll pull 
the rest of them down,  for starters.  You also want parallel-connected 
batteries to be all of the same type from the same mfr. and about the same 
age.
 CH> I may have that pack run into a couple common car batteries, 
 CH> which will handle surge needs, and be replenished by the deep 
 CH> cycle pack.
I wouldn't use car batteries at all in this application.  There's no problem 
with pulling lots of power out of a deep cycle battery if you need to,  most 
of the ones I sold were "dual purpose" in their labeling.  For example,  the 
two that I have sitting here are "Exide Nautilus Gold NG-27" types,  which 
are rated at 670 MCA (Marine Cranking Amps,  rated at 32 degrees instead of 
zero degrees).  That's quite a bit of kick,  and should take care of what you 
need in that respect.
You use car batteries,  and they can also deliver surge power,  but their 
design is such that they go down,  and get charged back up right shortly 
after that.  It doesn't sound like this will happen in your application.
Of more concern regarding the idea of pulling large amounts of power out of 
your setup is whether the inverter will deal with it or not.
 CH> The car batteries would handle not only surge demends, but 
 CH> would take the main charge from the panels, as the charge would 
 CH> vary with clouds passing and such. The car batteries would act 
 CH> similar to an extremely large capacitor, in effect, to smooth 
 CH> out the load (either inbound or outbound) for the deep cycle 
 CH> batteries, which would handle the main storage.
Nope,  I wouldn't do it...
 CH> Could I just connect them all in parallel to accomplish this, 
 CH> or would I need to use a special circut to prevent the deep 
 CH> cycles from handling surges?
 CH> I was thinking that I could hook it up like this:
 CH> +++++..++L+
 CH> DDDDD  CC
 CH> -----..--L-
 CH> Where "D" is the deep cycle batteries, "C" is the car 
 CH> batteries, "L" is the load or charge, and "." is either cable 
 CH> or special circut (if needed). As the load or charge would hit 
 CH> the car batteries first, I am hoping they'd buffer the deep 
 CH> cycles.
 CH> What do you think?
I don't know what kind of a circuit would be useful when it came to isolating 
batteries,  just on the basis of me wanting to keep them separate from each 
other.  I see no reason at all to get into using car batteries for this sort 
of stuff,  and wouldn't bother with them for the most part.
The only "isolators" I know of that are out there are typically two diodes 
mounted in a heat sink and connected between your vehicle's alternator and 
the starting and "rv" batteries,  so that running accessories off the latter 
one won't keep the former one from starting the motor when you need to.  
These are ok to an extent, but you still lose some power in those diodes,  
anyhow.  And I don't see where that applies to what you want to do.
--- 
---------------