I was actually going to post this as a comment to Zmoz in his inquiries about sla's in "Is voltage an accurate measurement of remaining capacity?"
http://www.fieldlines.com/story/2003/8/20/145253/750
But, I thought I would post it here as a new fresh post and reiterate some of the comments I have made in earlier posts.
There really has been a lot of information posted here on SLA batteries, which has ranged from totally incorrect to very good. I have watched with interest to see how
how many posts actually come nearer the real world of SLA's
First, I am presently a Biomedical Electronic Technician (BMET) at one of the largest hospitals in the Midwest here in St.Louis. I also used to live off grid for 2 years in New Mexico before I moved here in 2000.
We use nearly 2000 SLA batteries in medical electronic equipment that I maintain, from small SLA's at 1.4 amp/hrs to 26 amp/hr ones. I maintain and repair a wide varity of charging systems in these systems. So, basically I have seen every SLA issue one can imagine. Bear in mind, that my SLA experience only extends to SLA's of 36 amp/hr or less in capacity, so I can not make any claims about larger ones.
At the moment I run the Otherpower IRC server on a bank of used sla's totalling over 200 amp/hrs in capacity, using anywhere from 12 volt 7 amp'ers to 12 volt 26 amp'ers, all wired in parallel. This consists of a Coleman 400 watt inverter, and a Dell 133 Latitude LM laptop running Linux kernel 2.4.20. The avg load is 1.5 amps 24/7 out of the used SLA bank.
The bank is charged from my 6 panel PV array using an ASC 12/16 controller to the SLA bank, and to my Trace C-40 , which charges my t-105 bank out in the shed.
How I determine if a used SLA is good:
When I go thru the pile of used sla's looking for good candidates to take home, I consider several factors..
- The age of the battery by date code
- The unloaded voltage should be above 13 volts (means it was charged when pulled out of service).
- Physical condition of the battery (swelling, terminal damage, etc).
- And , if i can , determine the kind of service it had seen (standby use, heavy charge/discharge use, etc.)
Then, after I gather up my good candidates, the next step is to fully charge it, using my variable 5 amp power supply set to 14.45 volts. I look at initial charge current, and insure it drops down to a nice safe level after a few minutes. After an overnight charge, I then measure the current being drawn. If any of the batts draw more than a few millamps in float state, I discard them.
Now, Im down to the load test..I use a 12 volt automotive tail light bulb (1157 I think), which draws about 1.7 - 1.8 amps. I then do a time test to see what the battery voltage is under load, depending on the amp/hr capacity of the battery.
Using these methods, I have never had a bad SLA battery taken home from the shop at work..(they are heavy, and I commute on light rail!)
Also, over the last few years, I have used a variety of charging systems to charge sla's , anywhere from small wall worts, to simple LM317 and LM723 circuits I have built, to variable power supplies, to sophiscated charge controllers.
12 volt Sla's basically like 2 charging voltages depending on your use for them. When used in a system that regulary cycles them, they like to be charged between 14.5 volts and 14.9 volts. In a system where they are used for standby service, e.g. UPS supplies, they like to be kept charged at 13.7 to about 13.9 volts. And this voltage is not that critical really, but is one of the factors in SLA charging that extends the life of the battery.
The 1 amp trickle chargers will work, but need to be kept an eye on, since they are unregulated, and tend to float the SLA voltage up to 16 volts or higher. SLA's have no tolerance for over voltage charging, due to the fact they can not outgass, and tend to swell and eventually short out over time.
Also, a note on voltage as a determintation of state of charge (SOC).
The used bank, in normal 1.5 amp draw, has never seen a voltage below 12.78. Some here have stated that 12.7 is 100 % charged, but by the time the sun hits the panels in the morning, I am still well above this level, and estimate that the bank is down to 90 % SOC. Troy has made some excellent posts in the past to why this can vary .
The last thing I want to add, is that I have never been able to fully recover a dead SLA..In my experience, once a SLA goes dead, its useful life is near an end.
TomW and I have been working on some realtime online graphs of the Otherpower IRC PV system, and have some graphs now online, with a sample rate of every ten seconds. I am presently using a Radio Shack 22-805 multimeter linked to rs-232 with custom software to generate these plots in realtime. Soon to come will be PV "IN" amps, and battery (OUT) load amps as I build the meter multiplexing circuit
Check out the current graph at
http://oneota.com/~earthsourcepowr/jimgraph.php
or on the IRC server itself at :
http://66.140.203.101/solar/data.shtml
and you will see how the controller kicks in at 14.5 volts , shuts off the panels till the voltage hits 14.0 volts, and then cycles till the sun goes down, unless of course I fire up a fan or two off of the bank ;)
Hope this helps about on SLA's . Any questions, dont hesitate to give me a yell.
JimU , The Solar powered Otherpower IRC server Admin
IRC server: otherpower.serveirc.com on port 6667 channel is #otherpower
