Don and all, I work as a computer IT technician and have used older computer UPS supplies for a number of different applications. It's easy to hook up a couple of standard lead-acid batteries from Wal-Mart to these and gain extra run-time capacity than the original batteries provided.
But a couple of things I've noticed. The charging voltage the UPS provides to the batteries is a bit higher than I like and usually slow-cooks the batteries giving a life expectency of about 3 years. Also, even when lightly loaded, the run-time of a typical UPS is not very impressive. I used one of these set-ups at my repeater site but found it didn't work as well as I hoped. I reasoned that if I ran the repeater and controllers on just batteries I didn't like what happened to the controllers as the battery voltage faded into the sunset - I wanted everything, especially the controllers to completly power off at a certain voltage - so I tried a UPS. Well, at the first power outage the controller and repeater ran a couple of hours and shut down as planned but when the AC returned it never came back up. A trip to the site found that the UPS was squaking its alarm. Hitting the power-up button brought everything up for a few seconds but everything shut back down when the UPS tried to do a self-test. I guessed that the batteries had not recharged enough to pass the self test. I bypassed the UPS to get everything back on the air. I've since gone back to having the 12v batteries connected directly to my repeater power supply but I've built a simple little comparator circuit that monitors the battery voltage and if it falls below 11.5 volts it removes the power to the controllers - effectively shutting down the repeater. I've also put relays into the repeaters power amplifiers that drop the PA from full power to about 1 watt output. These relays engage when the AC goes out and greatly reduce the current draw from the batteries when the AC is off. This works much cleaner than using a UPS and the repeater has run off of batteries for DAYS instead of HOURS with this configuration. YMMV Lee --- In [email protected], Mike Morris WA6ILQ <[EMAIL PROTECTED]> wrote: > > At 09:37 AM 05/23/07, you wrote: > >Hello all, > > > >I'm in the process of rebuilding one of my repeaters and I'm looking > >to place a UPS in line for backup. > > > >I'm looking for information on make and models - what's good and > >what to stay away from. The VA range is 1300 to 1500, and I'm > >willing to spend no more than $200.00. > > > >Thoughts and comments appreciated. > > > >TIA, > > > >Don, KD9PT > > For a cheap UPS talk to any of your friend in the IT world. There > are enough uninformed "technicians" (and it use that term generously) > in the IT universe that have no idea as to which end of a screwdriver is > the handle... > When a UPS dies (at about the 3 years point) they just throw it away > and buy a new one. In fact I just recently "re-batteried" a several APC > "Back-UPS PRO 1000" units for a client. > > Three comments of warning about APC: > 1) They pass all the AC line junk through the unit (via relay contacts) > until there is a brownout (i.e. low AC voltage) or an outage (no AC voltage). > Then the inverter fires up, and the relay switches over. You WILL have an > outage as long as one to two tenths of a second. I've seen desktops and > servers reboot under those conditions (usually mans that the computer > power supply has dying caps, which is a whole 'nother problem). > 2) APC runs the batteries hard. Plan on replacing them every 3 years. > 3) They over-rate their units. Rarely does any APC give you over > 25-30 minutes of run-time. They are designed to provide graceful > shutdown time, not to ride out an outage of any length. Graceful > shutdown means that the server power cord is plugged into the UPS, > and a serial port is plugged into the DB9 jack in the back of the > UPS (via a special cable they sell, one that has a transistorized > circuit buried in one of the molded rubber shells). When the power > fails the UPS sends a signal to the server over the serial port > announcing power fail (they simply change the CD signal from > active to inactive). > In a properly designed network the server messages all the users > and gives them 5 minutes to save the current work and log out. > Then the server shuts down before the battery runs out (or in > large data centers the generator starts up). > > The biggest problem with using an IT UPS in a solid state > repeater environment is efficiency. You are running > AC to 12 or 24 or 48V DC and then back to AC into the > repeater power supply which makes 12vDC. > > The better way is to simply use the batteries directly on > the repeater - just use a Absorbed Glass Mat battery > (also known as an AGM battery) and float it across the > repeater power buss. The efficiency goes up, if you are > paying the AC power bill your wallet is happier, and you > get zero switchover time. > > A compromise design that I've seen has the repeater > power supply and the backup battery diode OR'd to > the repeater itself with 75amp stud-mount diodes > (from a dead fire truck alternator), with an IOTA brand > charger on the backup battery (which was a bank of > glass-cased Exide telephone central office style > batteries - 5 gallons of electrolyte per cell) > > So my recommendation, if you have a solid state repeater, > is an AGM battery with a IOTA power supply / charger. > Don't save money on the charger - a good one will last, > a cheap one won't and may take your battery bank with it. > A good IOTA power supply / charger is NOT a waste of > money and a cheap one is false economy. > > Mike WA6ILQ >
