DC power UPS?

[daurtanyn]

All good thoughts and comments, RAL.
 
Since I already have the power supply, the new cost for me is the OpenUPS and the fuse block.  Both total about $150.  The ebay option you found was gone by the time I read your post.   Another was available for about $80.   So, in cost, the Altronix AL600ULPD4 is a clear winner, especially used.
 
Still, I have a history of paying premiums for a long tail payback and operational versatility.   The OpenUPS can be configured to appropriatly charge several battery technologies.   SLA being the default, of course.  It can also charge and monitor six cells at once.
 
The tossup, in my mind, is the Li Iron Phosphate battery vs the Sealed Lead Acid battery.  I have found another, cheaper, source than the one I posted, from WholesaleBatteries.  It's still four times more than the SLA battery you suggested, but it should last 4 times longer AND it doesn't have issues with being fully discharged. 
 
All food for thought.

 

[RAL]

Ok, if you already have some of the components, that changes things a bit.
 
Keep an eye out on that auction.  Items that don't sell often get re-listed pretty quickly. 
 
Another alternative would be to buy an AL600ULX, like this one, that doesn't contain the PD8 distribution board, which can be purchased separately.
And here's another used AL600ULX for even less.
 
One thing to consider about the Li Ion batteries vs SLA.   Battery lives are usually measured in terms of charge-discharge cycles.  Li ion batteries certainly win over SLA in this respect and Li ion batteries are great for an application like a laptop, where you charge and discharge it every day.
 
But in this application, the battery will almost never go through a charge-discharge cycle.  It will be charged all the time, and discharged maybe just a few times a year (unless you live somewhere where the electric distribution system is really crappy!).
 
An SLA battery will last 3-5 years in this type of environment.  It will reach the end of its service life even if it isn't discharged all that much.  But so will a Li ion battery.  Like an SLA battery, they also have a shelf life where they will wear out even though they aren't being constantly charged and discharged.  So I wouldn't spend the money for a Li ion battery for this application even at just 2 times the cost of SLA.  It just isn't going to pay for itself.
 
You're right that SLA batteries don't like to be fully discharged - that shortens their life.  But there is an easy, cheap solution to that, which is a low battery cut-off.  It will disconnect the load from the battery when the output voltage drops to around 10.5V, which is the point where an SLA battery has exhausted most, but not all of its charge.  This greatly extends the useful life of the battery.  Altronix makes those, too - it is called the Altronix LPD.
 
Hope this helps.

 

[pete_c]

Keep an eye out on that auction.  Items that don't sell often get re-listed pretty quickly.

 
Historically I have asked sellers in similiar unsold item auctions to post items again in a timely sync'd manner for private sale of sorts on Ebay.
 
If they really want to sell the item; then they will sell it reasonably.  If they are just fishing for sold pricing then typically they may not.
 
Off on a tangent....
 
Here I utilize both APC and Cyberpower UPS's.  I personally liked the best bang for the buck Cyberpower UPS's.  Well one day cleared off a shelf of "on sale" Cyberpower UPS's at Frys.  Commercially (at work stuff) has always been APC.

 

[RAL]

Pete -
 
Agreed on the APC UPSs.  At work, we use APC Smart UPS models that put out a pure sine wave.  Nice, but expensive. They last for years and run well.  On the other hand. the consumer models don't seem to live up to that same level of quality.  But still better than other inexpensive brands that I've tried. You get what you pay for.

 

[pete_c]

You get what you pay for.

 
Yup.  :hesaid:

 

[NeverDie]

An SLA battery will last 3-5 years in this type of environment. 

As a practical matter, how do you decide when to replace a backup battery?  By the calendar?  Or, are there good ways to monitor the health of a battery "in situ"?  Or do you have to periodically pull it out and run a discharge test on it to really know?  Obviously, the worst time to find out you have a bad battery is when it fails prematurely during a power failure, which is what typically happens (without prior warning) with the SLA backup battery in my alarm panel.  I'd like to find a better way.

 

[pete_c]

Or do you have to periodically pull it out and run a discharge test on it to really know?

 
Truly en vivo here with the OPII batteries (primary / secondary panel). 
 
I say this; but in practice its really live when I happen to be there during a power failure and paying attention.
.
I do see a number value of the battery with the OPII panel PC Access program and I get a voltage reading with the Homeseer OPII plugin.  The values never change and typically I just change the batteries every 3 years.
 
I've switched over to higher mAh hour "better made" and higher priced batteries that promise more time and a longer replacement time spans.  Don't know if I am getting what I paid for as I still just change the batteries every three years.
 
Really then the best test is to use the actual load and unplug the HV from the device and let it run on the LV load until it fails.
 
In a commercial environment the batteries were just replaced after a time frame; no matter what. (fire alarm stuff).  Real DR testing was done once a year.  Noticed many places while documenting stuff like DR never really practice it which is a bad thing.
 
I have sat there and watched the Emerson / Liebert customer service guy test every battery in a UPS in a dedicated room of UPS's for hours checking each cell. 
 
Gets really time consuming.
 
I do question that "you get what you pay for" stuff relating to residential "computer" / automation stuff these days.
 
I mean many folks purchase a residential Linksys combo router and always use the name Cisco when referring to it when in fact it is mostly a LInksys router and OS with just the name Cisco imprinted on the OS GUI and box or they purchase a GE light switch based on a TV commercial and the GE logo these days.

 

[NeverDie]

Really then the best test is to use the actual load and unplug the HV from the device and let it run on the LV load until it fails.

 
Makes sense.  So, is this something best done by a manual test trigger and the measurement of time-to-failure is automated?  e.g. on my rechargeable battery charger for non-SLA AA, AAA, 18650, etc. rechargeable batteries, there's a function where it measures battery capacity by draining a battery to failure at a constant current rate while measuring the watt-hours and amp-hours to failure (which it then displays), and afterward it automatically recharges the battery so it'll be ready to use when I come for it.  I guess  an alarm panel's SLA battery tester could measure those metrics too for a trendline, as well as the actual clock time from the moment  of cut-over to battery power to the moment of battery "failure" (10.5v according to RAL).
 
Is that the "best practice"?  I haven't looked hard, but I haven't noticed any battery chargers/monitors  that actually do that for alarm panel SLA batteries.
 
For all I know there may be a lot of different ways of testing.  So, what I'm really wanting to know is simply:  1. what sort of gear might best  automate the process for accurate in vivo battery capacity testing, and 2. if sharp folks like you all are actually doing it that way.  I don't want to be like the Emerson guy (above) who spends hours manually testing batteries.

 

[pete_c]

Makes sense.  So, is this something best done by a manual test trigger and the measurement of time-to-failure is automated?

 
From personal experience the best guess automated "time-to-failure" on my Cyberpower UPS's never is right.  (nor was the APC run times).
 
Where as this was on a little digital display on the UPS or a software connection.  It does though provide me with a warm and fuzzy. 
 
Best though for me is just to use a watch with a second hand.  I don't do that today. (must be an age thing these days).
 
Relating to the OPII I just watch the LED Keypad. When its dead then I know the battery backup is also dead.  Way back a long time ago did do that power draw adding up of all of the power sensors stuff.  You can also do little live measurements with an Amp meter with or without a load.  Well then you have to start to play with numbers, loads, "time to death of battery" to get some good numbers.  My one VOM has a serial port on it.  Some one wrote a quickie script once on Homeseer so that you can save the numbers over to a CSV file and even write them to a database and make a graph.  I used the serial port on it once for just looking at the HV values of my home.  It never really stayed exact.
 
Its a PITA but even if you do a little of this and that rather than nothing it is better.
 
I have done it for my automobile (checking amperage draw) when trying to figure out what exactly was drawing power from the battery when the car was off.
 
Last new car, 3 battery replacements in first year all due to a "firmware" glitch in the the bluetooth module sucking the life out my car battery because it worked too well.
 
Just relating to the carpc / automation in a car / music stuff that whole battery "time of death" thing is really up there.  Really with fancy power supplies with very granular displays, tank battery stuff, et al.  There if you have no battery or kill the battery then you will have some issues. 

 

[RAL]

A good way to test SLA battery life is through a conductance test.  Conductance is the reciprocal of resistance.  Elk sells a battery test meter that does this, though it's on the expensive side - at least for occasional use by a non-professional like me.
 
As a battery ages, it's conductance changes and when it gets to a certain point, it's time to replace the battery.  This nice part about this test is that it can be done in a few seconds, without discharging the battery, although you do have to disconnect it from the alarm panel while you do the test.
 
The bad (inconvenient) thing about this test is that there is no universal number that tells you when a battery is bad.  Every battery has a different conductance value when it is new, even batteries that are of the same type, brand, capacity, etc.  And that initial conductance can vary quite a bit from one battery to another.
 
So in order to be able to use this test, you have to test each battery when it is new, and record its conductance.  Then, from time to time, say once or twice a year, you repeat the test and compare it to the original baseline reading.  When the conductance has dropped to around 70% of its original value, then it is time to replace the battery.
 
Unfortunately, if you didn't make an original baseline measurement when the battery was new, then the meter can't help you much.
 
In theory, the circuitry to do this could be built into a device like an alarm panel so that it could make the test automatically.  You would just have to tell it when a new battery was installed so it could reset the baseline value.   But I don't know of any panel that does this.
 
You can measure the internal resistance of a battery with a modest effort. To measure the internal resistance of a battery, you connect a resistor to create a modest test load (e.g. 1 amp) for the battery and then use a multimeter to measure a couple of voltages as well as the exact resistance of the load.  Then you do a few calculations to come up with the internal resistance of the battery, which can then be converted to conductance.
 
Here's a description of how to do it.
 
This method is not as good as what the Elk BLT meter does, but will give you something close.   But it probably is nearly as much of a PITA as simply doing a discharge test.

 

[RAL]

The ebay option you found was gone by the time I read your post. 

 
The new AL600ULPD8 has been relisted, with a lower starting bid price of just $49, and a lower buy it now price, in case anyone is interested.

 

[NeverDie]

Thanks, guys.
 
By the way, something related that I've been wondering.  When an SAL battery claims its capacity is, say, 5Ah, is that measured from full-charge down to 10.5V, or from full-charge down to a flatline 0V?  Obviously, the 10.5V number is the number that's of practical use, not the 0V number, but the marketing department probably wants to toute the bigger number.
 
Also, when it comes to finding a "qualilty" SAL battery, what does one even look for?  It's literally a sealed black box.  Also,  since the things are so dense and heavy, who is going to pay the cost to ship it back for a warranty claim unless it's DOA?  It seems like an invitation for  manufacturers (especially those from you know where) to sell junk.  Which brands sell "the good stuff"?

 

[pete_c]

Here recently (last 5 years) went to purchasing Power-Sonic batteries.  Not sure though if I am getting my monies worth.  They work.
 
Maybe I like the blue cases?
 

 
I do not work for this company nor endorsing this company; just a statement relating to what I am currently using.
 
Power-Sonic

 

[RAL]

The battery capacity for an SLA battery is usually measured using a load of C/20, down to the point where the battery reaches the cutoff voltage (usually around 10.5 volts, but I've also seen other values of 10.x).   So, for a 7Ah battery, that would mean drawing just 350mA over a 20 hour period in order to obtain the full capacity. 
 
If you discharge the battery at a higher rate than C/20, you will lose some of the capacity.  For example, discharging at a rate of C/10 will reduce the capacity by about 5%, and discharging at a rate of C/1 will reduce it by about 40%.
 
Power Sonic has a technical manual that tells you more than you probably care to know about batteries.  It has lots of good information on battery discharge characteristics.
 
Amstron is a top quality brand for SLA batteries.  I've had good luck with Power Sonic, which tend to be a little less expensive than Amstron.  The least expensive often seem to be UPG batteries.  I don't have much experience with them to be able to say whether they are any good or not.

 

[pete_c]

Geez thinking about what Verizon used for the ONT UPS relating to OP.
 
Thinking it was Verizon branded.  Curious though what MFG is being used.
 
Originally the tech was going to install it in the master bedroom.  I moved the installation to a walk in closet. 
 
I almost purchased a Rasberry Pi battery backup a couple of days ago. 
 
Did purchase an RTC clock for the Raspberry Pi for playing.
 

 
I like the newer pico PSU's today with the monitoring voltage stuff on them.  The last 4 boxes that I have built are all using pico psu's.  The two carpc's today are using most current versions of pico psu's and they have worked fine.
 
 
 
# Monitor input and output voltage levels:
$ ./m4 -diag loop
-- 1262486581
VIN: 11.34
IGN: 11.23
33V: 3.33
5V: 5.06
12V: 12.18
TEMP: +28
-- 1262486582
VIN: 11.34
IGN: 11.34
...

# Display your PSU's configuration:
$ ./m4 -config
IGN_HIGH: 5.84 V
IGN_LOW: 5.50 V
IGN_DBC: 500 ms
PSU_DELAY: 1 sec
...

# Check or update specific variables:
$ ./m4 -config VIN_MIN_ON
9.55 V
$ ./m4 -config VIN_MIN_ON 10.0
$ ./m4 -config VIN_MIN_ON
9.99 V
 
A pico UPS (tiny thing).
 
PicoUPS-120 12V DC micro UPS system / battery backup system optimized for low V(In)

- 12V DC micro UPS system
- 12V DC battery backup
- Works in 12V vehicles
- Build-in 1.2A fast charger
- Handles up to 10A (12A peak)
- Low loss, "diode OR" MOSFETs
- Quarter Brick (58x36x20mm)
- manufactured by mini-box.com
- ROHS compliant
 

 
 
Currently using Meanwell DIN mounted 120VAC to 12VDC power supplies for my LED lighting.  I noticed a matching DIN mounted power supply with battery charger.  Thinking of giving one of these a try.  Well I could just stick on the DIN rail configured.
 
Thinking DIN PS/UPS's were mentioned this earlier in the post.
 
DRC-60 Series - 60W Single Output Power Supply with Battery Charger (UPS Function)
 

Features

    Universal AC input
    Secondary output for battery charging (UPS Function)
    Battery low protection/Battery polarity protection by fuse
    Alarm signal for AC OK and Battery low
    Can be installed on DIN rail TS-35/7.5 or 15
    Short circuit, over load, over voltage protections
    Cooling by free air convection
    LED indicator for power on
    3 years warranty