Car battery as UPS backup

[MarianB]

Greetings.

As the title says i would like to know your opinions about using car batteryes as UPS backup, one would think that they would provide more uptime for the UPS cus of theyr lager capacity but i have heard some saying that even a new car battery after a few deep discharging cycles will be good to throw away cus they die, is it true? and if so then can u please explain what happens to it?

Any thoughts on this will be much appreciated.
All the best.
Marian.

 

[Mooly]

Car batteries aren't really designed for "cyclic" use that involves deep discharge. Deep discharge causes sulphation of the plates and that reduces capacity which is hard to recover from unless you recharge immediately and with sufficient (say 16 volts) voltage to break down the sulphation. That over voltage needs careful monitoring and controlling.

I believe there are "leisure" batteries (caravanning and boating) available that are more suited to this cyclic deep discharge treatment.

Having said all that... is your UPS really going to be called on to run that often and to discharge a battery that deeply. Given the relatively low cost of many car batteries it might be worth considering them anyway.

 

[MarianB]

First of all thank you for the interest on my topik and for the detailed explanations.

Now, to be more specific, i have operated my UPS for a few months now with a car battery but it was near it's end when i removed it from the car, at the beginning it has held up real good, all i need from it are maximum 5 minutes to allow me to save work and turn of the PC, but now it really needs changing and i need to know what i can do before a decide what to buy. My UPS has a 50Hz power transformer, i did not got to measure it's efficiency but o think it is rather low cus even with it's new battery it has discharged it real fast, and when the line drops the transformer buzzes up real loud-too loud... i intend to modify it to a ferrite core transformer to boost up the eff, that intails loads of stuff but i am not worried about that, but what i do need to know is that with all modifications a good car battery could hold up, as i sayd i only need a few mins out of it ( saldy in my area the power drops often... too often ) so maybe it will not get to discharge that deep. It has to power a desktop PC main unit and an LCD display. I will measure theyr combined power consumption and post it here as soon as the time allows it.

As i understand from what you sayd there is already a reason for the battery fail, the charging circuit of the UPS maintains the voltage to a stable 13,8V, even if i charge the battery from a external charger, the UPS drops it's voltage to that 13,8V, as you say i will need to provide much more voltage from the charging circuit, that can be done, no problem, all i need to know is exactly how the charger would need to act.

Again thank you.

 

[aspringv]

these may help

Some possibly helpful reading for you...

Upgrade your UPS! (with some car batteries...)
The DIY (or, if you must, "Ghetto") UPS

 

[evanc]

As mooly said if buying new get a deep cycle/marine battery.

 

[multisync]

Check out the de-sulfination thread here somewhere

 

[FoMoCo]

Something to think about:
Many UPSs only run for a few minutes at full-load before the battery poops out. What if the inverter is designed with that in mind? Could it overheat because it wasn't thermally designed for more than a few minutes use?

 

[Mooly]

Lets put some quick numbers into all this...

UPS delivering ??? 500 watts output and that's a lot for a PC to use but lets stick with that.

The battery must supply 500 watts plus losses due to convertor inefficiencies. Lets say 700 watts consumed.

60Ah battery. That 700 watts is 58 amps at 12 volts which is an hours running. So 30 minutes running is 50% discharge, 15 minutes is 25% and so on.

Put your own numbers in for your convertor and see what you get.

 

[MarianB]

@aspringv thanks for the links, i will go trough them as quick as possible.

@evanc i realise now that deep cycle batteryes are the obvious choice but there is the price issue, for example a medium 40A car battery is way cheaper than the deep cycle one at the same capacity ( less than half here in my country ) so you see it's worth trying to make do with the car one...

@FoMoCo i think you might be right, but i am not worried by that for 2 reasons: 1-I will never let the UPS's battery backup the PC more than 5 to 10 mins max, as soon as the line drops i instantly hit the shutdownd command on the PC, so in fact it only has to cope with less than 5 mins... and the 2'nd reason is that as i sayd i plan to modify completly my UPS, i will design a new converter using a ferrite core and make sure to get as high eff as possible from it, it is much work i know but i am not worried by that.

@Mooly again i am in your debt, your imput is greatly appreciated. Now i went on and checked some numbers, first of all my PC main unit is something like this:
-The power supply is a HW-ATX2.2-V1.2, i do not have much info on it, it was in a verry sorry state when i found it, just a board withouth any cables power devices heatsinks, case... etc, i have reconditioned it cus i liked it's architecture and put it in some case i had, it's performance was real nice, so i have put it in my PC and it does it's job verry well.
-The main board is Biostar G31M+;
-An InteL(R) Pentium(R) Dual CPU, E2180 @2,00GHz;
-3Gb of RAM;
-An onboard VGA;
-A Compro DTV TV-tunner;
-A Genius sound board;
-2 HDD's, one 80Gb and the other 320Gb, both SATA2 interface;
-An Assus DVD-writter;

And that's about it, along with the main unit i got an Fujitsu-Siemens LCD display 19"-wide, i have just measured about 100W of power consumption from both of them combined with the PC on idle ( well not exactly idle, a net browser and some advanced sound processing software running ), i am no gamer so i do not anticipate more than let say 200-250W of power consumption, i usually run easy stuff. Now as for my UPS's converter, it runs at a low freq and as i thought it's eff is verry low at about 50%, i am not sure what is the reason but as i sayd it does not matter, i will design a new one to make sure i get at least 70% eff ( i am hoping for more but it remaines to bee seen what and how ). Now judging from your examples and allowing for a 70% efficiency the battery would have to provide about 350W of power, let round it up to maximum 400W ( i doubt it verry much it will go any higher ) so about 33A, if i choose a 40A car battery it would be about 1 hour and 20 mins for the battery to compleetly discharge, running it for 10 min would mean about 12% discharge, and at my 5' stated would be less than 10%, am i right?

 

[kenpeter]

Never more than 13.8V for sealed absorbant glass mat (AGM)
batteries that were probably the original type in your UPS.

Batteries that are vented on top, may charge at up to 14.4V.
Any higher, you will then have to monitor and replace water.
Playing with water and acid isn't exactly my idea of fun...

Make or break your final connection well away from the battery.
There can be explosive gasses (ready to blow acid everywhere),
especially if you suspect a problem or overcharge. You do not
want any opening or closing spark to occur at the terminals of
the battery, with your hands and face in working proximity.

Car batteries should be vented outside.

-----

Most UPS have a serial port on back, that software can detect
and shutdown the PC in a controlled manner, even unattended.

 

[FoMoCo]

@FoMoCo i think you might be right, but i am not worried by that for 2 reasons: 1-I will never let the UPS's battery backup the PC more than 5 to 10 mins max, as soon as the line drops i instantly hit the shutdownd command on the PC, so in fact it only has to cope with less than 5 mins... and the 2'nd reason is that as i sayd i plan to modify completly my UPS, i will design a new converter using a ferrite core and make sure to get as high eff as possible from it, it is much work i know but i am not worried by that.

If you only need 5 minutes, why modify anything?

 

[MarianB]

because of what i have sayd earlyer, the efficiancy it's ridicoulus, actually a bit under 50%, that is unacceptable, i am sure it can do way better, also i can desighn the charging circuit exactly the way i want to.

 

[Mooly]

Your numbers sound about right. A 40Ah batteryrunning for 80 mins is 0.5Ah per minute capacity used.

I'm not 100% sure how well car batteries tolerate 24/7 float charging. If your designing your own charger then I would perhaps arrange it to cut out after say 18 hours if it applies around 13.8 volts and to cut in every 7 days or so or whenever a mains interruption occurs.

 

[MarianB]

Wouldn't it be easyer to just set up a monitoring circuit that cut's off the charger after reaching about 14,4V on the battery? this way i can ensure that the battery would not have to handle permanent charging, just recharging when needed, i could also provide some histeresis on the on/off charger thresholds if needed.

LE: @kenpeter, i keep my battery outside so it is well ventilated, also far from the UPS and any spark source, i am using heavy duty cables and conectors to get the power to the UPS.

 

[Mooly]

That doesn't work in practice.

14.4 volts or even higher voltage can be reached very quickly but the battery is still drawing substantial current that tapers down as the battery charges.

 

[MarianB]

Then the monitoring circuit can be set up to "read" the current drawn by the battery as it is charging and when it reaches a set vallue it cut's off the charger, or some combination of both current and voltage, whatever you think it is best cus i do not want to complicate stuff timing it...

 

[Mooly]

Exactly... keep it simple

Measuring the current is problematic because it varies greatly with temperature and the condition of the battery as it ages.

Applying a known voltage for a certain minimum time has to be the simplest solution.

 

[Sorento]

There's yet another reason why car batteries / starter batteries (flooded type) won't make you happy.
They need the vibrations from the road to stay healthy !
If kept stationary the electrolyte will separate over time, sulfuric acid being denser than water it will accumulate on the bottom and pure water on top. Plate corrosion is going to happen even with trickle charge.
Some float chargers apply a very high charge current (and voltage) once per day to deliberately cause gassing, the bubbles stirring the mixture. But I haven't seen any data on that yet.

 

[MarianB]

@Mooly, but as i understand if the battery draws no or little current it means it is charged... am i wrong?

If allowing your suggested solution then how do i go about the timing circuit? i think it would be too complicated to be worth it...

@payloadde, good point and yet another reason to drop the ideea but still this battery has done it's job good enough despite the UPS's low efficiency, looking at batteryes prices all i can afford of the deep cycle stuff is arround 10-12Ah capacity...

 

[Mooly]

There's yet another reason why car batteries / starter batteries (flooded type) won't make you happy.
They need the vibrations from the road to stay healthy !
If kept stationary the electrolyte will separate over time, sulfuric acid being denser than water it will accumulate on the bottom and pure water on top. Plate corrosion is going to happen even with trickle charge.
Some float chargers apply a very high charge current (and voltage) once per day to deliberately cause gassing, the bubbles stirring the mixture. But I haven't seen any data on that yet.

Interesting thought. I've never heard of that and it sounds plausible although batteries I have seen all have movement of the electrolyte when fully charged and even slight gassing occurs.

Interesting non the less...

@Mooly, but as i understand if the battery draws no or little current it means it is charged... am i wrong?

If allowing your suggested solution then how do i go about the timing circuit? i think it would be too complicated to be worth it...

@payloadde, good point and yet another reason to drop the ideea but still this battery has done it's job good enough despite the UPS's low efficiency, looking at batteryes prices all i can afford of the deep cycle stuff is arround 10-12Ah capacity...

The timing circuit you would have to design. Something like a CMOS oscillator and divider chip (4040 and 4060 chips from memory) perhaps with diode gating on the appropriate outputs to get something like the charge time required. Or a simple mechanical timer if it outside, the sort with pegs in to set the on and off times.

The only correct way to really determine the state of charge is to measure the specific gravity of the electrolyte but that's not practical.