Replacing UPS Batteries with Marine/RV Batteries

[dontosaw]

I had the UPS working fine for years with the two 33 Ah batteries, so I don't think the capacity of the batteries is causing my current issue. I think something has failed in the electronics and that is what is causing the UPS to slowly drain the batteries. My line power is good.

 

[jim hardy]

I had the UPS working fine for years with the two 33 Ah batteries,

Your logic is good.

All i can suggest is to measure the voltage it puts across the batteries while charging
and

if you can find the charging traces on the circuit board, they have to go through something that controls the current.
Probably also a low ohm resistor that measures the current as well.
If you're lucky you'll find something not conducting. Check voltage across anything in series with battery charger .

You'd be surprised how many things get fixed by a close visual exam. Look for cracked solder joints.. I recently fixed cracked solder joint at a 1/2 watt surface mount resistor .

You need a jeweler's loupe...

www.bourns.com/ProductLine.aspx?name=pulse_power_resistors

I'm out of ideas - good luck.

 

[dontosaw]

OK, both new batteries are not fully charged. Now to find a good used Smart-UPS to replace my failed unit. Anyone have any advice and/or suggestions? I would imagine that all the APC 1500 VA units have 24 volt battery input, but will have to confirm.

Don

Correction. I meant to say both batteries are now fully charged. Used external charger. I will do autopsy on my UPS. If I don't find anything obvious, I will replace it. Will update what I find.

Don

 

[tfr000]

A few hints on gel-cell battery salvage, for what it's worth. (I salvage the gel-cell batteries from some of the throwaway UPS's at my work. Yes, they throw them away when they get old, rather than replace the batteries.) This is for 12V batteries, there are some 6V out there.
I first measure their open-circuit voltage. It should be at least 11V. If it's less than that, it is likely the battery is damaged by over-discharge.
Next, I try charging at 13.5V at a 1 amp current limit (from a power supply). Unless it's fully charged, it should take the full 1 amp. If not, it might have an open cell. Charge until the current drops off.
Next, I try applying a heavy load... maybe 3-5 ohms. Be careful here - you need a high power resistor and heavy wiring. It's easy to melt a clip-lead. Damaged batteries will not be able to supply the current. A good battery should have no trouble. Do not discharge 12V gel-cells below about 10-11V open circuit voltage or you will damage them.
If it passes all of this, it should be usable. I find about one out of three from the dead UPS's are salvageable.
BTW, I have tried several methods of reviving dead gel-cells. I have gotten it to work once in many attempts. It's generally not worth bothering.

 

[jim hardy]

Next, I try applying a heavy load... maybe 3-5 ohms. Be careful here - you need a high power resistor and heavy wiring.

an automobile headlamp would be handy for that test

 

[dontosaw]

Thanks for all the great tips and info. This forum is really great.

I opened up my old UPS and didn't see anything obvious. Going to give it to a friend who sometimes is able to fix electronics, provided the problem is something easy and cheap to fix, like a failed diode or resister. In the case of my old UPS I'm thinking it is in the control circuitry for the charging circuit and probably not a cheap fix, but if he wants to try, have at it.

I've been looking online. An APC Smart-UPS that puts out a real sine wave would be great, but they are just too expensive (about $500 new and at least $200 for a used one likely to be in good working condition). I found what appears to be a very good value in a UPS that puts out a sine wave from CyberPower. It is the
CP1350PFCLCD. I found it at a local retailer for $180.
http://www.cyberpowersystems.com/products/ups-systems/pfc-sinewave-series/CP1350PFCLCD.html

If anyone has an opinion on this unit, let me know. Looks like it takes 2 12 volt batteries in series for 24v, so will be able to use with my batteries, but will do in a way that does not leave a permanent sign in case it fails during warranty.

Don

 

[dontosaw]

Update: Bought a CyberPower CP1350PFCLCD UPS. This UPS puts out a pure sine wave with 810 watt capacity. So far, I am impressed with the features and performance of this unit. Excellent value for $180 I think. Also, unlike some other brands, CyberPower will let you use their professional (Business Edition) software with this consumer grade unit.

Going to wait to make sure everything continues working well for another day before I attempt to connect external batteries. I installed a 30 amp inline fuse on the plus side near the batteries for safety as suggested. Thank you.

Don

 

[dontosaw]

OK, I wired the large external batteries to the new UPS and everything appears to be working fine. However, during a test in which I cut the power using the GFCI test button, the fuse and fuse holder got pretty hot. All the rest of the wiring stayed cool, but the fuse got pretty hot. Is that a sign that something is wrong with the fuse holder or that 30 amps might not be a big enough fuse? Since the UPS has to step up the voltage quite a lot (24v to 120v), perhaps there is enough current when drawing on the batteries to get the fuse hot. The load (PC) was drawing about 230 watts I think. Will a 40 amp fuse still provide adequate protection? Thoughts?

Thanks,
Don

 

[jim hardy]

230 watts at 12 volts is ~19 amps
Divide that by inverter efficiency , maybe .85, and we're up to 22.5
Fuses do generate a bit of heat, it's how they melt...
They depend on being cooled by convection and conduction along their connecting wires

So, does your installation allow free air movement around fuseholder and the first several inches of adjacent wire?
A 40 amp fuseholder will probably have more area and bigger wires
You could plug a 30 fuse into one? Or go ahead and use a 40 if your wires are at least #10 and in free air
http://www.houwire.com/products/technical/article310_17.html

Lastly, inverters may draw a "peaky" current waveform which generates more heat in wires than a DC current of same average value.
Comparing readings between a true RMS ammeter and an average responding one should show that,
While i don't think you need that level of 'esoteric' for your installation, it wouldn't be the first time that phenomenon has bit..
"Crest Factor" would the term to research.interesting project, thanks for sharing

 

[dontosaw]

Thanks Jim,
Yes, fuse holder is in open air. I'm using a 40 amp, as I'm thinking that the heat must indicate that the fuse is resisting the current flow too much, but maybe not. I also replaced the fuse holder, as I wanted to be sure that wasn't the issue. Wires to and from batteries and UPS are 12 gauge. The wire connecting the two batteries in series is 10 gauge. The fuse is near the batters on the red wire.

I conducted a test by cutting the supply power at the GFCI outlet with the test button. The test lasted about 50 minutes before the UPS software put the PC in hybernation. The fuse holder got warm, but not too hot. The inverter in the UPS got pretty hot as well. Lots of waste heat with the process it seams.

It seems that the UPS is confused by the larger batteries. After about 50 minutes, the UPS interface software shows the batteries at 89% and that they are charging. The internal charger puts out 1 amp, so it is going to probably take a day before the batteries show full charge. The interesting thing, is that for the first half hour or so, the UPS monitoring software showed my batteries at 100% charge, which I know is not true, so am worried that the UPS might not charge the batteries fully. Might have to disconnect them and put them on my smart charger to check the state of charge once the UPS is done charging them.

If I hook up my 3 amp smart charger to the batteries, can I just turn everything off and pull the fuse, then hook the 12 volt charger up to one of the two batteries while it is still connected in series to the other battery, or should I disconnect all wires from the batteries to be safe?

Don

 

[dontosaw]

Update: According to the interface software, the batteries are fully charged. The voltage at the batteries is 26.4 volts, which would be 13.2 volts each. I think that is consistent with fully charged batteries. So at this point I think I am set.

I got worried right after the discharge test yesterday, as the interface software first indicated 89% charge right after the the system was put back on utility power, then some time later went down to 86%. I think the bigger batteries throw some of the monitoring circuitry for a loop.I'm sure they behave a little differently than the 7 Ah batteries that were supplied with the unit.

Don

 

[jim hardy]

I got worried right after the discharge test yesterday, as the interface software first indicated 89% charge right after the the system was put back on utility power, then some time later went down to 86%. I think the bigger batteries throw some of the monitoring circuitry for a loop.I'm sure they behave a little differently than the 7 Ah batteries that were supplied with the unit.

Hard to guess what the software does. Some systems count coulombs , bigger batteries would confuse one of those.

You said the unit was warm after discharge test ?
My guess is the computer assumes the battery is same temperature as itself
so expecting a warm battery that would have fewer volts for a given state of charge
and getting volts from a nice cool battery outside the enclosure
it reports higher state of charge .
As the unit cools down and computer's temperature approaches that of your batteries, its estimated state of charge gets better.
http://www.solar-electric.com/deep-cycle-battery-faq.html

Observe 3% charge is only about 30 millivolts,
maybe 2 degF on that chart ?

Check my thinking - do things move the right direction ? I'm notorious for "sideways thinking" .

 

[dontosaw]

Thanks for the info and links Jim.

I have some additional info. I turned everything off, disconnected the batteries and charged them separately with my smart charger. The charger indicated 95% charge for each of them. I also put the two 7 Ah batteries that were supplied with the UPS on the charger and each of them indicated 95% charge. So, it appears that the UPS charges the batteries to 95%.

Does anyone know whether having the batteries maintained long term by the UPS at 95% charge will adversely affect their lifespan? If so, I can put them on my external smart charger every few months. Other than that I think I'm good.

Thanks,
Don

 

[dontosaw]

Update: Based on the voltage at the batteries while connected to the UPS, I know believe that the 1 amp charging circuit trickle charges the batteries at 2.3 volts/cell. I measure 27.6 volts at the batteries and this seems to be consistent. I also notice that the case of the UPS is warm in one place at all times when it is on standby and indicates that the batteries are fully charged. I am guessing this heat is coming for the charger. So, even though the UPS indicated that the batteries were fully charged at 95%, I'm guessing that it continues to trick charge them and that they do indeed reach a full charge. In fact, I thinking that overcharging is more of a concern than undercharging.

In any case, the setup seems to be working quite nicely and I have like 1 hour and 45 minutes backup without draining the batteries down too far. I have a 40 amp fuse on the red wire near the batteries and I have the UPS sitting on a fairly thick aluminum plate to help spread the heat during a long power interruption as it gets pretty hot.

Don

 

[jim hardy]

I'd just have to try placing a heating pad under the battery to see if warming Mr battery made Mr UPS change his opinion about state of charge..

 

[Mohammed Sahib]

Your post is educational. Thanks, Can you elaborate on how to make the UPS re-calibrate and recognize the larger capacity of new batteries?

 

[rbelli1]

Sometimes it is as simple as fully charging and then fully discharging under a steady load. Sometimes you need to enter a special calibration mode. You will need to call customer support for that option. Sometimes it is impossible.

BoB

 

[Mohammed Sahib]

Sometimes it is as simple as fully charging and then fully discharging under a steady load. Sometimes you need to enter a special calibration mode. You will need to call customer support for that option. Sometimes it is impossible.

BoB

Is re-calibrating essential to ensure correct handling of the batteries by the UPS or is it just for correct reporting of remaining time and charge to the computer?

 

[rbelli1]

I guess that would depend on the model. I had one that would work fine but for the entire battery powered duration the low battery alarm would sound. The only solution was to disconnect the buzzer.

If you are using the battery capacity measurement to properly shut down at end of discharge then you need that to be accurate. If it just a matter of making power then often you could get away without proper calibration.

BoB

 

[Mohammed Sahib]

I guess that would depend on the model. I had one that would work fine but for the entire battery powered duration the low battery alarm would sound. The only solution was to disconnect the buzzer.

If you are using the battery capacity measurement to properly shut down at end of discharge then you need that to be accurate. If it just a matter of making power then often you could get away without proper calibration.

BoB

Thanks for the help. I have an APC Back-UPS RS 1500 wired to two 65 AH 12 v Kung Long VRLA batteries. I will follow your advice and disconnect it from my computer as I believe completely draining then fully charging these will take a long time.

 

[KD6ICZ]

Great thread!

I'm about to do this with my Minute Man PRO1000RT UPS. It's just over a year old but one of the two 9 Ah SLA's swelled up and are dead. I'm not surprised with the level of heat that was inside the steel box. So going external will let the batteries stay much cooler.

I have been eyeing the 35 Ah Duracell AGM's that Batteries Plus sell. However with coupons the 35 Ah's at Harbor Freight look appealing as well. For sure a fuse will be added. I will most likely go with a 40A fuse since any fuse will blow in a short circuit situation.

Another battery caught my eye a Batteries Plus. It's also a Duracell but it is speciffically made for UPS applications. Here is a link:

https://www.batteriesplus.com/battery/sla-sealed-lead-acid/duracell-ultra/12/durhr12=100

Here is what the description says:

• 92.8 Watts per cell
• Best suited for float or emergency backup applications
• 10 year design life - 36 month warranty
• Thinner & more plates than general purpose; plate structure delivers power quickly
• Flame retardant
• Duracell Ultra High Rate AGM battery 12V 92.8WPC

I was hoping some of you that are more in the know could tell me if thiss sounds like something that would be worth a few extra dollars and a bit less capacity?

 

[anorlunda]

They sure whack you on price for those sealed AGM models. Batteries plus price for the 27 AH AGM battery you linked is $99, their price for a $75 AH flooded deep discharge battery is $88. That is nearly 3:1 ratio in $/AH.

If you need the sealed battery for safety, go ahead. But if you can put a flooded battery in a battery box someplace with ventilation, it is a much better deal.

 

[KD6ICZ]

They sure whack you on price for those sealed AGM models. Batteries plus price for the 27 AH AGM battery you linked is $99, their price for a $75 AH flooded deep discharge battery is $88. That is nearly 3:1 ratio in $/AH.

If you need the sealed battery for safety, go ahead. But if you can put a flooded battery in a battery box someplace with ventilation, it is a much better deal.

Oh believe me! If I could I would! I'm in the telecom industry and would love to have flooded cells that I could maintain! The wireless company I work for has some flooded cells that are 33 years old and still going strong! Unfortunately, there is an open pilot light about 2 feet away. So AGM's are a must!

The link I provided says these are designed for the constant trickle charge and come with a 3 year warranty. If they are really better suited then it might be worth the extra money.

 

[dlgoff]

... But if you can put a flooded battery in a battery box someplace with ventilation, it is a much better deal.

Not only that, but with a good documented maintenance procedure, you can make them last longer and have proof of proper care; should one fail.

 

[dlgoff]

... would love to have flooded cells that I could maintain!

You posted too fast.

 

[rbelli1]

I usually get about 3 years of standby out of the cheap ones. I just bought good ones so in a few years I will know if they are better. The original ones lasted over 10 years.

BoB

 

[KD6ICZ]

I ended up going with the 35Ah batteries at Harbor Freight Tools. With the 20% off coupon they were $60 each. They look as good as any other Chinese made battery. At least I have a core to turn in next time I need to replace them.