UPS Battery replacement help

[silkstone]

65°C (149°F) is well above the threshold that would cause 2nd degree burns to the skin. That temp does not seem reasonable to me.

The UPS should not be placed in a confined space but it does not need to be in a place that receives a "flow" of cool air ventilation. It only needs to be placed where air can freely radiate off it.

I would not trust that UPS in my home.

The UPS itself is not 65C. Just the iron core on the transformer when opened up! That is meant to get hot. Ambient here is around 30 C anyway

I switched it with my 6 month old one and it exhibits the same behavior, getting warmer in that location, but I haven't opened it up to measure the temp on the transformer as I don't want to void the warranty, but I can tell that the warmth is coming from the transformer area.
I may open up my one with 2 batteries again and add some sort of (non-flammable) insulation between the transformer and the battery compartment. Though the batteries are cool to the touch, they are closer to the transformer compartment since making the modification.

I've moved the dual battery system to my main rig for the time being anyway and I will keep it turned off at night or when not in use.

 

[E-Bear]

If ups ia big enough drop in a bigger all terrain vehicule or lawn mowing tractor battery inside as long as the battery is a sealed one. At limit go to scrapyard and buy a car battery you could plug it with some adapters.

 

[Bill_Bright]

The UPS itself is not 65C. Just the iron core on the transformer when opened up! That is meant to get hot. Ambient here is around 30 C anyway

I knew what you meant and no it is not "meant" to get "hot". They will get "very warm" when under load, but you indicate it is too hot to touch even when the batteries are already fully charged and the UPS is not operating on battery backup power.

When a UPS has not kicked over to battery backup, the AC power just passes through. A small amount of power may be used to make minor adjustments (shaping or "automatic voltage regulation") of the waveform, but the UPS should not be "working" hard, therefore, it should not get "hot".

If your room (ambient) temp is 30°C (86°F), then that is very warm to start with and you are right to assume that will have an adverse effect on the UPS temp (not to mention your computer/CPU/GPU temps). But IMO, that is just more reason to not to trust that UPS.

Certainly turning it off (or even unplugging it) at night is a wise precaution but frankly, in the heat of the day would concern me more.

 

[silkstone]

I knew what you meant and no it is not "meant" to get "hot". They will get "very warm" when under load, but you indicate it is too hot to touch even when the batteries are already fully charged and the UPS is not operating on battery backup power.

When a UPS has not kicked over to battery backup, the AC power just passes through. A small amount of power may be used to make minor adjustments (shaping or "automatic voltage regulation") of the waveform, but the UPS should not be "working" hard, therefore, it should not get "hot".

If your room (ambient) temp is 30°C (86°F), then that is very warm to start with and you are right to assume that will have an adverse effect on the UPS temp (not to mention your computer/CPU/GPU temps). But IMO, that is just more reason to not to trust that UPS.

Certainly turning it off (or even unplugging it) at night is a wise precaution but frankly, in the heat of the day would concern me more.

I've compared it to 3 other similar models now . . . a mixture of new and old in-line UPS models of similar capacity that I have. . . They all run at roughly the same temp, with hot iron cores on the transformer.
I have an online 2000VA model now too, but it would be unfair to compare against that as it has active cooling.

Honestly, my biggest concern now is that when one battery starts to fail then it will cause a lot more heat due to them being in parallel. I am thinking of adding a fuse between the two, but I' need to work out the conversion from the 240V rating to the 12 V rating and figure out a way to install it properly.

 

[Bill_Bright]

Honestly, my biggest concern now is that when one battery starts to fail then it will cause a lot more heat due to them being in parallel.

No. That should not happen. If that was normal, it would be an issue with every UPS ever made that uses more than one battery/cell - regardless if strapped in parallel, series, or series/parallel.

If you have an extended power outage and you run all the cells in your battery down, when power is restored that likely would put a greater strain on the charging circuit.

If a cell in a battery dies, yes, the entire battery will try to level out among the cells. But the other cells would be helping the charging circuit with the load. It is also likely you would see the UPS is constantly charging (assuming it has indicator lights - many don't).

Also, if one of cells dies catastrophically by developing an internal short, it is more likely the UPS will shut down or fail to operate. That said, because UPS batteries are not normally jostled about (like a car or golf cart battery might be), UPS SLA batteries don't normally develop shorts unless they are otherwise abused. They tend to just wear out after 3 - 5 years.

BTW, I got curious about your temps so yesterday I checked the transformer plates on my APC 1500VA UPS here, a smaller 800VA APC, and on a 1500VA Cyberpower UPS. I first checked using my fingers with the batteries fully charged. The plates were warm, but definitely not hot. Then I unplugged them from the wall and let the batteries run down until the UPS started beeping warnings that it was about to run out completely. During that time, the inverters got a little toasty (as expected) but not the transformers (also as expected).

Then I plugged them in and let them charge for 15 minutes with the connected hardware still running. This time I measured the transformer plates using my IR thermometer. The hottest (the Cyberpower) got to 132°F (55.55°C). That was uncomfortable leaving my finger on that for longer than 5 - 10 seconds, but I did not get burned. Now in fairness, my ambient (room) temp is 70°F (21.1°C), but still, heat is heat and as seen by the 4th line in my sig, the "bane of all electronics".

So I checked the published specs of these UPS and they have specified operating environment (ambient) temperatures of 32° to 104°F (0° to 40°C).

So it seems, at least on paper, you are okay.

 

[silkstone]

No. That should not happen. If that was normal, it would be an issue with every UPS ever made that uses more than one battery/cell - regardless if strapped in parallel, series, or series/parallel.

If you have an extended power outage and you run all the cells in your battery down, when power is restored that likely would put a greater strain on the charging circuit.

If a cell in a battery dies, yes, the entire battery will try to level out among the cells. But the other cells would be helping the charging circuit with the load. It is also likely you would see the UPS is constantly charging (assuming it has indicator lights - many don't).

Also, if one of cells dies catastrophically by developing an internal short, it is more likely the UPS will shut down or fail to operate. That said, because UPS batteries are not normally jostled about (like a car or golf cart battery might be), UPS SLA batteries don't normally develop shorts unless they are otherwise abused. They tend to just wear out after 3 - 5 years.

BTW, I got curious about your temps so yesterday I checked the transformer plates on my APC 1500VA UPS here, a smaller 800VA APC, and on a 1500VA Cyberpower UPS. I first checked using my fingers with the batteries fully charged. The plates were warm, but definitely not hot. Then I unplugged them from the wall and let the batteries run down until the UPS started beeping warnings that it was about to run out completely. During that time, the inverters got a little toasty (as expected) but not the transformers (also as expected).

Then I plugged them in and let them charge for 15 minutes with the connected hardware still running. This time I measured the transformer plates using my IR thermometer. The hottest (the Cyberpower) got to 132°F (55.55°C). That was uncomfortable leaving my finger on that for longer than 5 - 10 seconds, but I did not get burned. Now in fairness, my ambient (room) temp is 70°F (21.1°C), but still, heat is heat and as seen by the 4th line in my sig, the "bane of all electronics".

So I checked the published specs of these UPS and they have specified operating environment (ambient) temperatures of 32° to 104°F (0° to 40°C).

So it seems, at least on paper, you are okay.

Thanks for the additional info and the further research. I'm think that these models keep a current running through the transformer at all times as the output is a stepped sine approximation and the idle power draw is 30 W. This draw seems common amongst cheaper APC units, from what I have read.
Out of an abundance of caution, I added a 5 V 120 mm fan, with filter, to the outside of the case. It should blow a small amount of air into it through the vents to keep the internals a little cooler, I just need to keep an eye on dust and possibly clean it out every 6-12 months as it is sitting on the floor and, even with the filter, will accumulate dust.
There's no indicator status for charging or not, but I am familiar with the charging power draw now so I can check periodically to ensure it's working fine.

One thing that surprised my about the inside of the UPS was the diameter of the battery connector wires . . . I couldn't fit a regular spade clip back on, once I'd cut one off . . . then I remembered that it's only 12 V and I^2.R it made sense.
I spliced the UPS wires with copper 7x 0.67mm diameter wire that is commonly used in household wiring here., This should be able to handle the max current rating as it'll be at least equivalent to 14 AGW and is short (I'll do the calculations at some point). I added solder to the splices as well, just to ensure good mechanical contact.

 

[Bill_Bright]

I'm think that these models keep a current running through the transformer at all times as the output is a stepped sine approximation and the idle power draw is 30 W.

You are confused. Whether stepped approximation or pure sine wave output, that only applies when running on battery during a full power outage - which typically is just a tiny percent of the time. The rest of the time, as long as the line voltage is within the set intervention thresholds the line voltage just passes through. There may be some minor voltage regulation (AVR) to slightly attenuate slightly elevated voltages, or to boost slight sags, but if not excessive (to where it kicks over to full battery) the output is still a sinewave.

That said, as I noted before,

Some consumption is normal due to the UPS monitoring and regulation circuits which are active whenever the UPS is plugged in and turned on - even if nothing is connected to the output side of the UPS. So in normal operation, some heat is also normal.

Only with very expensive UPSs is the output being supplied by the batteries full time. This might be with critical life-support or life-signs monitoring equipment in hospital operating rooms and intensive care units.

All other UPS monitor the line in voltage and when it drops below or exceeds set thresholds, the UPS "cuts-over" or transfers to battery power. This needs to happen very quickly. The better UPS will do this in less than 5ms. Note the ATX Form Factor standard requires all computer ATX Form Factor power supplies maintain ("hold up") output for at least 17ms if power drops below 90 volts (for 120VAC mains). Sadly, not all PSUs pass this critical requirement. Also sad is most review sites don't test PSU "hold-up" times either. TPU does!

 

[silkstone]

You are confused. Whether stepped approximation or pure sine wave output, that only applies when running on battery during a full power outage - which typically is just a tiny percent of the time. The rest of the time, as long as the line voltage is within the set intervention thresholds the line voltage just passes through. There may be some minor voltage regulation (AVR) to slightly attenuate slightly elevated voltages, or to boost slight sags, but if not excessive (to where it kicks over to full battery) the output is still a sinewave.

I see. . . then it is likely the AVR functionality. I know that my more expensive online UPS has a Green Mode option that allows it to either use AC directly from the wall (when within a certain range) when Green Mode is on, or regulate the voltage on all current input. From the Logs on that UPS, I can see that the power delivery here isn't the best, with swings from 230 V up to 240 V. Still acceptable for most devices.

A small update. With the small airflow now coming from the fan, there is no warmth coming from the UPS at all. It seems that the minimal airflow generated, is more than enough to cool the internals of these units.

 

[Bill_Bright]

It should be noted that even in green mode, the UPS still provides essential surge and spike protection. And it should still kick over to battery should you get an excessive dip or sag, or complete power outage. This means some monitoring (and thus some power consumption) is still going on. But you are right in that constant AVR is not. And that's fine "if" your grid is stable and your line voltage is relatively clean. If you have a quality PSU, it should provide sufficient regulation for the connected devices.

However, if me, I would keep AVR enabled. To me, the AVR function is the most important feature of a quality UPS with AVR. Back up power during a full power outage is a minor bonus feature. While my power grid is stable most of the time, I do live in Tornado Alley. Complete power outages are not common, but not all that rare either. Surges and spikes, sags (opposite of surges) and dips (opposite of spikes), brown-outs (long duration sags) and "flickers" during severe weather, however, are pretty common in my neighborhood. But also, it is important to note that not all power anomalies originate from the grid. A $15 1500W hair dryer can dirty up a circuit too.

I paid for "good" UPS with AVR. I'm going to keep the AVR function enabled.

Note it is those sags, dips and brownouts that make surge and spike protectors little more than fancy and expensive extension cords - since they do absolutely nothing for those common "low voltage" anomalies.

I can see that the power delivery here isn't the best, with swings from 230 V up to 240 V. Still acceptable for most devices.

That's pretty typical. I note according to the ATX Form Factor standard, the AC Input Line Requirements are as follows:

Vin = 115VAC, 90VAC minimum to 135VAC maximum.
Vin = 230VAC, 180VAC minimum to 265VAC maximum.

So any ATX compliant PSU designed for your region should be happy with your 230 - 240VAC voltage in.

A small update. With the small airflow now coming from the fan, there is no warmth coming from the UPS at all. It seems that the minimal airflow generated, is more than enough to cool the internals of these units.

Good.

 

[silkstone]

It should be noted that even in green mode, the UPS still provides essential surge and spike protection. And it should still kick over to battery should you get an excessive dip or sag, or complete power outage. This means some monitoring (and thus some power consumption) is still going on. But you are right in that constant AVR is not. And that's fine "if" your grid is stable and your line voltage is relatively clean. If you have a quality PSU, it should provide sufficient regulation for the connected devices.

However, if me, I would keep AVR enabled. To me, the AVR function is the most important feature of a quality UPS with AVR. Back up power during a full power outage is a minor bonus feature. While my power grid is stable most of the time, I do live in Tornado Alley. Complete power outages are not common, but not all that rare either. Surges and spikes, sags (opposite of surges) and dips (opposite of spikes), brown-outs (long duration sags) and "flickers" during severe weather, however, are pretty common in my neighborhood. But also, it is important to note that not all power anomalies originate from the grid. A $15 1500W hair dryer can dirty up a circuit too.

I paid for "good" UPS with AVR. I'm going to keep the AVR function enabled.

Note it is those sags, dips and brownouts that make surge and spike protectors little more than fancy and expensive extension cords - since they do absolutely nothing for those common "low voltage" anomalies.
That's pretty typical. I note according to the ATX Form Factor standard, the AC Input Line Requirements are as follows:

Vin = 115VAC, 90VAC minimum to 135VAC maximum.
Vin = 230VAC, 180VAC minimum to 265VAC maximum.

So any ATX compliant PSU designed for your region should be happy with your 230 - 240VAC voltage in.
Good.

Just an update to this. The charging circuit on my UPS died after a few power cuts. I guess the 2x batteries in parallel were just pulling too much current. I replaced it with a ups that was designed to use 2x batteries. It's a 2nd hand SMC1000I but seems to work well.

 

[Bill_Bright]

I guess the 2x batteries in parallel were just pulling too much current.

Nah! Doesn't work that way (assuming nothing is faulty). If charging batteries could demand too much current for the connected chargers, you could not use a tiny "trickle" charger to charge a monster car battery, for example. So something else must have been faulty.

Anyway, I am glad you found a solution that works for you.