Smart UPS hacking with larger batteries -- Canceled

[1clue]

Hi,

I want to know about real-life pitfalls of attaching oversized batteries to a UPS. I'm looking for somebody who tried it at least once, and has real-world input as to the expected result. I can come up with all sorts of theoretical reasons why it MIGHT be a good or bad idea.

Please bear with me. I chose this section because it's hardware, and a serious inquiry. A Gentoo box is hooked up to this thing, and the serial cable is connected to shut the system down, that's as close as it gets to Gentoo.

So I have a Tripp-Lite Smart 1500 LCDT UPS. The UPS was bought for the devices attached to it, meaning they were all new at the same time. It's a SOHO-grade UPS with an LCD readout and some sort of low-function computer in it to manage things. It uses 2x 7Ah sealed lead-acid gel cells, nothing odd compared to other UPSs of similar size.

I just got a power outage and found out that the batteries give me about 2 seconds of backup power. I still want the systems hooked to it to run, and I want them on a UPS. But it's not production systems, and it's not mission critical. It's all old hardware, so I don't want to buy a new UPS. For professional installations my hardware is retired before this point.

Shopping around, I see that the gel cells for the thing cost not that much less than an automotive sized deep cycle battery, so as I have done every time I get into this situation, I'm tempted to hack the thing with a huge pair of batteries. I've been curious about this for literally decades, but never tried it. Now seems to be the time, but I don't want to waste money.

Specific questions:

1. Are these UPSs smart enough to expect a specific battery capacity? In other words, if it charges for too long will it decide the battery is dead? Or will it continue to charge until the thing is full?
2. Does the UPS correctly anticipate low power state and send a shutdown signal right before the threshold gets too low (5 minutes before loss of power, for example)
3. Does the UPS correctly show state of the battery?
4. Is there something I'm not thinking of that makes this a bad idea?

Specific things I know:

1. I know that automotive lead-acid batteries discharge hydrogen and oxygen during the charging cycle.
2. I know that those gases are explosive and not suitable for an enclosed environment. I can compensate by venting to the outside.
3. I know that the surge/spike protection is cumulative and the whole works may die right after I set this up.

[Jaglover]

I used a motorcycle battery as a replacement for many years, this was an APC UPS and these are as smart as they get. Finally the battery went dead and the UPS killed itself trying to charge it. Was my fault really, I saw the battery was dead and left it connected. When the battery was good it gave about triple discharge time compared to stock capacity battery. Of course, deep cycle is no good for car battery. In particular when left discharged for some time it will have devastating effect. Modern car batteries do not require venting.

[1clue]

Thanks.

I see a bunch of blogs about this (instructables etc) but an enthusiastic blogger is usually only enthusiastic about blogging until the project is built, not so much after the thing went dead. So you get all sorts of hurray! but don't find out about the end-of-life scenario.

This is already what I would call a dead UPS. It still is powering the hardware but if the power goes out it's done.

I have another one in close to the same state -- about 10 minutes life, should be an hour -- and I think it has a single battery. This one is just a button, no screen. I have a boat (deep cycle) battery laying around somewhere, I wonder if it has life in it? Might be worth a shot on that one too, maybe as a practice run?

I've wondered about this for as long as I have known about UPS systems. I know that the output power can't be raised, but my argument with UPSs has never been about output capacity but rather how long they can go without line power.

I have a charger for remote controlled cars/aircraft/whatever, for LiPo and other chemistries. This charger wants to know the battery capacity, and if it charges too long it will stop and tell you the battery's no good. If my dead UPS is like this then I'm screwed with this project, and would be better off buying a gel battery compatible with the UPS.

[1clue]

Has anyone had experience with a smarter UPS? Like with an LCD screen?

So modified question list:

1. Are these UPSs smart enough to expect a specific battery capacity? In other words, if it charges for too long will it decide the battery is dead? Or will it continue to charge until the thing is full?
2. Does the UPS correctly anticipate low power state and send a shutdown signal right before the threshold gets too low (5 minutes before loss of power, for example)
3. Does the UPS correctly show state of the battery?
4. How did your setup fail at end-of-life?
5. How does the battery lifespan compare to standard AGM battery?
6. Would/did you do it again?
7. Is there something I'm not thinking of that makes this a bad idea?

Modified things I know:

1. I know that automotive lead-acid batteries discharge hydrogen and oxygen during the charging cycle.
2. I know that those gases are explosive and not suitable for an enclosed environment. I can compensate by venting to the outside.
3. I know that the surge/spike protection is cumulative and the whole works may die right after I set this up.
4. I know that not-so-smart UPSs can and are regularly hacked like this.
5. I know that some higher-end UPSs allow for external batteries to extend the uptime of the pack

[xdarma]

Hi,
I've modified a old APC Smart-UPS 750 (SUA 750) some years ago with two car lead-gel battery.
I no longer work there, but I think still working.

[*]Are these UPSs smart enough to expect a specific battery capacity? In other words, if it charges for too long will it decide the battery is dead? Or will it continue to charge until the thing is full?

In my experience, UPS was not so smart. UPS worked as usual even with 10 times bigger batteries.

[*]Does the UPS correctly anticipate low power state and send a shutdown signal right before the threshold gets too low (5 minutes before loss of power, for example)

Never reached that limit. Longer power out was around 1 hour and UPS has charge for another 30 minutes, so no reason to shutdown.

[*]Does the UPS correctly show state of the battery?

Yes.

[*]How did your setup fail at end-of-life?

Don't know.

[*]How does the battery lifespan compare to standard AGM battery?

I think the life of car batteries should be longer but mines was "rugged" batteries.

[*]Would/did you do it again?

Yes, but only for personal use, or strictly controlled. Just to avoid legal problems.

[*]Is there something I'm not thinking of that makes this a bad idea?

Maybe is better to remove internal batteries, IMO.
And pay attention to avoid short circuit of hanging cables.

[rpruen]

I too have modified a UPS to have bigger batteries. This was an APC UPS so about as smart as it gets.

I had no problems. I used two 100AH car batteries so as to have a run time of 20 mins or so.

Richard

[bunder]

how heavily are you loading the UPS? if it shuts off in 2 seconds it might be something with the UPS not liking active PFC power supplies (or rather the supplies not liking the voltage the UPS is putting out)... long story short, the UPS thinks nothing is attached and it shuts off when the power goes out (or the power supply turns off, it's kind of a weird problem)

[1clue]

how heavily are you loading the UPS? if it shuts off in 2 seconds it might be something with the UPS not liking active PFC power supplies (or rather the supplies not liking the voltage the UPS is putting out)... long story short, the UPS thinks nothing is attached and it shuts off when the power goes out (or the power supply turns off, it's kind of a weird problem)

It's loaded the same way it was when it was new. It was sized to provide 30 minutes of battery power. All the equipment hooked to the supply was bought at the same time as I bought the supply.

Edit: Everything connected to this UPS was new when I bought it the UPS (no used equipment attached, new UPS, new box, new monitors) and I know it supplied backup power for a solid 30 minutes, and then safely shut down the box via the serial cable. This "pod" of equipment has been configured the same way since like 2011 or so. There are no new devices hooked up, although the computer has had drives changed out over the years. But mostly to SSD from spinners, so current requirements probably have gone down a bit.

The UPS is nowhere near the maximum current capacity rated for the box. This isn't my first rodeo. I don't cram a device into every outlet on the UPS, or keep adding stuff until the circuit pops. There is no new technology in power supplies being used. It's the same old 600w supply for the box.

The system is running now, nothing is hot and nothing is failing to work. But it's not on battery power either.

[1clue]

I decided to only experiment with one UPS at this time. The bigger one I decided to go the OEM route. It's back up with fresh batteries in it.

My other one is going to inherit a boat trolling motor battery I have sitting in my garage. I'll let you guys know how it goes.

[1clue]

My marine battery won't take a charge. I bought an OEM battery for the second UPS. So the entire thread is for nothing.

Thanks anyway guys, sorry for wasting your time. Maybe I'll try it next time.

[msst]

Code: Select all

It's loaded the same way it was when it was new. It was sized to provide 30 minutes of battery power. All the equipment hooked to the supply was bought at the same time as I bought the supply.

Then the original battery has simply died. Happens after a several years, I had the same just not with 2 seconds remaining time. And my UPS also told me that the battery was going bad (low capacity).
By replacing with a new original battery you should again get 30 mins standby time. But I guess you just did that and it worked, right?

[1clue]

Code: Select all

It's loaded the same way it was when it was new. It was sized to provide 30 minutes of battery power. All the equipment hooked to the supply was bought at the same time as I bought the supply.

Then the original battery has simply died. Happens after a several years, I had the same just not with 2 seconds remaining time. And my UPS also told me that the battery was going bad (low capacity).
By replacing with a new original battery you should again get 30 mins standby time. But I guess you just did that and it worked, right?

Replacing old batteries is old news for me. Everything is working and tested. Thanks.