Question Cyberpower CP1000AVRLCD UPS, Voltage, Battery up-size (hack), runtime

[LinwoodFerguson]

I am tired of replacing the little UPS batteries in my three Cyberpower UPS's. I just replaced one in the CP1000AVRLCD with a 35ah AGM Deep cycle, and it now has about 90 minutes runtime (maybe more). But this caused me to look more closely at some aspects and I am confused.

First, before and after replacement, the Cyberpower Software Powerpanel for Business shows the battery at 20v. It is a 12v battery, it should show about 13v +/-. Again, before and after my change, so unrelated.

Is the software just flakey? Is there some kind of buck converter in front of the inverter that is raising the voltage and measuring that?

After replacing the battery with the large one I ran a test (see the events below). It ran almost 90 minutes but then gave the really strange error "not completed, the battery runtime is insufficient". After all that time.

My guess is the firmware or software is just hard coded to work with tiny batteries, and giving false information. As it is recharging it's got nonsense numbers also, the 42% is probably right, but the 8 minute runtime is nonsense.

Has anyone else hacked Cyberpower UPS's with larger batteries, is there any fix to get the estimates correct, so I don't get alarms that are bogus if it starts to discharge (I've simply turned the shutdown of the PC off, but I guess it might shut the UPS output off incorrectly as well).

Linwood

 

[Lutfij]

Has anyone else hacked Cyberpower UPS's with larger batteries,
A crude search off the www lead me to this reddit. There are videos on YT of people modifying their UPS's as well but I'd take that with a grain of salt.

 

[LinwoodFerguson]

Sorry, I found lots of examples of people doing it, but none seemed to explore the interaction with the monitoring software. The actual "hack" in this case was just a piece of wire to hook up the battery, hardly exciting, and that's also what most of the youtubers are doing.

I'm interested in people who might have looked into the charging circuitry and/or firmware and in particular how the larger battery is failing to be reflected in estimated runtime?

 

[bill001g]

Hard to say they can't actually measure the charge in the battery other than to run it at some fixed amp rate and see how long it lasts. Would be nice if there was some way to know but a old battery will look perfectly fine but just suddenly drop in voltage when you use it.

It would not be unexpected that they is some kind of software restriction they built into the calculations. Not likely something you are going to change easily. Most UPS have no ability to update the firmware you would have to get very ambitious and unsolder the flash chip figure out how to change the software and then solder it back.

In general the USP will work fine even if it reports the wrong numbers. Most do not shut down until the voltages themselves get too low no matter what runtime number says.

Be aware this is not really a good idea to put really large batteries in these units. The parts have a duty cycle. The charging circuit and inverter circuits are only designed to be run for a certain amount of time without any break. Not sure it is a heat thing or some other limit. This will likely greatly reduce the life of the UPS electronics.

If you really want to do this you buy what is called a online or double conversion UPS. These have parts that are designed to run all the time. They tend to be very expensive. lately building your own online UPS with part designed for solar use can be cheaper especially when you are trying to use larger batteries.

 

[LinwoodFerguson]

In general the USP will work fine even if it reports the wrong numbers. Most do not shut down until the voltages themselves get too low no matter what runtime number says.

Be aware this is not really a good idea to put really large batteries in these units. The parts have a duty cycle. The charging circuit and inverter circuits are only designed to be run for a certain amount of time without any break. Not sure it is a heat thing or some other limit. This will likely greatly reduce the life of the UPS electronics.

If you really want to do this you buy what is called a online or double conversion UPS. These have parts that are designed to run all the time. They tend to be very expensive. lately building your own online UPS with part designed for solar use can be cheaper especially when you are trying to use larger batteries.

Thanks. I'm watching it recharge. It's really slow, but it's going steadily and at least at the case it's not overheating. But sure, the lifetime may suffer, but (other than regular battery replacements) it's cheap if it quits. From the steepness of the discharge the overheat risk is probably much larger during discharge (and I didn't look at it then), but it has a fan then I am pretty sure.

That's why I only went up from 9 to 35ah (and was actually aiming more like 30). Wasn't even sure it would have enough current to charge, say, an old car battery from a deeper discharge.

But so far so good.

But yes... my longer goal is to find a good charger/inverter and do a DIY UPS for both this room and (one wall over) a network closet. Biggest problem there is haven't found one I can monitor over the network, they all tend to have fancy displays or remote controls or panels, but not ethernet or USB monitoring (I think they are mostly for RV's and similar, at least the ones in the 2kw range).

But, so far so good with this approach. And not much money at risk, and figure one good deep discharge/recharge cycle will show if there's a real overheat risk. If it prematurely ages the UPS - no big loss. I usually buy a new one after 2-3 battery replaces, this one is from 2018 or so.

 

[LinwoodFerguson]

Just in case it is interesting to someone later some observations (this is all not terribly useful info, so feel free to move on if not a trivia person).

I looked at the State of Charge as indicated by the UPS itself as it charged the new battery. It had a weird step function. I assumed this was based on voltage (as opposed to coulomb counting), but now I wonder:

The voltage it measured however was not a step function (at least not such large steps), but fairly smooth. It also didn't rise to the pre-swap voltage. And indeed it did not, I measured the no load voltage of the failing battery and it was 13.3 and the AGM was 13.0, and the below is roughly proportional. Still don't know why it shows such high battery voltages for a 12v (class) battery.

But it's SOC algorithm must not be a simple function, it must have a table of values with some pretty big gaps? Or it is including time as a factor?

Regardless, 13.0 is a pretty good fully charged voltage for AGM. If I adjust the bottom voltage (remember it gave up without full exhaustion) it was about 11.2, but that's still probably in the 10% range.

The recharge cycle took about 22 hours. I was worried it might run for days. By comparison I had a 2-battery 15kw Cyberpower with stock batteries, did an exhaustion test on it, and it took 12 hours, which was much longer than I would anticipate, but at least the charge time is not radically (like days) longer. interestingly there was no step function in the other (it's of similar vintage and type just larger). Makes me think the step might be related to time in some fashion.

I do think this was, so far, successful enough when that other UPS' batteries go, I'll get two 30ah-ish batteries for it as well.