Question Brand new CyberPower 540W UPS - shutting off with a load of 415W

D

Deleted member 2922629

Guest
I bought a brand new CP900EPFCLCD and let it charge for 24hours, and tested it on my PC (nothing else plugged in).

With a load of 100W, it works fine. However, if I run a GPU benchmark to get it up to 415W, the UPS just shuts off with an F02 error, saying "Overload", despite the fact the little graph of "load" on the LCD is not full.

I opened the battery compartment and tested with a multimeter, that showed 13.28V, which seems very healthy.

Needless to say I'm pretty disappointed - I know the VA can be wildly exaggerated, but surely it should work on 76% of the rated powe?

This is sinewave too - so I don't think it's anything to do with the computer PSU.

Is it just a case of these units not working in general, or have I just been unlucky with a dud?
 
D

Deleted member 2922629

Guest
Welcome to the forums, newcomer!

We're going to need a little more information. Please list the specs to your build, like so:
CPU:
Motherboard:
Ram:
SSD/HDD:
GPU:
PSU:
Chassis:
OS:
Monitor:

I think the only thing that matters here is the UPS model and the PSU. The PSU I have is an SF750.
 

Ralston18

Titan
Moderator
This UPS?

https://www.cyberpower.com/eu/en/product/sku/cp900epfclcd#overview

This PSU?

View: https://www.reddit.com/r/sffpc/comments/sxrpfo/a_corsair_sfx_psu_died_on_me_again/


https://www.pcgamer.com/corsair-recalls-compact-sf-power-supplies-following-a-rash-of-failures/

Sources - where purchased?

= = = =

The computer specs do matter.

Someone may recognize a problem component or some other factor that needs to be considered.

What GPU benchmark did you use?

Any other devices in the power connection path; Wall outlet, extension cord, power strips, surge protectors, etc.?

What all else is the UPS serving - monitor(s) perhaps?

An overload can be very sudden spike that occurs and shuts down the UPS before the "little graph" etc. can even "catch up".

And just because the battery has the voltage that does not mean that the battery has the necessary stored power.
 
D

Deleted member 2922629

Guest
This UPS?

https://www.cyberpower.com/eu/en/product/sku/cp900epfclcd#overview

This PSU?

View: https://www.reddit.com/r/sffpc/comments/sxrpfo/a_corsair_sfx_psu_died_on_me_again/


https://www.pcgamer.com/corsair-recalls-compact-sf-power-supplies-following-a-rash-of-failures/

Sources - where purchased?

= = = =

The computer specs do matter.

Someone may recognize a problem component or some other factor that needs to be considered.

What GPU benchmark did you use?

Any other devices in the power connection path; Wall outlet, extension cord, power strips, surge protectors, etc.?

What all else is the UPS serving - monitor(s) perhaps?

An overload can be very sudden spike that occurs and shuts down the UPS before the "little graph" etc. can even "catch up".

And just because the battery has the voltage that does not mean that the battery has the necessary stored power.

Yes those are the correct UPS and PSU models.

The machine is a running a 5700X CPU on a B550 board. 64GB DDR4. RTX3080 founders edition.

There's nothing else attached to the UPS - as I just got it I was wanting to make sure it worked.

As to power I have a good high-frequency power monitor showing the load going no higher than 415W.

As to the charged state - yes I do realise the voltage doesn't necessarily mean it's charged. However, it does give confidence that the charge/voltage curve model is correct. It's usually when the battery is duff that the charge state can be reported erroneously.

Given the fact it's new and was charging for 24 hours before, I don't think it's a battery issue.
 

Ralston18

Titan
Moderator
How was the UPS sized?

How long does the benchmark test last?

From the UPS User Manual:

"HOW TO DETERMINE THE POWER REQUIREMENTS OF YOUR EQUIPMENT
1. Make sure that the total Volt-Amp (VA) requirements of your computer, monitor, and peripheral
equipment does not exceed the 550VA/900VA/1300VA/1500VA.
2. Ensure that the equipment plugged into the battery power-supplied/surge outlets does not exceed the
UPS unit's rated capacity (550VA/350W for CP550EPFCLCD, 900VA/540W for CP900EPFCLCD,
1300VA/780W for CP1300EPFCLCD, 1500VA/900W for CP1500EPFCLCD). If the rated unit capacities
are exceeded, an overload condition may occur and cause the UPS unit to shut down or the circuit
breaker to trip."


Take a look at the UPS's full specs and its' Runtime chart (interactive).
 
D

Deleted member 2922629

Guest
Well, then I'm out. Good luck to you.

I replied to @Ralston18 with the specifics. So it's clear to me you're just making a point, and flouncing out - making sure we know you're flouncing out. I see you're a mod, so it's surprising to me you behave that way. I couldn't see why a UPS dropping out due to a load I'm monitoring closely with a good power monitor should have anything to do with anything other than the PSU. That said, I provided the further info anyway.

Anyway, in order to, you know, actualIy be helpful, I contacted CyberPower. They say they don't "recommend" using more than 70% of the stated capacity, and since I'm using 76% they said I should get a larger one. I asked why they bother stating 540W, and they didn't have a meaningful answer.

It's tricky to find reviews that actually test UPSs at various loads, so makes choosing one difficult, especially as the sizing guidance suggested I was fine at my planned use.
 

Ralston18

Titan
Moderator
Many product tests are done under ideal circumstances.

UPS/PSU manufacturers do so to claim as much possible output wattage as they can.

Component manufacturers do so to claim as low as possible use of power/wattage.

And is very important to pay attention to the fine print, the caveats, and the details of any product claims, EUAs, warranties, RMA terms, etc.

Stuff not printed on the box. And difficult to read in the documentation. May be hard to find on manufacturer websites.

Very much the same here: the UPS is probably not able to support the PSU which in turn is trying to support system components.

Not uncommon that some product works well at measurable parameter "X" but crosses a threshold and fails at "X + 1" or "X - 1"

Sizing guidance for many products is not always objective. Consumers may be steered towards more profitable versions. I.e., " a larger one".

It appears that CyberPower seems to have done just that.....

It does go beyond "I think the only thing that matters here is the UPS model and the PSU. The PSU I have is an SF750. "

That is why we ask about such build components and specs as @Lutfij did from the beginning.

I do not view it as @DSzymborski "flouncing out".

If "I think the only thing that matters here is the UPS model and the PSU. The PSU I have is an SF750." is your bottom line then all you can really do is to work it out with CyberPower.

And I will have to flounce out as well.