APC UPS perfect sinusoidal output - Tested for Power Factor at ITTSB Lab 2021

[Kiriakos GR]

APC SUA1000XL this is a UPS that at previously made tests, this show great delivery of perfect sinusoidal output.
Perfect sinusoidal output this translates to significantly stable delivery of volts output.
But in the end of the day every UPS this is an inverter.
DC/AC inverter this trying to do it best, but results will vary according the age of the technology in use.

I will punch some numbers from my test for you to review.
Test system this is PC computer just the main unit.

AC from grid: 235V 0.542mA 114 Watt  PF=0.75 (PC computer main unit alone)
APC UPS load sensing 17,5 % in VA ( Sum of PC computer along peripherals)

Test starts ....  Grid power OFF 

AC from UPS : 233V 1.020A 90 Watt PF=0.40  (PC computer main unit alone)
APC UPS load sensing  23 % in VA ( Sum of PC computer along peripherals)

Power supply under test this is Corsair CX750 (the successful revision)   83% Efficiency at 15%of load. (highest efficiency 88.49% at 50% of load).

I am now wonder of what I did witnessed in this test?
It appears that the APC UPS while this indeed deliver perfect sinusoidal output, it circuit this delivering current its not that perfect, or clean, Power Factor damage by 45% this is indication that inverter losses are high, Efficiency score this is Low.

I am now using an budget power meter energy recorder, and I will not dare to speak of measurements with accuracy that I can trust.
But no matter the measuring error, 45% of lower PF this is a fact.

APC UPS load sensing this is UPS internal sensor / measuring of VA, PF reduction this increases VA consumption by 5.5% at this test setup.
Better conclusions they would be possible, if I had available an scientific grade power meter, but this can wait for now.     

[Kiriakos GR]

Three years later I have to add few corrections on this topic.

SUA1000XL this is designed to deliver at best PF=0.8 , this was considered as high-end at the year 2003~2004.
 
From the side of the AC load, it is a matter of product age and electronics in use, especially for Non-Linear loads, that is a switching power supply.
I do love to perform measurements, but especially at switching power supply, they are specific conditions so the test to be considered as valid.
The accuracy of the digital power meter, this is the last parameter so to be examined.

SUA1000XL this has 1000VA label and it is 800W.
An Computer PSU at the size of 750W, it should be stressed at 40% at least, for a proper Power factor evaluation.

 
In conclusion, if you are up to evaluate uninterruptible power supply Power Factor delivery, you should use a small resistive load instead.
A pure resistive AC load due it nature will operate at PF=1.0, from the other hand any UPS inverter (on battery), this will deliver as max the supported Power Factor according the potentials of it own electronics.

Old in age or low cost, uninterruptible power supply they might deliver at their best PF=0.6

Now it starts to matter the quality of the AC power measuring device.
Low cost plug-in power meter (16A) , they are famous about not delivering Power Factor measurement with accuracy and at high in digits resolution.

So I have to declare my above testing as not a valid one, but I will leave this topic standing, as reminder of past made mistakes.     

-------------------------------------------------------------------------
APC recommends UPS calibration this to be performed at not less than 40% of UPS Max loading capacity, what ever this be.
For the SUA1000XL this translates to 320W of load, and I did make my own load tester by combining 2 x 150W =  300W by the use classic style bulbs.
Then I did added and a third 40W of classic style bulb, and now I do perform UPS calibration with 340 Watts that is the recommended by the manufacturer. 

Personally I will follow this recommendation, load testing at 40% or higher, and for power factor evaluation, when I am testing a UPS.