Sine Waves and Efficiency of different UPSs


#21

Hey, did u manage to check Luminous UPS. How much Systek 1500V UPS will cost?


#22

To add load to UPS, please use energy saver bulbs. That will give you near accurate load. For instance, adding four 50 watt energy savers means 200 watts in total (am i right?)

If you use computers, load will keep on varying.


#23

I could get data of those UPS which we have in our plant or which I could get for demonstration.

So, for efficiency of Homage UPS, I will inform within 3-4days.

People will also astonish like me after seeing the low efficiency of Siemens while it has pure sine wave output. After discussion, I have been informed the reason that there are many filters in this UPS to purify the sine wave. So each filter loss some amount of power which then cause by low efficiency.

For testing the switchover time of Systek, I will inform within 3-4days as I am making discussion with their engineers.


#24

@signode

Great Job mate! so systek is the winner.


#25

Do all good quality inverters have similar losses like 68%?


#26

[quote=“ahmedrazaqa, post:5, topic:16913”]

We purchased Excelsior Classic series for our office but our experience was not good. Computers used to get restarted on power failures. That was the only purpose of them for which they were bought as building was on Generator supply. Multiple visit by their service reps did not resolve the issue.

[/quote]

This guy has a negative review of this Systek series… Its claimed switch-over in brochure is 2ms… if it was really the case, his PC would’ve remained ON!


#27

Yes. I am talking with Systek team about the 2ms switchover time if it is real. I am getting data from them then i would be able to calculate the time. Moreover, i will do an experiment in my home. So need some time.

I also snubbed them why they quoted 90% efficiency while i could get 68%. But nevertheless, the UPS, with its features is good.


#28

@Signode

Interesting comparison, though the measurements and results don't seem correct.

I can not even imagine a Siemens UPS to be 65% efficient - and out of curiosity I checked the technical specs of Masterguard 1000VA and they claim it to be 86%.

But as a comparison, this data might be useful.

Nice work..........

I can't promise but I will try to take out some time and measure the Mercury 1000VA (it is like Homage 1000VA) and the APC Smart UPS 1000VA


#29

I agree with Ijaz Ahmed. To understand how a sine-wave inverter works you must realize, generally, a sinusoidal PWM is fed to an H-Bridge. While the PWM will have some harmonics they are nothing as compared to a square waveform. To do test this, I did a FFT of a sinodial PWM and a FFT of a square wave. The results are below:

uPpUf.pngus7TC.png

Note how large the harmonics are in the first figure as compared to the PWM FFT. These results are VERY crude, by the way, My test jig was basically a AVR Mega 128 that I had on a breakout board, The AVR didn't have any decoupling caps(!), no pull-up on Reset etc. I just slapped some code onto it and measured the output. So, with a proper setup, the PWM FFT willl be better and the FFT of the square wave will be worse (because the AVR will output a better square wave).


#30

for god sake stop discussing about UPSs ups game is over now it can't charge itself with 20hours of load shedding, now lets move on some other energy sources like wind solar if you want to be live in country like this

solar is best for punjab regions and wind is for sindh areas please guide on how to make where to get and how to install :-) forget Ups


#31

@armada

Even with solar or wind solutions, you need a UPS / Inverter to convert the DC (from batteries) to AC because all your appliances are AC.


#32

Homage data added.

Dear PsychedelicBreakfast

I do have digital oscilloscope but i am trusted in analog for better results. Testing a same thing with digital and analog oscilloscope provides different results.

Dear Ijaz,

After doing many experiments, it has been proved that for measurement of output energy of UPS, digital meters don't work correctly (although i have very fast True RMS meters). So you are recommended to use Analog Disc Type Energy meter.

meter.jpg


#33

[quote=“signode, post:32, topic:17125”]

Dear Ijaz,

After doing many experiments, it has been proved that for measurement of output energy of UPS, digital meters don’t work correctly (although i have very fast True RMS meters). So you are recommended to use Analog Disc Type Energy meter.

[/quote]

Thank you for the recommendation. Your recommended instrument is a Kilo Watt-Hour meter, not a precise power meter.

Also this meter is designed for sinusoidal AC, how can you trust it to be accurate on non-sinusoidal waveforms ?


#34

@Ijaz

Looking forward to your efficiency test report


#35

#36

[quote=", post:33, topic:17125"]

Also this meter is designed for sinusoidal AC, how can you trust it to be accurate on non-sinusoidal waveforms ?

[/quote]

Dear Ijaz,

This is matter of understanding. I try to explain you.

The rotating disc energy meter are so precise that one can easily calculate the wattage of even small electric appliance. Just route the wires through meter, switch-on the load and count how much time disc takes to complete one revolution. Note down the constant written on meter (the meter i have,400rev/kw/hour written on it ) Now calculate the wattage by using formula.

I run a fan on wapda and measured the energy by using two EMs and results were as following;

Digital energy meter = 131W

Rotating Disc energy meter = 130W.

RPM of fan = 1280

Now run the fan on a UPS and measured;

Digital energy meter 171W

Rotating Disc energy meter 133W.

RPM = 1270

Now which one meter was correct? First we should know that our requirement from fan is 1280rpm and it consumed 130W to rotate at that speed. By seeing this, I ignored the difference in speed and assumed that fan was using 130W energy from UPS to rotate at 1270rpm and 03W was wasting in fan in the form of heat. So the rotating disc energy meter correctly measured the power.

Same theory applied at other UPSs.

I did that experiment after talking with different engineers about energy measurment of such non-sinusoidal wave form. The comments of one of those engineers is as following;

"A good digital will have a very high sample rate and execute a large number of samples and instantaneous volts times amps calculations which are then summed.

A lower cost unit will measure the current, and the voltage and the angle between them and multiply volts times amps time power factor, but this is only true for a sinusoidal voltage and current waveform.

One way to get an idea is to use a rotating disk power utility type meter and count the disk revolutions. This should be pretty accurate."

Test results of Switch-over time of Systek

As committed, I put 516W (86%) load on my home UPS ECS-202-600W including a desktop PC, switched off the wapda and found no PC re-start problem. I did several times but found no PC re-start problem while the UPS out voltage were almost 202V.

The guy who faced PC restart problem with Systek may be the cause of low battery or having some problem in wiring.


#37

Nice explanation signode

Just asking, what is the price for an average rotating disc meter? I am no engineer even by any stretch of imagination but i need something to measure current of different home appliances.


#38

[quote=“generaluser, post:37, topic:17125”]

I am no engineer even by any stretch of imagination but i need something to measure current of different home appliances.
[/quote]

How about this type of meter ? Plug in anything and see the power drawn by the appliance, it will also calculate how much units the appliance has consumed a certain time, if you keep the appliance plugged in through this power meter.

33o2kqf.jpg


#39

@ijaz

Its nice! How much is this for?


#40

Dear Ijaz,

Please test that device with load connected with Wapda and then on UPS (with same load) then inform us about results.

The price of a rotating disc energy meter is almost Rs.2500/ (Syed Bhai or PEL not a Chinese brand) and available at Al-Madeena electric company Branderth Road Lahore.