UPS sizing

[ubuysa]

This isn't at all scientific but it might help those considering a UPS (Uninterruptible Power Supply)....

Here in Crete we get the odd power flick and even a full power cut from time to time and so several months ago I decided to invest in a cheap UPS. All I wanted it to do was to hold up the PC and monitor long enough for me to do a proper shutdown, I didn't want any bells and whistles.

I bought one locally (it's a Tescom Leo) and chose an 800VA model, it's basically a large black brick with a green led on the front and it cost a few cents under €50. My hope, based on nothing more than gut feel, was that it would be able to power the PC and monitor for at least 5 minutes - long enough to see whether it was a power flick or an outage and to do a proper shutdown if necessary. I actually have it powering the router as well so that a power flick doesn't impact any downloads I might be doing at the time.

Today we had a scheduled 30 minute power outage right across the town so I decided to see how long the UPS would hold everything up. The PC was idling during the outage, and the monitor went into standby of course, the router was being used by our phones but that draws very little power.

I have to report that the test was a failure. The power was restored (30 mins later to the second) and the UPS was still powering everything perfectly well. Frankly I'm staggered, I never expected a 30 minute run time!

Although all PCs and monitors differ of course, and if the PC is working the power draw will be greater, my experience shows that 800VA is more than good enough.

I hope this help anyone considering a UPS.

 

[fnf]

This isn't at all scientific but it might help those considering a UPS (Uninterruptible Power Supply)....

Here in Crete we get the odd power flick and even a full power cut from time to time and so several months ago I decided to invest in a cheap UPS. All I wanted it to do was to hold up the PC and monitor long enough for me to do a proper shutdown, I didn't want any bells and whistles.

I bought one locally (it's a Tescom Leo) and chose an 800VA model, it's basically a large black brick with a green led on the front and it cost a few cents under €50. My hope, based on nothing more than gut feel, was that it would be able to power the PC and monitor for at least 5 minutes - long enough to see whether it was a power flick or an outage and to do a proper shutdown if necessary. I actually have it powering the router as well so that a power flick doesn't impact any downloads I might be doing at the time.

Today we had a scheduled 30 minute power outage right across the town so I decided to see how long the UPS would hold everything up. The PC was idling during the outage, and the monitor went into standby of course, the router was being used by our phones but that draws very little power.

I have to report that the test was a failure. The power was restored (30 mins later to the second) and the UPS was still powering everything perfectly well. Frankly I'm staggered, I never expected a 30 minute run time!

Although all PCs and monitors differ of course, and if the PC is working the power draw will be greater, my experience shows that 800VA is more than good enough.

I hope this help anyone considering a UPS.

Good bit of info, thanks. As far as I remember a back-of-the-envelope calculation for figuring out how much time you have running on a UPS is time_in_hours = UPS_VA * power_factor / PSU_wattage.

If we assume a good UPS then power_factor might be as high as 0.9. If the UPS is rated for 800VA and the PSU is rated for 500W then one can run this system under full load for at least 800 * 0.9 / 500 = 1.44 hours.

Considering that one will likely try to keep the system at a minimal load whilst being powered by a UPS, I think this system could go on for hours! . This makes sense if we think of a UPS as a huge battery pack.

 

[Tony1044]

You'll be lucky to see a PF of 0.9 in the home.

I cannot think of a single white good that does any kind of PF correction. For the gains, it's simply not worth the costs to try and implement (and no one in the home is going to have a synchronous motor running).

In a €50 UPS? Not a chance. Bet it's not even a true sine wave.

800VA is a fair old amount of juice though and I'm not the least bit surprised that it ran for as long as it did but thanks for posting the details, Ubuysa.

 

[fnf]

You'll be lucky to see a PF of 0.9 in the home.

I cannot think of a single white good that does any kind of PF correction. For the gains, it's simply not worth the costs to try and implement (and no one in the home is going to have a synchronous motor running).

In a €50 UPS? Not a chance. Bet it's not even a true sine wave.

800VA is a fair old amount of juice though and I'm not the least bit surprised that it ran for as long as it did but thanks for posting the details, Ubuysa.

Yes, as far as I've seen a PF of 0.6 is more likely so that 800VA will be able to power a 500W system under full load for about an hour.

 

[Tony1044]

Yes, as far as I've seen a PF of 0.6 is more likely so that 800VA will be able to power a 500W system under full load for about an hour.

Though I notice looking at the specs, they do a "SW" version, which as well as having a sine wave output, has a rated PF of 0.9

That's impressive for something like a UPS. So much so, I am a tad doubtful that true life would be quite so good.

 

[ubuysa]

Interesting. I have some experience with inverters, we had a 1.6kVA on the boat (with 400Ah of 12v battery capacity behind it). We used it often at anchor to run the DVD player and TV and even with those light loads the current drain from the batteries was over 10 Amps. We once ran a hair dryer as a test and saw 50A from the batteries. Scary numbers (and BIG cables). Inverters I know are terribly inefficient which is why I wasn't expecting that long a run time.

 

[Tony1044]

Interesting. I have some experience with inverters, we had a 1.6kVA on the boat (with 400Ah of 12v battery capacity behind it). We used it often at anchor to run the DVD player and TV and even with those light loads the current drain from the batteries was over 10 Amps. We once ran a hair dryer as a test and saw 50A from the batteries. Scary numbers (and BIG cables). Inverters I know are terribly inefficient which is why I wasn't expecting that long a run time.

Well if you take an average hairdryer running to be around 1500W for the heater element, that's 12.5A on the dryer. Since as voltage goes up, current comes down, a 4x current drain at the batteries doesn't seem untoward. Was it a 5 or 6 battery unit by any chance?

 

[ubuysa]

Well if you take an average hairdryer running to be around 1500W for the heater element, that's 12.5A on the dryer. Since as voltage goes up, current comes down, a 4x current drain at the batteries doesn't seem untoward. Was it a 5 or 6 battery unit by any chance?

It was 4 100Ah batteries in parallel.

 

[Tony1044]

It was 4 100Ah batteries in parallel.

Sure they weren't parallel-serial? It's normal for that level to try to have 24V as the source in the more powerful UPS's.

Well used to be. I am not really that up to date on this.

Assuming all in parallel, then the source voltage was only 12V...which means a 20x jump in voltage and the corresponding drop in current.

But that said...50A at the batteries would equate to 2.5A at the hair dryer or 600W - could be the UPS limiting the output of course in this situation.

24V at the batteries would seem to make more sense as it'd be 5A at the hair dryer, or (you guessed it) 1200W which seems far more typical.

Hmm...just noticed I seem to have slipped up with the arithmetic earlier and doubled the figure - 1500W is 6.25A not 12.5 (at 240V)

All of this is dragged from my long distant memory so could all be wrong, as well

One of my first pieces of practical work on my first year electronics was to research, design and build a working UPS. But that was <mumble mumble> years ago and pre-internet

 

[ubuysa]

Sure they weren't parallel-serial? It's normal for that level to try to have 24V as the source in the more powerful UPS's.

I lived on that boat for 13 years (not the one in my avatar), I think I'd have noticed.

Some boats use a 24v supply, it helps prevent corrosion on all the plugs and connectors, but the volume (and weight) required for the batteries is usually too big for boats under 50ft (ours was 45ft). We used a 12v supply which is much cheaper, it means that automotive products work out of the box too. They were four 12v batteries in parallel, and I replaced them about four times in those 13 years so I know very well how they were connected.

But that said...50A at the batteries would equate to 2.5A at the hair dryer or 600W

It was a 1200w hair dryer but we only used it on its low heat setting (it was only a test after all) and that probably was 600w, though I've no idea. All I measured was the current in the battery cables. It should also be noted that the inverter was in 'eco' mode where it's supposed to adjust its input current based on the mains load.

 

[Tony1044]

I lived on that boat for 13 years (not the one in my avatar), I think I'd have noticed.

Some boats use a 24v supply, it helps prevent corrosion on all the plugs and connectors, but the volume (and weight) required for the batteries is usually too big for boats under 50ft (ours was 45ft). We used a 12v supply which is much cheaper, it means that automotive products work out of the box too. They were four 12v batteries in parallel, and I replaced them about four times in those 13 years so I know very well how they were connected.



It was a 1200w hair dryer but we only used it on its low heat setting (it was only a test after all) and that probably was 600w, though I've no idea. All I measured was the current in the battery cables. It should also be noted that the inverter was in 'eco' mode where it's supposed to adjust its input current based on the mains load.

Only 13 years...not long enough for me to notice things, necessarily.