I measured the current on the AC cord to a file server with a clamp (on an extension cord that I broke the wires out of) and a Fluke 73.
Does it matter that the 73 is not true RMS? I thought the wave would be nearly a sine.
Since UPS's are rated in VA, not watts, is measuring the current alone good enough? Does that make PF irrelevant?
The peak current I measured while excersizing the server was 1.4 A. That's only 168 VA, but if that's true then UPS's must be rated for the load that would run them down in 30 seconds.
I'm thinking that this was just a useless measurement, but not sure why.
Yes it was. Size the UPS based on the rating of the server power supply or supplies. Add up the rating of all of the equipment you will attach to the UPS and that will give you a good idea of what you need. This does NOT apply if you want to connect the following to the UPS: motors, printers (particularly laser), or highly inductive loads. It does work if all you are connecting are computers, monitors, and routers.
Nowhere near, all switching power supplies do bad things to their power feeds, some worse then others.
Nope, since you are only seeing a rough average of current usage, you aren't seeing the peaks at all.
Useless? I wouldn't go that far, it did give you a rough (perhaps extrememly rough, depending on how "non sine" the wave was) idea of what average current is being drawn.
Basically use the ratings on the power supplies of the server, add them up, and then add an additional percentage, how much is up to you, I'd say
50% as a minimum.
Also remember that all UPSs basically have two ratings, a rating of the inverter used, and a Wh rating, which tells you how long it can hold out. Adding up the wattages will tell you what kind of inverter your UPS has to have, the Wh rating is up to you in how long you want it to run after power has failed.
Also, remember that the rating of a UPS is the absolute max it can supply, so if your device only draws 100W normally, but does peak at 1000W for a very short while (like laser printers) then you do need an inverter that can do that 1000W.
The power line is very "forgiving" of loads requiring short bursts of power way beyond what they normally use, UPSs are not. TTYL
Thanks. How do you convert the rating of those devices in watts to the UPS rating in VA? If the current is not even close to a sine then it seems optimistic to expect V and A to be in phase, right?
Actually the typical computer power supply has a displacement power factor (phase angle) of close to unity. That has nothing to do with the fact that the waveshape is not a perfect sine wave.
The server power supply should have an amp rating. Use this, along with rated voltage, to calculate the VA.
As for run time; it should be stated on the UPS documentation (so much time at full load). It is true that you will get longer run times for lesser load, but use the rating at full load as a worst case. As the UPS ages (actually the batteries) your run time will decrease.
Is this the load your UPS will carry? One server? No monitor, no printer, no nothing ever? Also please tell me how you calculated the 30 seconds number - I'm interested in what figures you (we) have to work from. In fact it sounds like you may already have a UPS. If so, let's take this from the other way around. Tell us what UPS you have and all the ratings / nameplate data for it and tell us your load and we'll tell you are thoughts. I can't even believe I just typed 'are thoughts'. Our thoughts. Also, if there is some specific problem you are experiencing, it's best to tell us that. j
Yes, the Ah * V of the batteries will give you a rough idea of actual Wh available, remember though that there is a loss with the inverter, so derate what value you get, I'd say to about 75%. TTYL
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.