1
\$\begingroup\$

IT guy here - I have a device called a WattBox in one of my networks so I can remotely cycle power on an outlet. It also provides me with alerts when something falls outside set parameters.

I've been getting amperage spike warnings from it - normal draw is ~3.2A, and it'll spike up to 20.6A for ~10s, and I'm not sure why.

All of the equipment plugged into it is configured to alert me if something goes wrong; I've confirmed that nothing's power cycling or otherwise functioning abnormally. The WattBox doesn't give me per-outlet power consumption readings.

The WattBox is fed by an APC UPS rated for 1000W/1500VA load and 12A input. At any given time we're sitting around 40% load on the UPS.

The WattBox support rep suggested I get the outlet tested by an electrician and also that the power spikes are 'coming from the UPS,' which seems unlikely to me.

My understanding is that you can mess with voltage all you want on the source and that might mess things up, but that amperage is based on draw, on the side of the receiver. In other words, you can't "send" more power than something draws; there has to be a vacuum to fill. So, an amperage spike would represent a device pulling more power. The exception being if you hike up the voltage enough, it would "push" so hard that something breaks, and then yes, you've got an open gate for energy to flow through.

Is my understanding incorrect, and is it possible that the UPS is actually at fault here?

My concern is that the issue is the WattBox itself. All of the most critical network gear is plugged in to it, so it represents a single point of failure.

Part of why I don't think it's any of my network gear is because these amperage spikes happen during off-hours when everything should be as idle as possible, and also because I'm not seeing any alerts, errors, or odd behaviour showing up in logs at the times I'm receiving the amperage notifications.

And, with something that's difficult to reproduce like this issue, how would one go about isolating it down to a specific WattBox outlet if it is my network gear?

\$\endgroup\$
  • \$\begingroup\$ I can't imagine anything drawing an extra 17 Amps for 10 seconds without blowing a fuse. \$\endgroup\$ – Oldfart Aug 28 at 18:26
  • \$\begingroup\$ @DaveTweed the UPS is probably the only piece of equipment that I don't have proper access to. I'll have to find it if I can pull logs from it. No breakers are tripping, and the UPS's audible alarm isn't being set off. \$\endgroup\$ – Harv Aug 28 at 18:36
  • \$\begingroup\$ @DaveTweed along this line of thought, given normal load is 3.2A, is it safe to assume if any one piece of equipment drew 17A for 10s, assuming the UPS could supply it, there would be a fire/dead equipment/blown fuses/smoke? I'm trying to ascertain the effects of that much power for that long on equipment that, all together (more than 12 devices, some of which are drawing 40W+), draws only ~3A. I had no frame of reference for how much energy these numbers mean. \$\endgroup\$ – Harv Aug 28 at 18:46
  • \$\begingroup\$ @DaveTweed I confirmed the UPS is rated for 12A. I had two separate and dedicated circuits pulled into my server room specifically for the rack, but I'm not sure if they're 15A. However I see the logic behind what you're saying now; for 20.6A to be pulled, the UPS fuse would blow, and likely the main circuit breaker, as well as any fused equipment plugged into it. It's more likely that the WattBox is giving me bad information. Thank you for your comments. \$\endgroup\$ – Harv Aug 28 at 20:18
  • \$\begingroup\$ Turning OFF an inductive load can cause a voltage spike - possible a very substantial one if it is badly designed. Equipment which runs in off/on cycles such as a refrigerator or thermostat controlled appliance are potential sources. How your monitor would see such spikes is unknown and the MAY be reported in unexpected manners. | aDDING A NOISE FILTER BETWEEN THE SOURCE AND MONITOR MAY HELP. aND/OR ON LINES TO APPLIANCES IT FEEDS. iF AVAILABLE A 1:1 ISOLATING TRANSFORMER can act as a spike filter. [Caps unintendd :-) ]. \$\endgroup\$ – Russell McMahon Aug 29 at 7:33
2
\$\begingroup\$

20.6 A would be a major overload for the UPS. If the UPS isn't reporting anything unusual -- and circuit breakers aren't tripping -- you have to entertain the notion that the reports you're getting from the WattBox are simply wrong.

20 A is almost 2500 W, so yes, something would be getting hot in that amount of time! Also, I'm assuming that the whole setup is plugged into a wall outlet that's only rated at 15A.

\$\endgroup\$
1
\$\begingroup\$

Some things you have said are not correct. I think you are mixing up voltage ratings and current ratings (aka current capability) for sources.

For a fixed voltage source, the load determines the current draw. But if you increase that voltage more current will flow in the load. The current capability of that voltage source will not have any role in how much current flows into the load other than whether it is within the capability of the source or not.

Think about drawing water from a tank and forcing it into the hose under pressure. As long as the pressure (voltage) remains the same, the diameter of the hose (the load) determines how much water will flow. But if you increase that water pressure more water will flow through the hose. The size of the tank and its outlet/mouth (loosely current capability) do not play a role in determine how much water flows through the hose so long as they can support the flow rate (i.e. as long as they are not the bottleneck).

\$\endgroup\$
  • 2
    \$\begingroup\$ This isn't really correct either. It's only true for a resistive load, which computer equipment is decidedly not. The switching power supplies in computer equipment are typically constant-power, more or less -- if the voltage rises, their current draw will FALL and vice versa, within the range of voltages they can tolerate. In general the real answer is "voltage and current in a circuit are related by the intersection of the VI curves of the source and the load"; anything more specific is an approximation that works for specific sources and/or loads. \$\endgroup\$ – Glenn Willen Aug 28 at 18:26
  • 1
    \$\begingroup\$ @GlennWillen Perhaps, but I'm not really sure how to clear up his misunderstanding without pulling that in since he seems to be mixing up a question like "will hooking up my load rated for 10A to a battery rated for 20A destroy my load?" with voltage. \$\endgroup\$ – DKNguyen Aug 28 at 18:28
  • 1
    \$\begingroup\$ This answer and its comments helped clear up some of my confusions greatly, thank you. \$\endgroup\$ – Harv Aug 28 at 20:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.