1

I wanted to test how much power a GFCI outlet uses while idle, so I hooked up my multi-meter to test the current flow. I put my test lead in the amp socket, and confirmed the current was on the order of a milli-amp, so I was safe to plug my lead into the milli-amp socket.

After doing so I measured a current of a few milli-amps. But then suddenly through testing I got a sudden current spike, and blew my 400mA fuse. I'm not sure what caused the spike... maybe tripping the GFCI, etc, but some GFCIs seem to operate this way.

Is there some way I can even this out, and absorb whatever the spike current is for that sub-second that it spikes protecting my fuse and multi-meter, but still be on the milli-amp scale? I know conceptually this is what a capacitor does, but I also know you can't simply hook a capacitor up in series and expect to get any current flow.

I'm thinking maybe there's some simple circuit, or inexpensive device I can hook up between the GFCI and my multi-meter that'll prevent the fuse from blowing, but I don't know enough about electronics to know if this is possible, or where to even begin.

EDIT: Another possibility is some sort of solid-state, fast trip 300ma circuit breaker that I put in series, before the multi-meter that'll trip before I blow the fuse? Does such a thing exist?

EDIT2: How about something like a Inrush Current Limiter? https://en.wikipedia.org/wiki/Inrush_current_limiter

On the face of it, it sounds like the solution to exactly the problem I'm having.

Steve Sether
  • 111
  • 3
  • how do you run something like a vacuum cleaner if the power outlet only provides few milliamps? – jsotola Sep 28 '20 at 18:42
  • There are slow blow fuses which can tolerate short pulses above their ratings. I would never suggest using a higher rated fuse, say 1A, in such a role. –  Sep 28 '20 at 18:55
  • @jsotola I'm measuring the power/current usage of the actual GFCI, not the curent flowing through some sort of load hooked up to the socket. The GFCI itself uses a little bit of power when monitoring the circuit for ground faults. – Steve Sether Sep 28 '20 at 19:04
  • @AndrewMorton I'd like to measure this myself. It's not particularly dangerous as long as you know what you're doing and take basic precautions. – Steve Sether Sep 28 '20 at 19:06
  • 1
    @BrianDrummond Wouldn't this put my multi-meter in danger? Whomever designed the multi-meter put in a fast-blow fuse for a reason. Presumably to protect the electronics from damage from say even a brief, 1 amp surge. – Steve Sether Sep 28 '20 at 19:08
  • It would to some extent. You could consider putting a couple of 3A diodes in parallel with the meter, (one each way round) along with the slow blow fuse. That would limit the voltage across the measurement circuit to +/-0.7V or so. A 1V overload (on what is the 200mV or 400mV range) should cause no damage to the meter. –  Sep 28 '20 at 19:14
  • go to your breaker panel ... turn off all of the breakers except the one feeding the GFCI outlet ... make sure that all loads are unplugged from the circuit ... then watch your power meter – jsotola Sep 28 '20 at 19:32
  • @jsotola The power we're talking about is perhaps .5 watts. That's going to be difficult and long to measure. It's also a very impractical way to measure power usage since I have to turn off the power to the entire house. – Steve Sether Sep 28 '20 at 19:38
  • 1
    "*... since I have to turn off the power to the entire house.*" Did you not do that before you inserted the meter? Did the fridge turn on and blow your meter? I think we need a schematic of how you hooked all this up and maybe a photo too. Hit the [edit] link ... – Transistor Sep 28 '20 at 19:57
  • You can add in a schematic using the CircuitLab button on the editor toolbar. Double-click a component to edit its properties. 'R' = rotate, 'H' = horizontal flip. 'V' = vertical flip. Note that when you use the CircuitLab button on the editor toolbar and "Save and Insert" on the editor an editable schematic is saved in your post. That makes it easy for us to copy and edit in our answers. You don't need a CircuitLab account, no screengrabs, no image uploads, no background grid. – Transistor Sep 28 '20 at 19:58
  • @Transistor There's no fridge or anything plugged in to the GFCI outlet. This is simple enough that no diagram is necessary. 120V Hot -> multi-meter -> GFCI. Neutral and ground of the house current are plugged into GFCI outlet. There's nothing attached to the load terminals of the GFCI, and they're taped over. This is simply a matter of something causing a surge. The first GFCI I tried didn't create a surge, so it must be something of the design of that particulat GFCI. – Steve Sether Sep 28 '20 at 20:25
  • A GFCI may have a built-in self test, which operates at power-up and then every few hours. That could be the surge you see. Rather than measure current, you could measure voltage across a known resistor (e.g. long lead wire) and shunt the DVM (in V-mode) with counter-facing Schottky diodes. I have not used diode protection in this arrangement, so best to be cautious and let others better guide you. – P2000 Sep 28 '20 at 21:46
  • 1
    @BrianDrummond could you sketch what you have in mind? I have a hard time visualizing how the protection is happening, and I think there's something I can learn here. If you have time to submit an answer that would be much appreciated. – P2000 Sep 28 '20 at 21:47

1 Answers1

1

I think you're barking up the wrong tree.

You are attempting to view a momentary surge with a common DVM. They're not made for that.

You would be better off installing an ammeter shunt, and then looking for millivolt rises across the shunt. Further, you should use some sort of data recorder that will sample at a very high rate, so you can observe the transient.

Most likely this was the GFCI device's internal electronics initially charging their capacitors. It could also be some part of the startup testing that some GFCIs do. For instance some are able to detect neutral-ground faults on power-up; they can only do that by attempting to flow some current on neutral. Whatever it is has to fit inside a recep form-factor.

Harper - Reinstate Monica
  • 21,639
  • 27
  • 68
  • I don't really care about measuring the surge. I'm just interested in how much power these things consume over minutes, not the first initial milli-seconds. Even some sort of cheap, sacrificial fast-blow breaker, or even I guess some cheap fast-blow fuses that cost a few cents would be OK. The fuses for the multi-meter are about a $1 or $2 a piece, which, while highway robbery is what they cost. I"m using the eevblog meter, so this isn't a cheapy from the big box store or harbor-freight. I also don't really want to buy a bunch of expensive equipment to do something relatively simple. – Steve Sether Sep 28 '20 at 20:38
  • I think the right approach might be just to avoid the surge in the first place. Hook the hot wire to a switch, and the other end of the switch to the hot of the GFCI. Power up the GFCI by flipping the switch. Hook the multi-meter between the hot and the GFCI, Turn off the switch. That should charge any capacitors, or get past any tests the GFCI does on initial start, and measure the current I'm interested in. – Steve Sether Sep 28 '20 at 21:09
  • @SteveSether or measure voltage over known resistor / long lead wire (see my comment above) but apply DVM in V-mode after a few seconds. Then you don't have to interrupt power to the GFCI to apply the DVM in I-mode. – P2000 Sep 28 '20 at 21:55
  • Right. That would be the best approach @SteveSether. You only need to shunt the meter leads until the GFCI powers up. – Harper - Reinstate Monica Sep 29 '20 at 00:15