Winter here in Northern California is approaching and I was pondering about keeping my  portable spa filled and running or shutting it down until Spring. It rarely goes below freezing in the winter here. but it’s often really wet and rainy.

Being the techno nerd that I am, it seemed like a good time to verify the safety of the spa electrical service by checking the operation of the GFCI (ground fault circuit interrupter). All GFCIs have a little test button, you press it and they should trip. Yes my spa GFCI has a test button, and yes it trips when I press it. It’s good to do this test on a regular basis, like every month, as water and electricity can be a fatal combination.

After pressing the test button and resetting the GFCI, I didn’t exactly feel any safer. Did this test actually prove the spa was protected from electrical ground faults? Well no, it just proved the GFCI could trip when the test button was pressed, and verified the internal operation of the GFCI. It didn’t verify the grounding to the spa or what the current setting actually was or if it worked correctly on both hot leads of split-phase 230V service. I first tried to find a commercial GFCI test tool that would work on a spa GFCI. I hunted in a variety of tool and industrial supply catalogs and websites, but the only thing I could find was a multi-thousand dollar sophisticated device possibly used by GFCI manufacturers, or a $25 self contained plug-in tester for GFCI outlet’s. Not to be deterred, I looked into what simple tests could be done, that didn’t require a fortune in equipment.

If you understand how to deal with electricity, and potentially fatal electrical circuits, there is an easy way to verify the GFCI operation in more detail. If you don’t know how to safely deal with live electrical circuits, then you will just have to trust the test button on the GFCI or find someone who can do more tests.

A little background on spa GFCIs, which are designed to protect people from hazardous electrical shocks. To be safe, a GFCI should trip with an imbalance of 4-6 milliampsflowing between anything other than the hot and neutral wires, or for split-phase 230V service, between the two hot wires. Here in the US, split-phase 230V electrical service is very common, withtwohot wires, a neutral wire, and a ground wire.  A spa might have a 230V heater, connected across the two hot wires, and a 120V pump connected across just one of the hot wires and the neutral wire. A spa GFCI needs to not trip if the 230V heater is drawing 17 amps from both hot wires and a 120V pump is drawing 10 amps between one hot and neutral. This would give a total of 27 amps flowing through one hot wire, 17 amps through the other hot wire and 10 amps flowing though the neutral wire.

The danger of electrical shock happens when current from either hot wire flows to ground, either via the 4th ground wire, which should be attached to any metal parts of the spa pump/heater assembly, or via a path through the earth(which includes water pipes, or ANY other conductor attached to ground, like the metal case of a power tool plugged into a grounded outlet, or your wet bare feet on the ground).

What we need to actually test the operation of the GFCI is to create an imbalance in the current flow on each of the hot wires. An easy way to do this is to connect an electrical resistance of the correct value between the ground wire and the hot wire on the spa. Both hot wires should be tested separately.

The question to be answered is: how much of a current imbalance is needed to trip the GFCI as seen from the actual spa. Given a specific voltage, 120V here in the US, a formula know as ohms law allows us to calculate the resistance needed to allow a specific current flow. By creating a controlled current imbalance we can measure the exact point the GFCI will trip. Doing the ohms law calculations, we find resistance in the range of 10,000 to 50,000 would be a good test range.

It turns out, a very appropriate adjustable resistance happens to be available at any local Radio Shack store for $5. I purchased  Radio Shack part number 271-1716 50K-Ohm Linear-Taper Potentiometer. I already owned a digital volt-ohm meter (DVM), which also can measure AC current in the milliamp range.

The potentiometer has three terminals, and you should think of the two outer terminals as the ends of a 50,000 ohm resistor, and the center terminal as a variable resistor that taps into the 50,000 ohm resistor at a variable place depending on the rotation of the shaft. I ran a wire to the center terminal and another wire to one outer terminal. You can then use the resistance measurement function of the DVM to set the variable resistance to a specific value using the below table to map resistance to current flow at 120v.

Current at 120V

Resistance

2 ma

60,000 ohms

3 ma

40,000 ohms

4 ma

30,000 ohms

5 ma

25,000 ohms

6 ma

20,000 ohms

7 ma

17,142 ohms

8 ma

15,000 ohms

9 ma

13,333 ohms

10 ma

12,000 ohms

So now the actual test procedure:

  1. Turn the potentiometer shaft to it’s maximum variable value of 50,000 ohms, as measured with the DVM across the two wires connected to the center and side terminals
  2. Configure the DVM to measure AC current (or if you don’t have a current function on your DVM, skip to step 4)
  3. Connect one wire from the potentiometer to one of the DVM leads
  4. Connect the other wire from the potentiometer to an appropriate ground point on the spa. On mine there as a large grounding post on the metal frame
  5. You now should have the variable resistor, set for 50,000 ohms in series with your DVM set for current measurement. Be sure to switch the DVM from measuring resistance to current, as it may be damaged otherwise.
  6. Open the electrical junction box on the spa, to allow access to the power leads. My spa requires removing two screws, and then there is a nice organized plastic structure where the power from the GFCI comes in. Don’t touch anything, as these are live electrical circuits (if you don’t already know this, you should probably not be doing this).
  7. Touch the lead from the DVM (or directly from the potentiometer) to the hot power terminal. This should cause a little more than 2 milliampsof imbalance in the current flow through the GFCI. It may trip the GFCI, but it also may not. If your spa has split-phase 230v power, do this test on each of the 120V line terminals. If your GFCI tripped, reset it and try the other split-phase line. If it trips on both, or you only have a 120V spa, you have proven the GFCI trip current is about 2 milliamps or less.
  8. Assuming the above test didn’t trip the GFCI, with the lead not connected to the hot line, turn the potentiometer shaft a small amount (like a really small amount), reducing the resistance a bit. Touch the DVM lead to the hot terminal again, and read the current flow on the DVM. At a maximum resistance value of 50,000 ohms, the current flow should be a little over 2 milliamps. Keep cycling back and forth, adjusting the resistance slightly lower each time, and touching the lead to the hot terminal. Each time you test it, the DVM may show 2.5 then 3 then 4 then 5 millimaps flowing. Around about 6 millamps, your GFCI should trip. If the current flow get up to 12 milliamps or so, your GFCI is NOT working correctly, even if it’s test buttons says it is. Do not keep reducing the resistance as the potentiometer can’t handle much current and may be damaged or catch fire if the GFCI is totally malfunctioning. Do this test on each of the hot lines for split-phase 230V spas. My spa GFCI trips between 5 and 6 milliamps, which is in the correct range.
  9. If you happen to have something like a digital oscilloscope, you can potentially also measure the time it takes for the GFCI to trip. Tripping within 25 milliseconds on about 5 millamps of imbalance would be typical.
  10. If your DVM doesn’t have a current function, you can still do this test. Slowly reduce the variable resistance, and when the GFCI trips use the resistance measurement function on your DVM to measure where it was set, and calculate or use the table to see what the current flow would have been at that resistance. If not using a current meter, you should be very careful as it will be harder to tell if your GFCI is malfunctioning. You might want to determine beforehand what shaft position will give about 12 milliamps, and not exceed that. You can also use a set of fixed resistors arranged in series and parallel to create different resistance values. The simplest testing would be to just use 20,000 ohms and verify your GFCI is tripping at 6 milliamps or below. A pack of 10,000 ohm resistors is Radio Shack part number 271-1335, and two of them in series will give the appropriate value. If you do this often, like as an electrician, you might just permanently wire some resistors to some test leads, specifically for testing GFCIs.
  11. After testing, close the electrical junction box back up.

If you have done everything carefully and correctly, you will know just how much current imbalance is required to trip your GFCI, and if you have split-phase power will know each of the phases is also triggering at a correct value. You will also have used the grounding wire to create the current imbalance, verifying the grounding wire is connected to the spa and at the other end. This will not verify the grounding wire is connected to ground and not neutral, which would be an incorrect installation.

For me, knowing my spa GFCI trips at just over 5 milliampson each of the split-phase lines, and ground is good, is a lot more reassuring than just pressing the test button and seeing it trip. Considering how common spas are now, I’m surprised I can’t get a little microprocessor controlled GFCI test device at Home Depot or Lowes. Doing a more involved test is certainly a lot more work than just pressing the test button on the GFCI, but I’ve seen the test button work on a wall outlet GFCI and yet if you plugged a self contained GFCI tester into it and pressed the test button it doesn’t trip. For something like a spa, where it’s guaranteed to be operating under wet conditions, I do feel a bit safer after I spent the 45 minutes to do a more complete test.  Now that I know how, and have the couple of parts needed, I may make it a yearly part of my home maintenance.