I've wired a circuit like the one at the top of the attached image. The wall switch is intended as an override on switch. I was not concerned about combing the wall switch leg with the relay leg because the source for both legs is the same 12/2 cable, and neither leg includes significantly capacitive or inductive components which would cause the voltage between the two legs to be out of phase.

I think I may have been wrong. I have observed 100VAC across the wall switch (between points A and B) when both the relay and switch are open. When the relay is closed and the switch is open, the voltage is 0 as expected.

Do you agree with my assessment that the voltages are out of phase between the two legs? I think something like this (https://www.desmos.com/calculator/gkqrbq72io) may be happening. However if that were the case, I would not expect the voltage to be 0 when the relay is closed.

Is this a problem?

The circuit at the bottom of the attached image seems like a good alternative. Unfortunately my local hardware store didn't have any 20A 3-way switches.

 

ac measurements on open circuits can be confusng, if you put a load across a and b, does the 100 folts still show up? a small load, like a night light.

 

I don't see any problem. Can you clarify what is the problem?

 

I've wired a circuit like the one at the top of the attached image. The wall switch is intended as an override on switch. I was not concerned about combing the wall switch leg with the relay leg because the source for both legs is the same 12/2 cable, and neither leg includes significantly capacitive or inductive components which would cause the voltage between the two legs to be out of phase.

I think I may have been wrong. I have observed 100VAC across the wall switch (between points A and B) when both the relay and switch are open. When the relay is closed and the switch is open, the voltage is 0 as expected.

Do you agree with my assessment that the voltages are out of phase between the two legs? I think something like this (https://www.desmos.com/calculator/gkqrbq72io) may be happening. However if that were the case, I would not expect the voltage to be 0 when the relay is closed.

Is this a problem?

The circuit at the bottom of the attached image seems like a good alternative. Unfortunately my local hardware store didn't have any 20A 3-way switches.

View attachment 74480

When the switch is open between points A and B and the relay contact is also open, you should be able to read 120 VAC due to the fact that no current flows. What is wrong with this?

 

When the switch is open between points A and B and the relay contact is also open, you should be able to read 120 VAC due to the fact that no current flows. What is wrong with this?

The problem is that I measure 100 volts, not 120 volts.

There is a varistor (http://www.digikey.com/product-detail/en/MOV-14D201K/MOV-14D201K-ND/2799113) between neutral and ground, as well as the motor side of the relay and ground. Perhaps they are contributing to what I've observed; I plan to investigate that tonight.

 

I've removed the varistors from the circuit, and still measure 100VAC across both the wall switch and the relay. It has occurred to me that "hot" side of the motor should be floating when not connected, but I'm not sure if this is contributing to what I'm measuring.

I've updated the diagram with measured voltages.

 

What type of motor is it?

 

I've removed the varistors from the circuit, and still measure 100VAC across both the wall switch and the relay. It has occurred to me that "hot" side of the motor should be floating when not connected, but I'm not sure if this is contributing to what I'm measuring.

I've updated the diagram with measured voltages.

It seems like your meter is low impedance, hence the 20V difference between the readings (across the load and then in series with the load). What type of voltmeter do you use?

 

What type of motor is it?

Actually, the motor is conneted via a 20A GFCI wall outlet. For almost all of the measurements I've taken so far, the motor hasn't been connected.

It seems like your meter is low impedance, hence the 20V difference between the readings (across the load and then in series with the load). What type of voltmeter do you use?

It's a very cheap Craftsman digital mutimeter.

 

Actually, the motor is conneted via a 20A GFCI wall outlet. For almost all of the measurements I've taken so far, the motor hasn't been connected.


It's a very cheap Craftsman digital mutimeter.

If the motor is not connected, what else is? To put it differently, what is inside the GFCI wall socket? Perhaps a neon indicator? Is the wall socket part of the circuit when you make the tests?

 

If the motor is not connected, what else is? To put it differently, what is inside the GFCI wall socket? Perhaps a neon indicator? Is the wall socket part of the circuit when you make the tests?

Ah ha. There is a orange indicator light incorporated in to the GFCI outlet. It indicates whether the outlet is in an interrupted/test state or not.

Yes, the outlet has always been part of the circuit during the tests.

The neutral and ground are obviously always connected to the outlet. When both the relay and wall switch are open, the hot 120VAC wire is not connected to the outlet. The orange light is not on in this case, and I measure 100VAC across the terminals of the wall switch. If I turn on the wall switch, the hot wire is connected to the outlet but the orange indicator light does not turn on. If I then hit the test button on the GFCI outlet the orange light appears.

 

Actually, the motor is conneted via a 20A GFCI wall outlet. For almost all of the measurements I've taken so far, the motor hasn't been connected.

So you are saying that the motor is not actually part of the circuit?

Then you are measuring leakage current through the GFCI.

 

If you assume that the internal resistance of your digital meter is 1MΩ, then the GFCI is presenting a resistance of 200kΩ to give you a reading of 100V.

 

by the way, what is the application of this circuit?

 

Thanks for the help.

The application of the circuit is to control power to a 120VAC water pump. The pump takes water from several rain barrels and sends the water into an irrigation network for my garden. Irrigation happens automatically according to a schedule, but will only occur if it hasn't rained lately, and isn't forecasted to rain soon. This year on several occasions, the garden was being watered during rain storms which was wasteful of the water in the rain barrels.