ok I did some calculations, at 1.2MΩ, you get around 100µA which I find way to low to make a difference.
Using the 24kΩ (as stated by @steve.maddigan), you get 5mA (600mW of heat dissipation) , this seems a lot better without over doing it, using a 3 Watt resistor is on the safe side if you consider that 120VAC has a peak voltage of around 170 and gives you 1.2W, so at least 2W resistor will do the job perfectly.
So I guess I'll only be getting some 24kΩ resistors or something in that ball park, I'll go see my local store that will most likely have something in that range, thanks for the info @steve.maddigan. 1$ of parts and 2-3 minutes of time and voila, bypass and more money in the pockets to get more HA stuff (don't tell the wife).
Thanks for the detailed post. Am I correct to assume that the set-up shown in the pictures above is for an install where there is a neutral in the switch box. Would you still use the same resistor if you were wiring a dimmer switch without a neutral? If so, where and how would you place the resistor.
Dimmers that don't require a neutral have a special circuit or TRIAC that let's some current flow, this is the path where the circuit board is powered. Dimmers that require a neutral wire will not have this leak to power the electronics, it's using the neutral for that. Then there's the hybrid that can do both, some will leak even if a neutral is connected because it's a feature built in the TRIAC and can't really be truly called a neutral required alternative since it will still leak current and have this possible problem.
Other thing to note is that this can occur with some LED lights and not with others, it all depends on how the LED light was built. So some places you will see a glow, and some others not. Your mileage will vary and believe me, it has nothing to do with how much you pay for the lights,
Thanks for the explanation, it is a little above my head. With no wire hooked-up to the neutral terminal on the switch, I find it hard to understand how placing any resistor between the load terminal and the neutral terminal would have any effect on operation, where is the path for the electricity to flow?
I assumed (most likely incorrectly), that a dimmer installed without a neutral would just have a small amount of current go through the switch and the fixture. This small amount of current would be enough to power the zwave electronics, but not enough to turn on the light fixture.
Your assumptions are right on the money, the problem is that the small amount of current flowing through the light was not a problem with incandescent bulbs because all they are in reality is a resistor. But since we almost all have LED lights (since they were almost all banned in Canada except for stoves and other places that can't use LEDs), this small current flowing can excite the LED enough to make it turn on and glow just enough to be annoying. Some LED lights have a sort of bypass when you don't have a minimum of voltage going through the, that makes them go completely off all the time, but most don't.
So basically that resistor goes in parallel with the lights, either at the switch (when neutral is present) or the fist light in a string when using multiple lights (because there will always be a neutral at that first fixture) to make sure that current goes through the place with the lease amount of resistance and stop the LEDs from glowing.
The resistor should always be place between the switches hot and the neutral. If there is no neutral in the switch box then the resistor can be placed in oct box of the light fixture.
Thanks for your help. I have no experience with resistors, but am I correctly in assuming the resistor you mentioned above would draw a constant 0.6 watts at 120 volts (ohm's law) and not the 3 watts posted on the package?
1000/0.6 = 600 hours to use 1 kilowatt of electricity
1 year you would use 14.6 kilowatts or in Manitoba $1.46 of electricity. (.10 cents per kilowatt hour)
The load resistor will only be burning this much power when the switch is fully on.
When the switch is off it pulls the RMs down to below 20v rms (it’s not symmetrical but is less than 60Vp-p). So when the switch is off it’s a sixth of this value.
Yeah we're definitely exploring that now as this has come up multiple times -- hopefully once funding comes in and we're able to revamp the line, we will do this.
I had wondered why manufacturer's did not do this already as well. It would DEFINITELY need a warning that it would only work on switches that have a neutral in the box. I am betting that would cause a bit of confusion if the switch was not one that required a neutral. (Basically, most people only need it if you don't have a neutral. In that case, a load on the switch would not work because there is no neutral to be able to put it in parallel to the load)
If you are doubling the voltage and still want to maintain the same ‘bleed’ current of approx 5ma then you need to double the value of the resistor - meaning 48k ohms.
In this case the same power rating of 3W would be acceptable. Resistor wattage is rated in open air so when you place it in a closed box you want to derate it a bit. Ultimately you don’t want a really hot load resistor in the box.
If you find that 5ma is not sufficient to keep your switch in a stable state, then you need to reduce the resistance value until you reach stability.
Bottom line - use a power resistor with a wattage rating of at least double of what you calculate that you need. Test on your bench first in open air to get your solution. With the load on, if you can’t keep your fingers on the power resistor because it’s too hot, then you need to increase the power rating.
-- hopefully once funding comes in and we're able to revamp the line, we will do this.
