Foot-operated dimmers

Foot operated dimmers. They're hardly a rare item. Last ones I have came (cheaply) from Ikea.

I guess they can be taken apart to wire in the old lamp.

What do these things (generally speaking, not these particular ones) do to the waveform of the current?

Thanks,

Daniele

A Google for 'foot dimmer' just worked for me. Though I don't want to dim my feet! :-)

They are known as phase dimmers - in contrast to the more sophisticated pulse width modulated power controllers.

The technique is to use an electronic switch known as a triac which is triggered into it's conducting state by a pulse to its gate. The pulse is progressively delayed for each half cycle of the mains as you dim the light so that less of each half cycle reaches the lamp. Hence less power to the lamp.

For an incandescent lamp, this is not too disastrous as the 'average' current isn't that significant but because the current at the switching point is into a large voltage there is an RF generation problem and filtering is therefore necessary. Equally this is not good for the power generators. This method of power control is therefore only used for light loads such as lamps.

Rob

So is less current drawn as it's dimmed? I guess so, because I don't see huge heat sinks on these things, which there would be if they were just shunting the current, though they often get quite warm, so there must be some waste.

Daniele

Yes they use less current when dimmed, but the reduction is not inline with the apparent brightness. So a lamp at 50% brightness may still be drawing 80% of the current.

I've not seen the Y shaped ones. Mine are in-line Koppla. I've some in- line Koppla that are thumbwheel operated too.

Less _average_ current is drawn. What they actually do (and why the heatsinks can be small) is to switch either fully-on or fully-off at all times (so theoretically no dissipation in the switch) and to do this more quickly than the thermal inertia of the filament, so you don't see it.

Old ones were phase-angle switches. They measured the rising voltage of the incoming sine wave and switched from off to on when the adjustable set-point was reached. The thyristors or triacs used for switching then couldn't be turned off afterwards, except by the supply voltage falling to 0, which happens twice every cycle (100 times a second). A very simple circuit, even for '70s discrete component construction.

The drawback of these is that they switch on _suddenly_ (100 times a second), which is always a recipe for electrical noise. A better solution was to use a "zero crossing detector" which only switches when the supply is at 0 anyway. The corollary of this is that they deliver (or don't) _entire_ half-cycles of the AC waveform, which quantises the average output they can deliver. If we assume that any change longer than 1/10th of a second can cause visible flicker, then that means 10 different discrete steps at most. In practice though this isn't a problem. These are the common sort today, just to reduce their radiated RFI. They're not PWM, but are very similar to it - they supply (or omit) many pulses in varying ratios, rather than PWM's constant pulse trains of varying width. PWM is quite rare in AC systems, except for slow-speed controllers for ovens.

I don't think Koppla exist any longer.

Daniele