Dimmers Page 3

Dimmers are not without problems!

As mentioned in Dimmers Page 2, dimmers can create EMI and annoying filament rattle. In some cases with long power lines you can get power induction causing lights to illuminate that have not had power applied to them! There are a several of work a rounds - and as usual, the more effective ones cost more money!

Snubber - This is nothing more than a resistor and capacitor placed in series from MT1 to MT2 on a triac. Again the purpose is to slow down the rapid voltage rise on a triac. These are best used on inductive loads.

Chokes - A choke is a coil of wire, often wrapped on a ferrite core. The idea of a choke is to slow down the rise time of a triac turning on. Chokes have two values of concern, the inductance and the current they can carry. Most chokes on commercial dimmers are 3-4 inches in diameter, an inch or more tall and can carry 15-20 amps.

IGFET - Since the above problems come from the rapid rise of suddenly turning the triac on, what would happen if you could allow the voltage to turn on at the zero crossing and then turn it off when enough power has been used? That is the concept behind using an IGFET for dimming. The gentle rise of an 60Hz AC wave generates little in the way of EMI. So why doesn't everyone use IGFETs for dimming - cost. They are several times more expensive than the typical triac.

Another problem area for triac dimmers is that they need a minimum load to properly function. A mechanical switch will pass any current/voltage (up to its rating) that is presented to it. A triac is a solid state device that has a minimum threshold of voltage/current that must exist before it will turn on. In some low current situations I saw in my planetarium days, we would parallel a 40w light bulb to the low current device just to give the triac dimmer enough load to run. This problem seems to raising it's ugly head in the world of computer Christmas lighting, where strings of low low wattage LED Christmas lights are becoming more readily available.

Triacs do not provide a linear output. As mentioned before, what is controlled by a triac is power (voltage times current). The amount of power delivered by a triac can be determined by looking at the area under the curve of an AC wave form.

The graph to the left is power of one half cycle of a sine wave. The x-axis (time) is divided into 256 steps. What should be noticed is that when the wave first starts it is almost flat for the first 20 steps before changing into a nice straight slope, but as it reaches the end the line flattens back out again. Not at all a straight line.

Actual values look like this:

% Input (0-100)	% Power Out (0-100)
1	0.05%
5	0.68%
10	2.62%
15	5.48%
20	9.66%
25	14.86%
30	20.96%
35	27.79%
40	34.61%
45	42.36%
50	50.00%
55	57.64%
60	65.98%
65	72.75%
70	79.54%
75	85.57%
80	90.70%
85	96.54%
90	97.57%
95	99.41%
99	99.98%
100	100.00%

So what do you do? Using Excel, I worked out a table of values that give a linear output. A graph of that table is below.

This graph is sort of the inverse of the previous graph. At both the beginning and the end, a small change in input results in a large change in output - thus offsetting the normal power curve. The Excel spreadsheet I developed these numbers and graphs from can be downloaded from here.

January 2007 - I would not waste my time with the values in the table - experiments made measuring the light value with this table did not work out at all! The story continues on Dimmers Page 4.

Please address general comments to web@dv-fansler.com