Michael Lane LC

archived articles

THD and Dimming Ballasts
November 1997

Okay, maybe I'm wrong, but didn"t we try to get rid of high THD (total harmonic distortion) ballasts about 5 years ago? They're all gone from the market aren't they? Well there back!

I feel like Chicken Little, you know "the sky is falling, the sky is falling". "THD is back, THD is back". Actually there is no need to be alarmed for general applications. However, there is a new bread of electronic ballast that has me concerned: the dimming electronic ballast designed to be controlled by an incandescent dimmer. In the past 2 years we have seen these ballasts from Solium (out of business), B+L, Advance, and Lutron. Ballast are available to dim the 4-pin compact fluorescent and the T8 lamp.

The idea of these ballast is great: just install them in an existing space and add an incandescent dimmer. No need to pull extra wires for the dimming control, it doesn't need any. Wham, bam, you've got a dimming system!

So what is the problem with these ballasts? Well at full light output the THD is a respectable less than 20%, but as you dim the ballast down to the low end, the THD can climb to around 100%! Actually it's not the ballasts fault; have you ever tried running a standard 75 watt A-lamp through a harmonic analyzer? It's prtey scary. At full light output the THD is around 25%, and at full dim the THD rises to over 160%. So there's the real problem, the incandescent dimmer is chopping the sign wave causing all kinds of distortion regardless of the lamp/ballast that it is controlling. With the fluorescent dimming ballast the problem is not the situation of a few ballasts in a conference or training room. Four or eight downlights on a dimming system will not add up to that much, even at 100% THD.

As an example, let's look at eight 32 watt CFL's dimmed down to 5% light output, producing 100% THD. The THD put back on the power distribution system is only 1.1 amps. An additional factor to consider is that the power factor at full dim drops to around 55%. However, a problem could arise in trying to dim an entire room of T8 lamps onthis type of dimming system. The following example assumes a 2000 watt incandescent dimmer, on a 120 volt system, with 22, 2-lamp T8 ballasts using around 70 watts at full light output. That's 1540 watts on the 2000 watt dimmer, or about 75% of the dimmers rating. At full light output the ballast produces less than 13% THD, or 1.63 amps of distortion.

Since ANSI C82.11-1993 electronic ballasts specifications allow 33% THD, there is no problem with this 13% THD. At full dim, 5% light output, each ballast still is using about .275 amps, or about 5.5 total amps on the circuit at 91.6 THD. Is 5.5 amps of distortion on the circuit a problem? Maybe of maybe not, it all depends on the number of circuits that are dimmed in the building, and whether a three-phase common neutral power system is used.

As for my data. I have evaluated copies of ETL reports on a 32 watt CFL, and a two-lamp F32T8 system. The CFL report show a THD of 12.2% at 100% and a Power Factor of 99%. At full dim the THD rises to 95.5% and the Power Factor drops to 55%. I also had the opportunity to see a competitors 32 watt dimming system in action a few weeks ago and their numbers were very similar to these ETL test report numbers. The test on the 2-lamp F32T8 system shows a THD of 12.7% at 100% and a Power Factor of 99%. At full dim the THD rises to 91.6% and the Power Factor drops to 50%.

Now, as I've said before, I'm not an engineer; my background is in architecture. Maybe I'm wrong here, and everything is okay with these ballasts? But I'm not convinced that this type of ballast and dimmer is the best solution for large area dimming. On the positive side, using a few of these ballasts in a conference or training room is not going to be a problem, and maybe is the easiest and most cost effective solution.