Dimmers and how they work

[RonHebbard]

It's way more complex than P.F. in the old days a capacitive or inductive load would shift the P.F. and factories would have rooms of capacitors to correct it but with wave chopping and switch mode power supplies the wave forms are being distorted to hell and neutral currents can be more than per phase current, dimmers load up the beginning or end of the waveform and switch mode load up the middle, all in multiple constantly changing ways.

@ David Ashton Posting in support. This is a major reason why 3 phase feeds originally installed to power a bank of (for example) 42 x 6 Kw SCR dimmers had their neutrals catching fire when the original racks were replaced with a greater quantity of appreciably smaller dimmers; racks of 2.4 Kw SCR dimmers for example. Here in Canada inspectors were initially caught off guard then, over a period of years ( and several fires ) eventually moved to double, and in some cases, even triple neutrals once the effects of triplen harmonics became more widely understood. I recall one senior electrician thinking something had to be wrong with his equipment when he found 60 Hz on each phase but three times 60 Hz on the single neutral. Thanks for your post.
Toodleoo!
Ron Hebbard

 

[TimMc]

It's way more complex than P.F. in the old days a capacitive or inductive load would shift the P.F. and factories would have rooms of capacitors to correct it but with wave chopping and switch mode power supplies the wave forms are being distorted to hell and neutral currents can be more than per phase current, dimmers load up the beginning or end of the waveform and switch mode load up the middle, all in multiple constantly changing ways.

Guy Holt is a studio mechanic/grip/gaffer from the Boston area and he writes articles for Protocol. One of the topics he's covered in depth is the harmonic content of a variety of film/TV lighting and how that affects generators (among other things). I suggest those interested in what happens when the neutral carries harmonics up to the 13th read his articles.

 

[BillConnerFASTC]

It seems like conservative design intentionally or inadvertently solves most of these problems. I think of all the high schools I did with 2-3 racks of 96 20 amp dimmers and an 800 or 1000 amp feed - and they probably never added much above the 125-150 fixtures. Those all at full is not 50%

I look at all LED system today and end up with 40 to 60 - 20 amp switched stage lighting circuits often - and two 48 circuit panels with house and maybe some motorized hoist on them. 200 amp feeds. And maybe 120 CS fixtures. That's about 20%. And plan for no more than 10 fixtures on a circuit - that's roughly 50% per circuit.

I believe that is all good planning. It makes me wonder when I see an engineer or sales rep (or combo) designed system that is close to fully loaded at opening. DImmer racks and incandescents fed with 200 amps. Those are where problems and compromises result.

 

[JD]

It's way more complex than P.F. in the old days a capacitive or inductive load would shift the P.F. and factories would have rooms of capacitors to correct it but with wave chopping and switch mode power supplies the wave forms are being distorted to hell and neutral currents can be more than per phase current, dimmers load up the beginning or end of the waveform and switch mode load up the middle, all in multiple constantly changing ways.

So true. You can correct a leading or lagging load, but switch mode supplies leave you with that sharp chunk right in the middle of the waveform! One technique has been to use undersized supply capacitors so the cap voltage does a major sag and the diodes conduct earlier in the next waveform. Still can't go to zero, and I wonder how that effects cap life as E-caps hate big AC ripple.

 

[DrewE]

So true. You can correct a leading or lagging load, but switch mode supplies leave you with that sharp chunk right in the middle of the waveform! One technique has been to use undersized supply capacitors so the cap voltage does a major sag and the diodes conduct earlier in the next waveform. Still can't go to zero, and I wonder how that effects cap life as E-caps hate big AC ripple.

Many newish switch mode power supplies have active power factor correction, which as I understand things often works by using what is basically a switched inductor filter (typically with a supersonic switching frequency), controlled by some clever control circuitry to force the current to more or less track the incoming voltage. This does tend to add RFI that must be filtered, and in many cases doesn't cope very well with input voltage waveforms that are not sinusoidal (including, but not limited to, the output of a dimmer or most inexpensive DC-AC power inverters). They can give excellent overall power factors.

Such a circuit isn't particularly suitable for use on a phase control dimmer, or so it seems to me, since the dimmer relies specifically on the switching of the waveform mid-cycle rather than that being a side effect of the desired need to rectify the AC waveform into something resembling DC power as a switch mode power supply requires.