Forward Biased

An AC mains filter and quality monitor, part 2

Alrighty, then.  This first drawing is Joost Waegebaert’s original filter design, as published in Elektor, so that you can see where we’re coming from.

After the AC connector, switch, fuses, and single-mode varistor comes the shunt differential-mode filter, consisting of six components across the line, including the capacitors between and after the transformers.  These mainly operate on anything from about 1KHz up, which eliminates most of the noise while being far enough away from the line frequency to leave it completely alone.

You can see how the final transformer is wired:  the primary winding is the power output, and it has the PE conductor connected to the center tap.  This causes any common-mode signals at the two “ends” of the primary winding to be of opposite polarity so that they cancel each other out.

The actual line power is differential-mode, not common-mode, so that it does not cancel itself out!  That would, obviously, be a Bad Thing, and I like to avoid Bad Things whenever I get the opportunity to do so.  This, by the way, accounts for why I did not connect the line and neutral conductors together after the fuses so I could watch all the pretty sparkles when the fuses blew to spectacular little smithereens inside their glass cartridges, which, of course never shatter during such an event, making it intrinsically safe not to wear safety glasses, right?

 

An AC mains filter and quality monitor, part 1

I’ve decided to “quickly¹” describe the AC line/mains filter for you, since we’re not including it in the lab power supply, and because it’s a really good one that overcomes the shortcomings of commercially-available filters—which I almost made the mistake of using.  Since I’m going to build mine with a line voltage and simple quality (THD+N) monitor, I’ll post the monitor circuitry soon and show how it should be interfaced to the filter, since differential-mode interference is shunted to ground (so that merely measuring “before” the filter will give the same result as “after”).  Common-mode interference is handled differently—the second of the back-to-back transformers is used as a common-mode choke.

I probably don’t have to tell you that the sort of quality-monitoring we’re talking about is far from the level that a Fluke 435 could provide without breaking a sweat.  This will be a reasonably-accurate readout of total harmonic distortion plus noise (THD+N) taken after a simple but carefully-trimmed 60- (or 50-) Hz T-notch filter, along with an equally-well-trimmed 3.5-digit voltage readout—reading true RMS, of course, not simply an averaged value pre-calibrated for a correct RMS reading with a sine wave at that voltage and frequency (i.e., cheating).  A few simple component value changes provide equal accuracy for 50Hz, 60Hz, 230VAC and 115VAC, as if you couldn’t figure those out for yourself, probably.