RFI/EMI Filters
Edited/Updated: Dec. 24, 2004 & 8-9-11
Terry Filter & Components
This is supposed to be the absolute best protective system for neon sign transformers, which have a nasty habit of failing in TC circuits
Schematic - http://hot-streamer.com/TeslaCoils/Misc/NSTFilt.jpg
L100J1.0K, 2 per assembly, power resistor, 1.0 k Ohm, 100W, Ohmite,
CFR-50JB10M, 12 per assembly, carbon film resistor, 10 m Ohm, 1/2 W, Yageo,
ECWH16332JV, 12 per assembly, capacitor, 0.0033uF, 1600V, Panasonic,
ERZ-V10D182, 12 for 9 kV, 14 for 12 kV or 16 pcs for a 15 kV Trans-orbs (transient surge absorbers), Panasonic.
I may be building some of these Terry filters and will put them on my For Sale page if that happens - John C May 8, 2011
RC Filter, http://hot-streamer.com/greg/filter.htm-
For a cheap RFI filter, BG Micro, www.bgmicro.com, has a 3 phase unit rated at 30 amps per phase (90 amps if you parallel them) for something like six dollars.
-
Yes, use a RF line filter. Also, use an RF "hash" filter in series with each line. This is 80 turns of # 12 AWG wrapped around a 2 inch ID PVC tube. Tube is filled with 1/4 inch dia. steel/iron rods --- epoxy or silicone in place (another hash filter is just 60 turns of #14 AWG house wiring wound around a 2 inch ID PVC pipe. Silicone some steel rods, or bolts with heads cut off, into the PVC tube). Corcom filter plus this hash filter will keep everything out of the AC line. Corcom (recently bought by Tyco), has a B series 20-amp model (cheapest series); the series that filters all the way down to 10 kHz is the T series, but it costs quite a bit more (~$88 vs. ~$23)). You can buy surplus Corcom's from C&H Sales, Pasadena, CA or Fair Radio Sales, Lima, OH, at around $10 each. Dr. R.
-
Home-automation products marketed by companies such as X-10 depend on power-line-transmitted RF signals occurring near mains-voltage zero-crossings. Electronic apparatus incorporating switching power supplies can put EMI on the power line that will interfere with those low-level signals. I found that that's what my computer was doing. X-10 markets a filter (their model XPPF) that is supposed to take care of that kind of problem. Finding that it did not in my case, I made up a simple trap that works very nicely and which might do the same for other coilers.
The X-10 products send and receive their signals within 200 us of the zero-crossings. With the nominal 160 V peak amplitude of (U.S.) mains voltage, about 4 V of amplitude is reached at 200 us. What the trap does is to keep the mains-circuit to the load essentially open for that 200 us by the simple expedient of connecting a string of back-to-back diodes in series with each side of the mains circuit. I connected 3 diodes in series, in each side, paralleled by another 3 reversed. That provides the "dead band" of about +/- 4 V. I added 0.15 uF capacitors to ground at the load-side of the diodes to further reduce the EMI. The trap is, of course, too good by half: It not only blocks the EMI during the interval but also the mains voltage itself. But not to worry: with mains voltage at nominal, most electronic equipment should not be adversely affected by the loss of 2 1/2% of the peak. Ken H. -
June 10, 2004. EMI Filters are low pass filter circuits. If you parallel them, you can double the current capacity of your supply, but you will halve the EMI protection (i.e., double the amplitude of EMI that will pass the assembly). If your series them you double the power dissipated within the filter, but you will halve the amount of EMI that would get through one filter. For most situations, connecting filters in series gives no great advantage. You can parallel them if you don't have enough current capacity with one filter alone, but you're probably better off getting a larger filter that can handle the desired current. HTH, Dave.
-
June 30, 2004. Even though pigs are built extra tough, it's a good idea to add 40-50 Ohms resistance in series with the output. This acts as a "Q-spoiler" and clips most harsh transients before they get into the sec windings. Use 4-6 pcs of 200 Ohm, 200 watt resistors in parallel to get down to the 40 Ohm value. Right at the input to the xmfr box we also use a 100 turn x 2 inch dia. PVC tube as a HF transient clipper before it hits the resistors. This an
air-core choke. This holds off the nasty high freq. RF that will crawl all over your HV sec windings looking for a ground. Many pole xmfrs will live through this but some die. Why take the chance? Dr. R.