X-Virus-Scanned: clean according to Sophos on Logan.com Return-Path: Sender: To: lml@lancaironline.net Date: Sun, 16 Apr 2006 00:28:44 -0400 Message-ID: X-Original-Return-Path: Received: from smtp113.sbc.mail.mud.yahoo.com ([68.142.198.212] verified) by logan.com (CommuniGate Pro SMTP 5.0.9) with SMTP id 1069541 for lml@lancaironline.net; Sat, 15 Apr 2006 19:28:52 -0400 Received-SPF: none receiver=logan.com; client-ip=68.142.198.212; envelope-from=elippse@sbcglobal.net Received: (qmail 67550 invoked from network); 15 Apr 2006 23:28:09 -0000 Received: from unknown (HELO Computerroom) (elippse@sbcglobal.net@71.148.2.252 with login) by smtp113.sbc.mail.mud.yahoo.com with SMTP; 15 Apr 2006 23:28:09 -0000 X-Original-Message-ID: <000501c660e4$417451f0$fc029447@Computerroom> From: "Paul Lipps" X-Original-To: Subject: Shielding X-Original-Date: Sat, 15 Apr 2006 16:28:05 -0700 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0002_01C660A9.930B5210" X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2900.2869 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2869 This is a multi-part message in MIME format. ------=_NextPart_000_0002_01C660A9.930B5210 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Well, let me contribute what I can from some of the things I've = learned. An electrical circuit requires two conductors; there is no such = thing as "ground" or "earth", although I've heard they found the = absolute reference ground at 0:00:00 Lat, 0:00:00 Long. If you throw = "ground" out of your vocabulary and instead substitute "signal return", = you'll have a much better time in wiring things, especially on = dielectric (plastic), as well as metallic airplanes. Every morning when = you arise, repeat ten times "There is no such thing as ground in = electrical circuits". That being said, when a signal goes from a source = to a load, there must be a way for that signal current to get back to = the source. A shielded wire, or a twisted pair, is a perfect medium to = do that, providing that the return is connected to common at the source = end. If it is not connected at the source end, then the current must = find its way back by some other path, often bringing with it other = undesirable currents, giving rise to common-mode coupling, so-called = ground-loops. These currents arise from the difference in potential, AC = and DC, between the source common and the load common. Obviously if the = load is tied to its common, the return conductor must be tied at the = receiving end also! This is where opto isolator circuits find good use = in keeping out unwanted signals; they only require being tied to the = opto circuit at each end without a common return connection between = them. One of the best things you can do to minimize interference is to = have a single point at which all power returns are connected. Keep the = main and common buss close together and run twisted pair or shielded = wires to each load. It also helps to have capacitors connected between = the main and return buss. Parallel a 4700 uF 35V with a 1uF 35V film and = a .01 ceramic disc across them. This keeps both the main and common buss = at the same AC potential. Under no circumstance should loads share a = return! Please don't "daisy-chain" returns from one circuit to another! = Down that path lies destruction, Alice! On my Lancair, I have two = terminal boards that are mounted horizontal and parallel, one below the = other, with a copper return buss below the bottom terminal board. The = return buss is made from flattened copper tubing with brass screws = through it at the same spacing as the terminal board terminals. The = upper terminal board is for the main, aux, and avionics buss. The one = below it is for load distribution. I run twisted pair from the main and = distribution terminals to each circuit breaker. Current flows from the = main, to the circuit breaker, then back to the distribution terminal. = That provides cancellation of the magnetic field in these conductors. = Then each load has a twisted pair from the distribution terminal and the = common. Here again, the currents go out and back over these wires = causing the magnetic fields to cancel, and all circuits are tied = together at one common point. An advantage of this scheme is that each = load's source wiring is easily found on the distribution board for = connecting, dis-connecting, and trouble shooting. Another thing you = might try is to mount an LED with each circuit breaker and wire it = directly across the breaker terminals with a series resistor, 1.2k 1/2W = for 14V systems, and 2.7k, 1/2W for 28V systems. Then if a circuit = breaker is open under a load, the LED will be on, showing that it's = open. Radio Shack and JameCo have snap-on ferrite interference suppressors = that can be placed over coax and wire bundles to cut down on = interference between circuits. They are tubular ferrites, about 3/4" OD, = 1/4" ID, and 1" long, sliced in half down their long dimension, and = secured in a hinged plastic cover that can be snapped in place over = wiring. They cost less than $2 each.=20 One more consideration is where to put the field circuit breaker if = an alternator circuit breaker is used. If your battery is run down and = you have to jumper it to start, and then shortly after you take off, the = charging current on the depleted battery could be very high, tripping = the alternator breaker. If your field breaker is connected to the main = buss, the regulator will sense the low voltage on the buss and feed = maximum current to the alternator field. With the alternator breaker = tripped and no load, the alternator will put out well over 100V! Then = when you push the tripped-breaker back on, the sudden inrush current and = high voltage could fry your electronics. A better place to connect your = field breaker is on the alternator-side of the alternator. That way, if = the alternator breaker trips, the alternator field will still be = connected to the alternator output and keep its voltage under = regulation!=20 For a much better look at the subject of shielding and grounding, = try "Interference Handbook" by W.R. Nelson, WA6FQG and "Grounding and = Shielding Techniques in Instrumentation" by R. Morrison. Morrison has = good illustrations of how interference currents get coupled in, = electromagnetically, electrostatically, and common-mode! Nelson's book = has interesting tales of the interference problems he analyzed and = solved! Whew! This started out short and got long-winded! Oh, well! Back to CA = form 540A! ------=_NextPart_000_0002_01C660A9.930B5210 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
    Well, let me contribute what = I can from=20 some of the things I've learned. An electrical circuit requires two = conductors;=20 there is no such thing as "ground" or "earth", although I've heard they = found=20 the absolute reference ground at 0:00:00  Lat, 0:00:00  = Long. If=20 you throw "ground" out of your vocabulary and instead substitute "signal = return", you'll have a much better time in wiring things, especially on=20 dielectric (plastic), as well as metallic airplanes. Every morning = when you=20 arise, repeat ten times "There is no such thing as ground in electrical=20 circuits". That being said, when a signal goes from a source to a = load,=20 there must be a way for that signal current to get back to the source. A = shielded wire, or a twisted pair, is a perfect medium to do that, = providing that=20 the return is connected to common at the = source end.=20 If it is not connected at the source end, then the current must find its = way=20 back by some other path, often bringing with it other undesirable = currents,=20 giving rise to common-mode coupling, so-called ground-loops. These = currents=20 arise from the difference in potential, AC and DC, between the = source=20 common and the load common. Obviously if the load is tied to its common, = the=20 return conductor must be tied at the receiving end also! This is where = opto=20 isolator circuits find good use in keeping out unwanted signals; they = only=20 require being tied to the opto circuit at each end without a common = return=20 connection between them.
    One of the best things you = can do to=20 minimize interference is to have a single point at which all power = returns are=20 connected. Keep the main and common buss close together and run twisted = pair or=20 shielded wires to each load. It also helps to have capacitors connected = between=20 the main and return buss. Parallel a 4700 uF 35V with a 1uF 35V film and = a .01=20 ceramic disc across them. This keeps both the main and common buss at = the same=20 AC potential.  Under no circumstance should loads share a return! = Please=20 don't "daisy-chain" returns from one circuit to another!  Down = that=20 path lies destruction, Alice! On my Lancair, I have two terminal boards = that are=20 mounted horizontal and parallel, one below the other, with a copper = return=20 buss below the bottom terminal board. The return buss is made from = flattened=20 copper tubing with brass screws through it at the same spacing as the = terminal=20 board terminals. The upper terminal board is for the main, aux, and = avionics buss. The one below it is for load distribution. I run twisted = pair=20 from the main and distribution terminals to each circuit breaker. = Current flows=20 from the main, to the circuit breaker, then back to the distribution = terminal.=20 That provides cancellation of the magnetic field in these = conductors. Then=20 each load has a twisted pair from the distribution terminal and the = common. Here=20 again, the currents go out and back over these wires causing the = magnetic fields=20 to cancel, and all circuits are tied together at one=20 common point. An advantage of this scheme is that each = load's=20 source wiring is easily found on the distribution board for = connecting,=20 dis-connecting, and trouble shooting. Another thing you might try is to = mount an=20 LED with each circuit breaker and wire it directly across the breaker=20 terminals with a series resistor, 1.2k 1/2W for 14V systems, and = 2.7k, 1/2W=20 for 28V systems. Then if a circuit breaker is open under a load, the LED = will be=20 on, showing that it's open.
    Radio Shack and JameCo have = snap-on=20 ferrite interference suppressors that can be placed over coax and wire = bundles=20 to cut down on interference between circuits. They are tubular ferrites, = about=20 3/4" OD, 1/4" ID, and 1" long, sliced in half down their long = dimension,=20 and secured in a hinged plastic cover that can be snapped in place = over=20 wiring. They cost less than $2 each.
    One more consideration is = where to put=20 the field circuit breaker if an alternator circuit breaker is used. If = your=20 battery is run down and you have to jumper it to start, and then shortly = after=20 you take off, the charging current on the depleted battery could be very = high,=20 tripping the alternator breaker. If your field breaker is connected to = the main=20 buss, the regulator will sense the low voltage on the buss and feed = maximum=20 current to the alternator field. With the alternator breaker tripped and = no=20 load, the alternator will put out well over 100V! Then when you push the = tripped-breaker back on, the sudden inrush current and high = voltage could=20 fry your electronics. A better place to connect your field breaker is on = the=20 alternator-side of the alternator. That way, if the alternator breaker = trips,=20 the alternator field will still be connected to the alternator output = and keep=20 its voltage under regulation! 
     For a much better = look at=20 the subject of shielding and grounding, try "Interference Handbook" by = W.R.=20 Nelson, WA6FQG and "Grounding and Shielding Techniques in = Instrumentation" by R.=20 Morrison. Morrison has good illustrations of how interference currents = get=20 coupled in, electromagnetically, electrostatically, and common-mode! = Nelson's=20 book has interesting tales of the interference problems he = analyzed=20 and solved!
 
Whew! This started out short and got = long-winded! =20 Oh, well! Back to CA form 540A!
------=_NextPart_000_0002_01C660A9.930B5210--