Dave, On Thu, 18 Sep 2003 06:35:58 -0700, Dave Tweed wrote: > Howard Winter wrote: >...< > > The system you describe is OK for normal domestic use, but if you ha= ve a > > radio transmitter (I'm a radio amateur) with an aeriel system which = has > > an actual Earth connection, transmitting may trip the GFCI. > > You probably don't want an RCCD in the circuit to the ham shack. Or ar= e > you saying that there's just one RCCD for the entire house? That's > different from here in the U.S., where they're used one per branch cir= cuit. There are various ways to do it here, as a user you can use socket-adapt= ors or mains plugs with an RCCD built in (cost about =A310, say US$15), or there could be one installed in the= Consumer Unit (the distribution box "under the stairs"). In the latter case the unit costs typically =A350 = (US$75) and would either cover all circuits in the house, or it would be "split" so that it covers only som= e of the circuits, usually leaving out the lighting so that a nuisence trip doesn't plunge the house into darkn= ess. Since the Miniature Circuit Breakers (MCBs) protecting each circuit are about =A35 each, and RCCDs o= nly come in high-current ratings (63A or 100A, usually) there's no way you'd have one RCCD installed per circuit.= Basically it's designed that the RCCD replaces the "Incomer" switch or main CB. > The only places the NEC actually requires them is on branch circuits t= hat > go near where water is regularly used, such as bathrooms, kitchens and= > pools/spas. Sockets aren't allowed in bathrooms here (except special "shaver sockets= " that have an isolating transformer built in) and I don't know anyone with a pool or spa to comment there :-= ) Electric and "Power" showers would have an RCCD if there isn't a "whole house" one. Kitchens, for some rea= son, don't seem to be any different from the rest of the house, but any metalwork such as the sink and pipes= would be earth-bonded. > On the other hand, you could take this as a clue that it's time to ver= ify > where those unbalanced RF currents are actually going -- that could be= a > dangerous situation as well! Well they may be going to Earth as part of the transmitter operation - b= ut I don't know enough to say whether this would be a fault or just the way it works! > > Personally I'd rather have a nice big Earth cable connected to every= thing > > metal, because relying on an active device to protect you isn't "fai= l > > safe", and I have an RCCD socket adaptor which has failed "dangerous= " - > > it doesn't trip when tested! > > Well, that's the whole point of the test! Yes, but my point is that if an active safety device can fail "dangerous= ", Murphy will get involved such that it fails between the latest test and the fault occurring! > Without a GFCI, you're going to carry >15A for hundreds of ms until th= e > overcurrent protection on the circuit trips. With a GFCI, you've lost > nothing (except a few bucks) and gained a large amount of additional > protection. Yes, but that fault current will be to Earth rather than through a perso= n. They are really dealing with different problems, in different ways. Having all metalwork earthed mea= ns that a fault to chassis (or pipe or whatever) will cause an overcurrent trip even if no people are involved,= and it means that any metalwork can't remain live, waiting to be touched by an unsuspecting human. An RCCD/GF= CI will keep a shock to a human within "safe" limits when it happens, but isolated metalwork which becomes acci= dentally live may not trip it until someone touches it. Having both types of protection is a Good Thing, ob= viously, but I don't like the idea of *only* having an active device, and I'm pretty sure the regulations here= wouldn't allow it. RCCDs are sufficiently "new" here that most houses won't have them, only new ones = or those that have been rewired in the last ten years or so. Cheers, Howard -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads