ET AL, An earth connection via capacitors cannot replace DC earth protection. As mentioned below by Ron Pickard capacitors are considered to be unearthed protection.
Long before UL or any other body that dealt with safety, it was a well known fact the only reliable protection against lightning strikes or contact with mains was a DC earth connection. I believe that is still the case today. As Ron Pickard has pointed out below capacitors do not provide DC protection. Capacitors used for safety are isolation capacitors or decoupling capacitors. Isolation is a form of safety, but it cannot be used to replace a DC safety connection to earth. As indicated in IEC 950 and UL 1950 capacitors used for isolation must be two capacitors in series. This is to insure there is still protection if one capacitor fails. At the Radio Science Lab at Stanford in the 1950/60's we performed a number of experiments in an effort to protect very expensive military radio transmitters and receivers. Capacitors produced limited protection at best. In the case of lightning or direct contact with mains the capacitor dialectric always punched through or were arced around. Lightning always finds the path of least resistance path to earth. If capacitors are used in this situation lightning will cause dialectric punched through. Creating a DC path that destroys many other components along with it. Also spraying molten metal over entire circuitry causing secondary shorting and possibly fires. Hopefully not killing anyone in the process. A DC earth connection during a lightning strike will sustain some damage, but the scenario more predictable. Duane Marcroft Telecom Consultant Peer Communications [email protected] > Jack, > > I suppose there are, in fact, no truly absolute unearthed circuits given >such > parameters as the earth's magnetic fields, numerous parasitic >capacitances and > atmospheric conditions. ;-) However, within certain defined specific > conditions (e.g., isolation voltage, spacings, insulation resistance, etc.), > isolation would then exist and then a circuit would then become unearthed if > that circuit is not physically connected to earth. > > Your suggestion for using capacitors might be valid, but only at F=0 (DC). > > IMO, a circuit is considered connected to earth via a capacitor because > capacitors have impedance of 1/(2*pi*F*C) [remember your basic circuits > classes]. And, when was the last truly DC circuit you observed? > > This is why DC hipot testing is recommended for circuits with capacitors >across > tested barriers (see note 2 of 5.3.2 of IEC60950). Leakage current >requirements > of 5.2 of IEC60950 might also need to be considered when using such >capacitors, > although SELV voltages would likely not present a safety concern. > > As for optos and relays, they offer AC and DC electrical >isolation/separation > from their primary to secondary circuits up to certain specified voltages >(most > transformers also apply here, too) and therefore, offer unearthed >connections. > > These are reasons why I feel that capacitors should not be considered as > offering unearthed connections. > > Well, so much for my temporal soapbox. And of course, comments are invited. > > Best regards, > Ron Pickard > [email protected] > > >______________________________ Reply Separator >_________________________________ >Subject: RE: protective earth >Author: <[email protected]> at INTERNET >Date: 7/14/98 2:05 PM > > >>From another angle: > >When is a circuit considered NOT connected to earth ground? In other >words, if the only connections to earth in a SELV circuit are through >isolating components such as 3kV capacitors (from point B to PE in the >drawing), is that circuit considered unearthed? I understand that a >circuit isolated from protective earth by optos and relays is considered >unearthed. > >Thanks, > >Jack >Netaccess, Inc.
