Hi Joe, Thanks very much for the very detailed response to my question.
It indeed looks as though digital telephones are SELV devices. However, I wonder what the legal basis would be. Surely, it would be insufficient to just measure the voltages on the cable to which each telephone set model is connected to its PBX extension because, theoretically, the telephone might internally generate higher voltages; or the measured voltages might not represent the highest possible under certain conditions. Regards, David Drori Novarex Enterprises Ltd., PO Box 45178, Room 204, 2nd Floor, Park Center, Hamarpe 1, Har Hotsvim Industrial Area, Jerusalem 91450, Israel. Tel: +972 2 540 0168 Fax: +972 2 540 0169 GSM Cellular: +972 54 828011 E-mail: [email protected] -----Original Message----- From: [email protected] [mailto:[email protected]]On Behalf Of [email protected] Sent: Tuesday, September 11, 2001 4:59 PM To: [email protected]; [email protected] Subject: Re: Safety question regarding feature phone adjunct device In a message dated 9/11/01, David Drori writes: Can anyone specify (quoting the relevant safety standard paragraph) the degree of safety isolation required between the portions of the device that are connected to the externally connected PC (that are referenced to PC ground), and the portions of the device that are connected to the four terminals of the handset socket on the feature phones? Hi David: I had to wrestle with this problem several years ago when I designed the interface circuits for an encryption device that was inserted in series with the handset cord. This type of series device introduces some interesting safety issues and some very challenging electrical interface issues, since handset electrical interfaces differ substantially from vendor to vendor. Following are some brief comments on the safety issues: 1) If the series device will be used with conventional analog POTS phones that are connected to the PSTN, the handset cord will be TNV-3. This is because there is no safety isolation between the PSTN cord and the handset cord of a conventional, loop powered telephone. 2) If the series device will be used on a conventional analog POTS phone that is connected to an analog extension port of a digital PBX, the phone line will be TNV-2 and the handset cord should be treated as TNV-2, since there is no safety isolation between the two cords. 3) If the series device will be used on a digital extension phone on a digital PBX, the incoming phone line will most likely be SELV, and the handset can most likely be treated as SELV also. As you can see, it is to your advantage to limit the specified applications of the series device to specific digital telephones and digital PBXs, so that safety isolation is not required. In my case, the device had to work with conventional analog POTS phones connected to the PSTN, so I had to provide full isolation from the handset cord. For analog POTS phones connected to the PSTN, using the new 3rd edition of EN 60950, you will need basic insulation between the handset cord and the USB port (I use 1.6 mm creepage and 1.0 mm clearance). The hipot test in clause 6.1.2.1 is now 1500 VRMS for TNV-1 and TNV-3. For Norway and Sweden, Annex ZB places some requirements on distance through solid insulation. These requirements are slightly different from the ones that appeared in the 2nd edition of EN 60950. In that version the hipot test for isolation from ground was 1000 VRMS, but the national deviations for Norway and Sweden called out supplementary insulation, which imposed a 1500 VRMS hipot and larger creepage/clearance distances. You may find that you are slightly better off with the 3rd edition, especially with regard to the mechanical relay you refer to. Lastly, your best bet for isolation from the AC mains is to use a Class II wall transformer with reinforced insulation. Depending on your internal architecture, there may be an issue with isolation between the handset cord and the (accessible) connector where the low voltage output from the wall transformer connects to the series device. If this becomes a problem, one solution is to make the power connector not accessible, either by careful mechanical design or by hard wiring the connection. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
