Hi Dwight, I had the same thoughts and I did some research on this topic. When there is no ground connection in the interface, you are relying on the other points of isolation within the system to provide longitudinal protection. The over-voltage surge is going to appear across tip/ring and ground. An example: With no ground connection in the interface, the over-voltage surge is trying to go through the telco transformer (about 1.5kV isolation), through the interface IC Tx/Rx pins, through the IC's Vcc pins, through the DC supply and through the mains power supply to ground. You are relying on the isolation of the power supply (about 3kV), the telco transformer (about 1.5kV) and the inherent isolation of other components within the system. These do not necessarily add depending on the layout of the complete system.
Some of the higher over-voltage surges such as with NEBS testing may exceed the isolation provided within the system mentioned above and you could get damaging voltage levels across sensitive parts. Metallic over-voltage surges appear at the interface due to the fact that the buildings primary protectors, one in each line, are not perfectly matched. If there is no functional ground connection in the interface, you should be able to get by with metallic protection only. Especially if you are not testing against NEBS, outside wiring-type requirements. Ground protection in such an interface will give you some added protection against the higher test levels found in Bellcore (or Telcordia) documents. I hope this helps, Jack ______________________________ Jack Murphy Sr. Compliance Engineer Brooktrout Technology 18 Keewaydin Drive, Salem, NH 03079, USA tel 1 603 890-7284, fax 1 603 898-1199 jmur...@brooktrout.com > Question regarding "two-legged" vs. "three-legged" overvoltage > protection circuitry: What are the pro & cons of the two? > > Background: Typically, to protect against overvoltages on a telco > interface protection circuit (analog or digital, such as POTS, T1, HDSL, > etc.), one sees either: > > 1) an MOV/varistor type device across TIP/RING, or > 2) two MOV/varistor/gas-tube type devices tied in series across > TIP/RING, with the center connection tied to earth ground. > > Of course, there are also typically PTC's or fuses in line for > overcurrent protection. However, my interest is the pros/cons of the > overvoltage protection topology. > > If the interface circuit has no path to earth (typically through > overvoltage protectors), then UL1950/UL1459 allows waiving of the > longitudinal(common) mode overvoltage tests, which makes sense, because > there is no return path for the fault energy. > > Since this waiver eliminates about half of the overvoltage testing, why > does one see the "three-legged" topology being used? Are there some > advantages to shunting energy to earth, rather than just back out the > TIP/RING pair? Certainly, one has to provide overcurrent protection to > prevent building telco wiring from burning (tested via the MDQ 1-6/10A > fuse), but are there other reasons for preferring a three-legged > approach? What are you missing out on if you elect to use the simpler > topology of just an MOV across TIP/RING? > > (To further stir things up, how about if we take into consideration > Bellcore GR-1089-CORE? Does that change things? I don't believe GR-1089 > specifically contains the same waiver as UL1950/UL1459, but certainly > the results are the same, and a test lab should consider waiving for the > same rationale.) > > I'm sure many of you have seen both topologies described in application > notes for various interface components, and have had to deal with both. > Any light shed will be appreciated by all. > > D > -- > > DWIGHT HUNNICUTT > Sr. Compliance Engineer > > **************************** > * <dwi...@vina-tech.com> * > * (510) 771-3349 direct * > * (510) 492-0808 fax * > * VINA Technologies,Inc. * > ****************************
