Correction to my last email - instead of 20*log_10(N) and 10*log_10(N), it should have been 2*log_10(N) and log_10(N) On Tue, Feb 3, 2015 at 10:53 PM, Sean Breheny wrote: > I don't know what the entire frequency range of modern cable TV/cable > internet is, but not all splitters work the way you described. They fall > into two general categories: resistive (lossy) and low-loss. The resistiv= e > ones cause at least 6 dB per two-way split or in the general case, > 20*log_10 (N) where N is the order of the split (i.e., an "N-way" split). > Their advantage is that they are cheap and very broadband. > > The low-loss ones usually consist of an array of matching elements plus > load resistors - which only come into play when outputs are not properly > terminated. Even these do not reach 10*log_10 (N) "loss" (power division) > but can come close, but they are limited to two or three octaves of > frequency at most (e.g., really good ones achieve a 5:1 max:min frequency > ratio), whereas resistive ones will work all the way down to DC. > > > > On Mon, Feb 2, 2015 at 9:41 PM, Neil Cherry wrote: > >> With each split there is a 3db (halved) power drop >> >> /-----------------[ n - 3.5db - 3.5db ]- >> -[ n ]-+ /-[ n - 3.5db ]--------- >> \-[ n - 3.5db ]-+ >> \-[ n - 3.5db - 3.5db ]- >> >> > --=20 http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .