said by djrobx:
The signal level immediately from the node is always high so the node will be a able to feed everyone to the end of the trunk. So to achieve the 'target' level for the first customer in the line (people closer to the node) a high level attenuation tap should be used but this attenuation is not unidirectional (upstream and downstream) it affects both signals the same way so the modem to compensate for the added attenuation it has to increase its power output (upstream).
I had to think about this one. At first I thought attenuation being equal it wouldn't matter. Either you have attenuation from cable length, or you have artificial attenuation added from a high value tap.
Then I realized that cable-length attenuation impacts higher frequencies (where the downstreams are) more than low frequencies (where the upstreams are).
»www.w4rp.com/ref/coax.htmlThat means if attenuation is added to each tap that "targets", say, 0db at 500mhz, the upstream will skew stronger (lower modem output levels needed) as you go further away from the amp/node.
I had to think about this one too
But there is another variable that we havent taken into consideration and that is the variable gain in the receiver (downstream) This number can take any value (gain) between -15 to +15 dBmV. I don't work in the cable industry so I don't know the details but from my experience I would extrapolate this sub-system as some sort of AGC (automatic gain control) ... well sort of.
The optimum value for the 'operating level range' as some motorola documents call it is 0dBmV, this means (to me) that the modem can compensate better in case the signal is too strong or too weak. The 'operating level range' can fine tune the receiver's gain in case the tap value is not an exact match. I'm sure taps come in standard values »www.taiwantrade.com.tw/tcc/produ···S/198003
Tap values available:11dB to 35dB in 3dB steps
So if a cable guy needed a 12.5db tap he would have to choose a 11db or a 14db tap. Unmatched taps could influence upstream power output too.
Another thing, the cable types used in comercial and HAM radio like in Richard's site (the link that you posted) are not the same used by CATV. I don't know the specifics or the type of cable but from what I've seen it looks like a ~1 inch in diameter hardline type of cable with a thick ~1/8 in. aluminum jacket. »www.therfc.com/hardline.htm
I'm sure that this type of cable has a much better RF characteristics and lower attenuation than those used in HAM radio.