i think i understand.
i'm curious if your cmts is able to tell all customers from a certain node to respond on a specific upstream freq. we tie each downstream freq to a certain upstream freq and then we run 4 downstreams per node.
that would put you in a position to test the concept 1 node at a time and also answer the question of whether your cmts can read the upstream freq (since the upstream data is dictated by the info that the downstream docsis channel gives it). this way you dont have to do any conversion and you know right away if the cmts is going to play friendly.
i am concerned though, that when you recombine your nodes, the noise will be power additive and your noise will be right where it was when you started, despite the freq allocation. if i'm misunderstanding your intent, let me know. if you plan on using some sort of "cut" mit filter that only passes the particular upstream (and killing all other noise on that leg) before hitting your combining point, this may help.
all things being considered, you are still down to 2 choices, in my opinion. pay for the studies and get your fiber there or work the noise down until it takes a MAJOR problem to kill your upstream.
if your upstream snr is below 32 at any point, then you have to kill more noise. 1000 subs with 200 data carriers is not over the top in my opinion. we run our system based on 250 data customers per node with about 40% penetration (so about 700 subs i guess), and we are considered conservative.
good luck with whatever you try
Re: need upstream data help
There are 0 replies to this message
