First off, I do think this C-Cor gear is total junk (mabee not in it's time, 70? ha,ha). It's not even the same ballgame as the C-Cor stuff I have worked with, and liked (flexnet750). But that's what we have and that's not going to change anytime soon. This gear is unreliable, the test ports are shot, the AGC is only a shade less then useless, repair/replacements are hard to find, gaskets leak (rf and water),ect. But I could deal with that if the area in question was broken up. The other half of the system that is a standard fiber set up runs great (with appropriate maint of course).
As to the setup for levels, That part of the system is in a rural area that is steep terrain everywhere and every house is very large (typically 4-8 outlets). So the design is for >20dB at the tap. Depending on drop length the data leg will usually have a DC-6 or 12.
The main problem is even if that area works good, which it does now, when something does cause a higher noise problem we have to start at square one with the process of elimination each time. That typically takes 2 hours (ideal weather, no traffic) just to find what neighborhood to look in due to the distances and slow roads involved.
I'm not sure what you mean about the freqs. To clairify, it's not the idea to get to a quieter part of the response, rather to seperate the paths from the nodes by putting each (BC&D) on it's own freq. so only the noise from each node (and the added noise from the coax hop) would be seen at each CMTS upchannel port. (I think that sounds right
)I guess unless I get some better advice / direction here, I am going to go experimental. I'm thinking about running one part of the return through a superhetrodine converter(s) and stepping it from 23 MHz to say 35 Mhz, 2 6meg channels and see if the CMTS will still recognise those modems. I'll do the experiment right at the CMTS with a node from the normal part of the system. I think Sientific Atlanta 6150/6250/6350 style input converters and output LO & mixer modules would have a flat response and be stable enough (although the filtering may not have sharp enough skirts).
If just converting freq. doesn't work then I may look into seeing if changeing the assigned upchannel freq. (for those in the converted area) to the expected freq @ the CMTS.
At this point that's all the further I was going to go with my experiment unless it was a success. Then I would figure out if there is a feesable way to implement it in the field, at least to divide the area once into 2 upchannel freqs.
And yes, I know this is a log shot, but what the.......
