here goes....catv 101
forward signal: the signal that leaves the headend which eventually provides service at a subscriber's residence.
satellite signals are received at head end. these signals are decoded and recoded or modulated (depending on digital or analog). each channel is combined with the other system channels and then sent to different forward lasers to go to their respective nodes. all of these channels are "broadcast" because they feed the same signal to all nodes. any signal that delivers different signals to different nodes at the same freq are called "narrow cast". node specific internet, voip, and vod are all examples of "narrow cast".
forward lasers convert rf signal to light. this light is transmitted down fiber to the node amp, where it is converted from light back to rf. the rf is then launched to different amps (or actives) which cascade the signal (by amplifying the signal) downstream (or towards the house). everything the rf signal goes through has an "insertion loss", the decrease in signal strength that occurs when pushing rf through a medium. taps are used to distribute the signal to the house. most taps are power blocking to the subscriber but some are intentionally able to pass ac.
as a general rule but by no means set in concrete, systems are designed to have about 14 to 16 db on the high/low end of the forward spectrum on each port of the system taps. this number is usually arrived at by planning for a 100 ft rg6 drop (or rg11 over 200 ft), with 4 planned outlets inside the house (yeah right). under this design, an "average" house should have no problems setting up 4 active outlets while maintaining fcc minimum signal level requirements.
entire systems are designed with little more than a map, desired level specs and knowledge of insertion loss. a good tech uses this knowledge to calculate expected levels at any location (either hard line or in the house, rf has a predictable level loss)
return signal: an rf frequency used by the subscriber equipment to communicate with the head-end.
the modem or digital box with modem capabilities becomes the beginning or "head end" of the signal. the signal is generated in the house and travels to the tap. once in the tap, it goes to the its upstream active. the active amplifies the signal to a desired level and sends it to the next active or the node amp. every return signal (and all noise/ingress) is funneled back to the node amp. once at the node amp, this rf signal is converted to light (it can be muxed at this time in a more advanced system) and transported down the fiber to the headend, where it hits a return receiver.
this receiver converts the light back to rf. the signal is then split to several locations. the cmts (for voip and internet), the rpd (for vod, etc), the sweep receiver, the return ingress monitor (pathtraks), and for us: the arcom hunter.
the rf signal is bound by all the same insertion loss rules as the forward (although the loss is lower at these freq).
leakage: rf signal leaking from the "closed cable system" into the open air.
most systems use a meter with an antenna in the tech's vehicle. this meter and antenna can be "cut" to receive certain freq and "ignore" other freq. the tech then drives out a route (or just does his work for the day) and documents any leakage he detects with his meter. this documentation (depending on level of leak) is tranferred to someone qualified to squash the leak or if the level is high enough, the leak must be fixed immediately.
ingress: rf signal penetrating the "closed cable system" from the open air
this is what you were asking about. unwanted rf penetration can do many things depending on its characteristics.
a simple example of rf interference (although its not actually ingress) is a digital box that constantly sends out an upstream request. lets say you request vod from the digital box on 11mhz. now, in node X, subscriber Y has a box that continually sends requests to the rpd (something wrong in the box so it doesnt time out properly). if this signal is constant, it will not let any other box in the node get through to the rpd, so in effect vod is killed in node X.
now lets take this further. node X is combined with nodes S,T,V at the back of the rpd. now you have vod killed on 4 nodes. now you have 4000 customers (for example) who can not order vod. the bean counters are pissed. =)
ok, now lets talk about rf ingress that can trouble voip and hsi. noise, impulse, cpd are all different types of ingress. each one is caused by different things, but the basic pattern is: loose connectors, broken shield or braid, loose seizure screws and corrosion build up (this is not all inclusive but good enough for this talk).
ingress (as a general term) can cause trouble in 2 basic forms. first, the level of noise becomes high enough on a certain frequency that it actually drowns out the signal that lives on that freq (kinda like ghosting on the forward). the other is "overdriving" or "clipping" the return laser. when the overall "power" of the signal becomes too great, the return laser will "clip" destroying the signal quality and shutting down the node return.
voip and hsi are generally combined behind the cmts. so, as a general rule, if you bring down 1 node, you bring down everything its combined with. this combining is usually connected to a "blade" or card on the cmts (usually 8 ports) and this entire blade can be brought down in extreme circumstances.
thats it for cable 101. there is a crap load of science and theory behind this technology so my description is as basic as i can make it. needless to say there is many details left out and lots of gaps to fill.
good luck
Re: How the System Works?
There are 5 replies to this message
