Viewing our return path on either an HP analyzer or with Pathtrak these beats(?) usually appear at a lower amplitude on the high and low sides of the data band, though sometimes (rarely) they can appear to be random impulse noise scattered across the entire spectrum.
The more typical “beats” viewed with max hold will build up “shoulders” on the sides of the data band or will form “horns” or “wings” to the left and right of the data band. (See the Pathtrak pictures I’ve included. One in this post and one in a reply.)
Several of us involved on this project have been on multiple mitigation projects with active return paths and have all been heavily involved with mitigation for VoIP this year. None of us have ever seen this anywhere else, not that that means anything!
The plant here doesn’t seem to be the issue. We have seen these beats on old unjacketed 750/500 plant with GI Starline gear, ADC nodes and ADC headend receivers. This lead to speculation by the operator that it was a reflection issue caused by old splices, seizure screws or any other plant flaw you can think of.
Now we are working on “new” QR 715 plant with SA III gear, SA III nodes and SA Prisma headend receivers and are getting the same beats. Both types of plant are fed from multiple hub sites so there seems to be no common head end issue. Could it be something common “further upstream” where all the hub returns are combined?
In both types of plant these beats only occur in some nodes perhaps just 5% to 10% of all of the nodes. So the question is not only what is causing the beats but why do they occur only in an occasional node and not all of them.
Even though they appear to be beats because of their amplitude being less that the primary carrier another theory is that they are not beats at all but instead are modems transmitting out of their allocated bandwidth.
Are there any CMTS experts that could answer whether that is possible? Could it be a simple matter of a modem not being told to limit its transmission to the proper frequency range? See the “wrongFrequency” shot. This seems clearly to not be a beat but a case of a modem(s) talking in the wrong place. The amplitude is only half of normal like the other “beats”. This happened immediately after a return pad was replaced and the modem(s) reestablished its upstream connection.
Changing out old RG house wiring has solved the problem on occasion but that doesn’t explain why this seems to be the only system in the country to have these issues.
Is it? Has any one else wrestled with this issue?
