CABL Forum - CPD Possibilities

I have been reading many postings on CPD related problems. Some problems may be a similar looking but not CPD.

Yes, most CPD is caused by the dissimilar metal corrosion at junction points in the CATV network, giving the old "mixer diode" effect that shows up as a spike every 6-MHz in a standard NTSC system and every 8-MHz in a PAL system within the return spectrum. This problem can be anywhere from the node to the "f" connector on the back of the TV.

I noticed many years ago that a similar type CPD condition will exist in a system who runs excessivly high forward outputs. All amplifiers have manufacturer specification outputs that reflect channel loading and distortions related as such. Any changes from this spec sheet will change distortions accordingly. What is normally NOT mentioned is the specifications on the DIPLEX filters.

In my experiences, depending on the equipment, forward output levels in excess of +50dBmV (usually in the bridger/LE) may "possibly" OVERLOAD the diplex filter and cause a condition similar to CPD by having the forward channels "bleed over" into the return spectrum. This is not CPD, but is similar when viewing with a spectrum analyzer. The isolation of the diplex filter can only do so much.

A simple reduction in forward levels corrected this problem. ALSO, this is ANOTHER reason for a FORWARD SWEEP of the CATV plant. Espically if it is the older "Trunk and Feeder" design with or without a node for cascade reduction. Anyone who sweeps knows about the infamous "HUMP". This "HUMP" can be excessive and actually not only add more CSO and CTB, but, this increase in level can also contribute to the "diplex overload" mentioned above.

EXAMPLE: You balance your entire system at the bandedges. 750/55-MHz at 48/38-dBmV respectively. You ASSUME that there is a FLAT LINE TILT between these two points. Well, think again...that is the PAPER WORLD. In reality, there is a "HUMP" or "HUMPS" that can be in excess of 1-10dB or more depending on the cascade, design, type passive, age and various other varibles. The assumption is that the RF never exceeds +48dBmV at anypoint in the plant. If the "HUMP" is around 450-540 MHz, and is a 7dB hump, that portion of the spectrum is actually about +53dBmV or more. And that is if your AGC's are working properly...but that is another story. This "HUMP" alone is enough to possibly overload the diplex filter and "simulate" the CPD issue mentioned above if levels are not controled.

Just though that possibly somebody might find this information helpful. I know what it is like to beat my head and get no results.

Greg

INSTALLATION

CONSTRUCTION / TECH

OFFICE / OTHER

CPD Possibilities