[Sephiroth]

Thanks for that. The downstream doesn't show any irregularity in the stats, as of course you knew.

The upstream displays the same problems as before - the effects of line noise (which you can see by the dropped packets.

Why has the modem flipped to low power? First, to explain that low power on the intended data stream gives noise (from any source) a good chance. We are seeing this.

When you had high power, it was commanded by the CMTS because it need the power to drive initial ranging data. The CMTS then decides on the power level it needs to overcome attenuation (anywheere in the works).

The CMTS is not going to command such low power because for a start, the CMTS expects to see a signal arriving at around no lower than the 20 dBmv mark (thereabouts, lower for QPSK - someone will correct me if necessary). There's more than 8 dB attenuation between your gaff and the CMTS.

So either the modem isn't seeing power raise commands from the CMTS at initial ranging, or it can't deliver the power because it's defective (and maybe made so by sitting for so long at the 60 dBmv earlier).

So I take back what I said earlier and your next move might be to get the modem swapped out.