Problems with CM1150V upstream on Comcast Xfinity

In a previous post, two years ago, I presented an analysis of problems with the CM1150V. (https://community.netgear.com/t5/Cable-Modems-Routers/Problems-with-CM1150V-upstream-on-Comcast-Xfinity/m-p/2030739/highlight/true#M29475)

I just returned to that post and saw the reply by decosse pointing out a fundamental error in what I said.  That topic is closed, so I can't reply there, hence this.

Yes, I realized the error sometime after making that post, much to my utterly cringeworthy and mortifying embarrassment.  It's an object lesson in dyslexia combined with falling prey to the seduction of beautiful but misguided theories.  I'll never live that one down, at least in my own mind.  Not to excuse my own sloppiness, but I'd argue that Netgear is being more than a little silly in displaying the upstream frequency in Hz, not MHz.  I guarantee those last few digits are completely meaningless.

Nevertheless, I'm on my third CM1150V and still having problems with the modem. When the second modem I received exhibited the same problems as the first, I spent hours dealing with Netgear customer support and eventually sent them a long video demonstrating that there was a very sharp temperature threshold below which the upstream amplifier would fail.  (This was about 62 degrees F at the surface of the metal shield can over the front end components.)  This was sufficient to get them to send me the third modem, even though this was long past the warranty and support period.  Of course, the third modem had the same problem.  At this point, having other things to do in my life, I just wrapped the modem up in a blanket of bubble-wrap to keep it warm via its own self-heating and moved on.

It's now December and even though I'm in Northern California, it's cold.  The room that the modem is in regularly gets down to the low 50's F, and the modem is failing again.  I've ordered some replacement NXP BGA3131 upstream amplifier chips, on the chance that there's a bad batch that Netgear got into.  Incidentally, that same chip is used in the XB6-T (aka Arris TB3482G) modem that I rented from Comcast for a while during all of this.  That modem performed fine, and I think I cooled the amplifier explicitly to test this, though that's a bit lost in the mists of time so I'd have to rent another one and repeat the experiment to be absolutely sure. If that was a valid test, it suggests that either Netgear got some bad (counterfeit?) chips, or their design is somehow abusing the chip.  I'll post something here if I do indeed get around to replacing the chip.

BTW, decosse, if you're listening, I apologize for not seeing and responding to your post of 2020-12-15 12:46 PM.  To answer your questions if you still care, yes those upstream power numbers look fine, and yes, the most obvious symptom is poor download speeds.  There's no easy way to see if your upstream packets are making it back to the head end, but if you're requesting downstream packets and your requests never make it upstream, then of course you won't get your data.  This is one of the things that made it so hard to understand early on.  The other notable symptom will be a large number of T3 Timeout errors in the log file.  However there seems to be a bug in their logging code in that most of the time I see it simply overwrite the last entry rather than accumulating errors, making the whole thing even more confusing.