Carlson is a gifted amateur; the kind who would probably have been a crackerjack TV repairman back in the hollow-state days. But his advice where modern stuff is concerned
can be just plain wrong.
His advice to follow where preinstalled cable-paks "tend to fall of their own accord" is a good one, which tactic I've used for decades when sleuthing out how a "borked unit in a box" is supposed to go together.
But his "fix" on the memory batteries made me cringe; not
just because he soldered directly to a Li-xx primary cell (which I've done too using the same "work hot & fast" technique), but his reasoning vs spot-welding a tab to the battery is just plain ignorant. The amount of heat delivered to the guts of the cell is still easily 3 orders of magnitude greater when soldering.
Now that I actually own a tab-welder I'd never do it again, and since nowadays they come as a cheap commodity PCB that can be had for as little as $12 delivered, I could not in good conscience ever recommend the soldering technique, just as a matter of personal safety for the person being advised.
That was cringe enough, but he didn't put any heat-shrink over the cell, just in case it did decide to puke its guts out, and to make a "finished" repair that doesn't have live battery terminals exposed with absolutely no strain-relief.
A lot of his "repairs" are similarly "half-finished".
And finally, he never even mentioned the
very obvious tactic (at least for those of us in here), and what should've been his
first step in determining the polarity of those memory battery plugs;
identify which pin is connected to the GND plane of the board.
While not absolutely true every time, at least with memory batteries it will be true most of the time... and if that coincides with your "small positive measured voltage" when probing the board, almost certainly correct. Of course, you can also usually easily follow said battery traces back to the IC they're backing up, and look up the datasheet for said IC. That is the way to be
absolutely certain.
His identification of the pads by shape is not what I'd consider good general advice, even if it is applicable to this model... the square pad
usually identifies pin 0 or pin 1 (depending on the numbering schema) which
often is the first "GND" plane pin, but
not always.
mnem
I do wish I could playback faster than 2X speed when watching one of his vids...