Background: I run FMJ spindles, 78 Cordoba discs and regular set 2/set 17 bearings on my 67 Coronet. I have maybe 10k miles on the current bearings, tops.

I hadn't checked the bearings for a bit and found them to be a bit loose - they clearly failed the vertical push-pull test with the wheel on.

The outer bearing and race looked OK when I pulled the disc off - no chatter marks or scoring. Nothing weird came out of the old grease when I repacked it. Wheel turns quiet save for noise of pads on the disc.

When I went to set the bearing by the book with the wheel on - 90 in. lb. while spinning the wheel, then back the nut off one slot on the adjuster - I could still get an audible knock out of the wheel if I tried to rock it.

I have seen other factory manuals, like same-year Chevy manuals, call for 12 ft lb/144 in. lb.

If I squinch things up this much and back off 1 slot, it is a bit better at first but I still end up with audible play. It looks like about .003 up-down on the nose of the hub (though I think my current dial is a bit busted) and I can still get a sound out of it.

Is it normal to be able to get a sound out of this adjustment, or am I doing something wrong/these bearings are hosed already? I know they need a little room to heat up, but it's honestly been years since I monkeyed with this & I'm not entirely certain if I noted this when they were set as-new.

I would consider that amount acceptable where you measured it. I'd rather have 'some' play then have them 'bind' at some higher operating temp. Disc hubs do run 'hot', and can be even hotter depending upon your driving manners. We're talking large bearings here, not something for computer drives or cnc machines.

Are they quality bearings? I had a issue some years ago with some China made bearings on my sons mustang.

Torque 50ft-lbs while spinning to fully seat the races and bearings, loosen, then torque to 90in-lbs. You should be able to line up the keeper without turning the nut.

If the spindle is worn from a possible seized bearing it may now be under size. The bearing preload can be correct but the bearing can be moving on the spindle.

are you using castellated nuts or nuts with the covers. The covers allow one to align the cotter key without disturbing the setting.
they look like the image

Torque wheel bearings?

This is wrong

From the FSM

Timken tapered roller bearings are design to be ran with preload on them, how much depends on the application. those on the 3rd member carriers in rear ends get a lot more preload than the front wheel bearings should have
That being said there is no way I would drive any of my early model Mopars with any front wheel bearing hubs having any up or down free play after adjusting the bearing preload, SOMETHING is WRONG with yours, I think
When i was a lot younger, back in the early 1960s, I worked in several different service stations, now know as gas stations, and did a lot of lube jobs which would include removing the front brake drums and hubs to remove and inspect and then clean and repack the good wheel bearings, replace any than had any obvious visual issues on the bearings or races
GM cars back then, especially Chevy, had ball bearing in the front hubs, they were way more sensitive to the proper front wheel bearing preload adjustments

Here's how Timken says to do it

I use a torque wrench on everything.

Except wheel bearings.

Cinch it down tight with pliers to squeeze the excess grease out, and seat the bearings, while spinning it. then loosen, Do that a few times.

The last time, just snug enough to take the play out, no more. Zero or tiny bit of preload. Done.

I agree. I have never had a bearing failure and have never torqued one.
A buddy of mine burned through two sets before telling me the procedure he was using. He mistook ft/lbs for inch/lbs.

Yep! Done it like that for over 40 yrs. Never had a bearing failure.

That's how my dad did it. So how I do it. He worked the flight line at an air base and kept unclecsams stuff in the air.
So he cared about lots of things being done right and did not take short cuts that were not safe

I guess some preload is OK, as long as heat expansion of balls/races still allows everything to spin "freely" when heated up.
But how would you know that except checking freeplay while it's still hot?
And disc hubs do get really hot sometimes (pad friction), more so then drums or the rear axle bearings.
That's why they say backoff a notch or such.
I'd rather have a bearing with "some" freeplay while hot, then one that's "binding".
Excessive preload, even on a quality bearing, will produce scoring, mainly on the race surfaces.

Me too

clearances loosen up with heat expansion

Timken recommends .001-.005 end play for tapered roller bearings. I got in the habit of using a dial indicator, shooting for the lower end of that spec because we were required to record the readings on work orders. I don’t pull wrenches for a living anymore, but old habits never die…

just a side note on timkin bearings.
i bought some for a bridgeport mill a few months back, and they were made in elcheapistan. they were not bashful about the price, however.

I used Timken bearings originally.

I had to get the old girl on the ground to make space for a daily driver repair, but I ought to just replace the bearings once I am done with that since they are cheap either way. The donor spindles do not look like they ever had anything get scorched on them, and I want to say the play feels and sounds like it is coming from the inner bearing, not the outer one - can't exactly just take that thing out and peer at it the same as the outer, so may as well just play it safe and redo them.

Also, I do not have in-out play, it is up-down play if I put my dial *under* the hub and rock the wheel.

When I first bought the car, the original drum knuckles and bearings were clearly audibly shot, so I guess I caught this problem before much worse wear happened in there.

If I get the same results with brand new bearings I will be real confused.