Hamilton 921- need some advice !

[Randy]

Not my first rodeo with this Hamilton caliber.
I bought it in non-running condition, but no explanation as to why.
The watch was not abused..perfect dial, case, original box.
Really nice example.


I saw ( and replaced ) a bent staff.
Also replaced the roller with another one I had, as it had a chipped impulse pin
No other chips or damage to any other jewels. Inspected with a 20 power loupe

Cleaned the movement in fresh Zenith solutions.
Hairspring cleaned in One-Dip
Oiled with brand new Moebius oils
9010 for the balance and escape wheel
941 on the pallet stones
9104 / HP1300 on the 2nd, third and fourth wheel
8200 on the new mainspring and barrel

Without the balance installed. the train worked perfectly, including the small rebound at the end.

So, I add the balance.
Pallet fork lines up dead center between the banking pins
Hairspring has no kinks, moves freely. Doesn't have a "riding" condition on the arms, no mis-shaped coils or magnetism. Coils look really concentric and aren't sticking.

Fires up after about the 10th wind, then picks up amplitude.

Stick it on the timing machine and it all goes to "south".

In dial down:
No rate registers..but the tracks show it's gaining a bit
Amplitude is around 250.
Beat error shows 5.1 MS ..( yet the fork rested perfectly centered at rest ) ?
Rate keeps showing 19800..not the 18000 it should be.. ???

Pendant down
Pretty much the same readings ..but a bit lower amplitude.

I know the machine is running well, as I put my personal 921 on it and it reads as it should;
Amplitude around 285
Beat error around 0.8
Rate at 18000

I've watched it with a 10P loupe while running. Can't see anything that looks off.

HELP !! LOL !!!!


Randy

[Randy]

Just realized I used the mainspring that came with the watch..not a new one. It was a white alloy, and hadn't taken a set so I cleaned it and put it back in. Now I'm thinking it may be a wrong strength, so I'll yank it out and put it a fresh one that I know is correct.

[Randy]

The mainspring appeared to be correct, but I put a new white alloy one in.

Same results..this has to be something in the balance..but I just can't figure it out !

[Bob Tascione]

Hi Randy,
I hate when that stuff happens.
You most likely have done everything I can think of here but just in case.
How's it doing in the other positions?
Could be the balance or staff pivot shape. Or maybe the roller that you replaced isn't correctly engaging the lever? Have you tried removing the lever and then vibrate and count the balance oscillations and observe hairspring movement with and between both index pins? That would help isolate the balance from the rest of the watch to test its action. If the count is correct and all else checks out okay then you could move to the roller/lever interaction. If all's well there then move on to the lever and then escape wheel.
Strange that your machine is showing 19,800 as that's a beat count (5.5 bps) used in some watches. Curious, if you put the hands on how much is it gaining as shown on the dial? I don't really know why I'm asking that question but I guess I'm just wondering if the machine is automatically determining the balance rate to it's closest programmed standard and then giving any variation from that number of 19,800 rather than 18,000. If someone popped a balance vibrated to 19800 in there in the past then...well I'm probably way off on that one. Again just pondering since I need something to do tonight.
Questions I ask myself when trouble shooting:
Any possibility that the escape wheel is distorted or escape arbor has a bent pivot? Timing machine could be sampling say a full escape wheel revolution and then spitting out an average beat error for that period.
Banking pins adjusted okay? How does the total lock look for for entry and exit pallet?
Pallet jewels and roller jewel and table secure?
Any excess pallet pivot play?
Can components be swapped from another good running watch to test? Like that other 921? lol Lazy way out I know but hey, works for me.
I'm sure there's a lot more that can be checked but the above usually reveals something that can be chased deeper.
Curious what you finally uncover Randy!
Good luck for now,
Bob

[Randy]

Hi Bob..thanks for all the great ideas.

The lock/slide looks good. The watch starts up after a few winds, and at rest, the lever is dead center between the banking pins. The jewels appear to be tight and clean. No shellac goop that could interfere with the action

The banking pins are the non adjustable type, and they appear to be straight, no damage or bends

The hairspring is oscillating fine between the regulating pins, and the pins appear to be smooth so no catching or dragging going on that I can tell.

I did have to close the hole on the roller, as it wasn't' fitting tight. After staking it down, it rests even with the balance arm, and when I remove the balance ( about 20 times now LOL ) it is holding it's position, and appears to be snug. Watching the jewel moving with the escape lever shows it to move easily, and the guard pin isn't dragging or interfering with the action.

I did put hands on it to verify what the machine was getting, and it was gaining almost 15 minute an hour !

I also have a couple of 917's that I could use for parts if I find damage to the escape wheel. Neither of them had good balances so I will find time today or tomorrow to swap out the balance from my wristwatch conversion 921, to see what effect that has. I'm hoping that will isolate something.

Thanks again..I'll keep you in the loop.

Randy

[Bob Tascione]

Hi Randy,
I remember your 921 wrist watch conversion now. That is a beautiful watch.
A few things come to mind after having this morning cup of coffee.
First it does appear that your timing machine is automatically referencing a programmed base frequency of 19,800 since the watch is running 15 minute an hour fast. The reason I asked how it is doing in the other positions was to determine the poising and the possibility that one or more balance rim screws may have been altered, replaced or removed in the past or that the balance was vibrated at the higher rate.
If the balance has been altered to be made to beat around 19,800 then that extra .5 bps equates to an additional 1,800 bph. That additional balance velocity (and the possibility of a bit too small roller jewel) would put the impulse a little further past the line of centers than the design called for as the lever has to catch up with the jewel. Might result in a late, partial impulse resulting in less balance amplitude.

The reason I mentioned the possibility of a distorted or maybe off plane escape wheel or bent escape arbor pivot is that, although the roller jewel is lining up on the center line, the timing machine could be sampling a number of beats before giving an average that doesn't necessarily reflect the alignment of the lever along the center line. I suppose if it sampled the beats of all 15 teeth of the escape wheel then the average might be correct but any sampling of a number less or a bit more than 15 teeth might give a larger cumulative error? I need a couple more cups of coffee to think that one through. Probably wrong but fun to think about.

Looking forward to seeing what you come up with.
Bob

[Randy]

Thanks Bob...I still like it ( the watch ).

Good stuff to ponder for sure. I'm hoping to get at the bench later this afternoon to smoke out the possibilities you've mentioned.
Thanks again for helping me to see this through.

Be well Amigo !

R

[Randy]

OK..isolated a few things
I used my balance in it.
Runs perfect..-1 sec per day..amplitude is 330, beat error improved to 0.5 !!

So..I started rechecking everything. I replaced to roller I had with another one from a 917 parts movement.

It all went together well.
But,...
Face up- 259, 1.5 error rate ( big improvement so the roller had something going on ) still reading 198000
Pendant left- 245,..2.3 error rate...19800
Pendant right- 261,..1.5 error rate..19800
Pendant down- 254,...1.5 error rate..19800

I counted the coils...same on both balances-

All the balance screws are evenly set around the rim. None appear to be missing, or altered fro timing
Also,..the numbers scribed on the balance arms don't match the serial number on the mainplate

I'm now of the opinion that someone crammed a 19800 balance into this..just to sell it.

Any other ideas,..thoughts that I may have missed Bob ??

Randy

[Bob Tascione]

Hey sorry Randy,
I didn't realize that you had followed up on your earlier post from yesterday.
Great, thanks for checking that.
I agree. Most likely not the original balance.
The timing data shows very little amplitude variation between dial and pendant positions which leads me to believe there still may be an issue with the pivot end shapes and or the increase in the impulse jewels angular velocity resulting in a delayed partial impulse. Is it possible that when the balance was replaced the staff length had to be reduced slightly to fit by grinding flat (or worse, ground at some angle off the cap jewel plane) pivot ends? Removal of the conical pivots will decrease the dial positions amplitude.

If the pivots are good can you turn the timing screws out a bit to see how that affects the timing and amplitude? If the screws are turned all the way in then maybe the person who sold the movement just put a balance in that fit and managed to time it at 19,800 on a machine never checking the dial reading afterwards.
Any extra hairspring tail sticking out past the stud pinning point?

Gracias Randy!
Bob

[Randy]

No worries Bob....We all get busy !

I replaced the staff with a fresh one from a sealed Hamilton package ( 607100 ).
I inspected the pivots after the new install, and they appear to be well shaped and without scarring or blemish.
I didn't need to alter or shorten them in any way. After install the balance dropped in with acceptable end-shake.

Your timing screw comment is something I need to play with. That could be the issue.

I have another way out of this I think within the week...I'm chasing down a new balance complete.

I'll let you know how it goes !

Gracias Compadre !

Randy

[Bob Tascione]

Ooops, I didn't realize that you had changed the staff. I just re-read your first post and see that you had mentioned the replacement.
I look forward to hearing what the problem turns out to be.
Que tengas buena suerte Randy!

Bob

[Randy]

Last night I swapped out the hairspring and roller onto a 917 balance arm I had in my spares drawer. Now it reads at 18000, and the amplitude is around 275. But the error rate won't register..appears to be racing off the screen still.
I can now only think that the original hairspring and balance were vibrated for 19800, so I'm just going to wait for the complete balance I think I've spotted. I'll let you know how that turns out !

[Randy]

Finally received the balance complete, from a 917 movement.
Cleaned it, and dropped it in,..and she took right off with decent amplitude and a rate I could adjust !

Thanks to all (and especially Bob for all the great ideas ) for the support !

Randy