Clock won't regulate

[Elliott Wolin]

I have repaired/cleaned/refurbished a number of clocks, but one of them continues to give trouble: it won't regulate. That is, I'm constantly adjusting the pendulum length by small amounts, often every week. It's slow, then fast, then slow, etc, just can't seem to get it right. This hasn't happened for any of my other clocks.

It has a wooden pendulum rod with a metal bob, same as a number of other of my clocks, so I'm not sure what the difference is.

Does anyone know why this happens? Does it need to be oiled again? Rough pivots or pivot holes? Dirt somewhere?

I don't recall for sure, but I think it's got a run-of-the-mill T&S movement, nothing special, possibly Emperor branded but made by Hermle.

 

[Willie X]

It's probably getting ready to stop on ya.
Willie X

 

[wow]

Are you getting a good wide pendulum swing?

 

[bruce linde]

i'm curious to see where this thread goes... i have a marriage clock with a gorgeous ST 68 movement, seconds pendulum, 8" 5lb bob... i have been trying to regulate that sucker forever. i'm not sure if it's the regulating washing, karma, the dogs playing tricks on me, or what....

it's possible i need to go back (it's been a while) and possible re-bush the escape wheel and verge pivot holes just to make sure they're perfect, and look more closely at the regulating nut (and anything else i can think of)....

but in the meantime, i totally empathize... it's the one clock out of many that is finicky as heck.

 

[Vernon]

Elliott, it seems like you'll be taking it apart again. I would look for: cracked mainspring, bent pivots, arbors, teeth and pinions. It doesn't sound as it you're confident in your pivot work? Maybe review the procedures for that and recheck the need for bushings. Don't forget the need for barrel bushing. Finally, check the anchor for pitting and wear. My shotgun approach.

 

[shutterbug]

Over swing is the best indicator of a healthy clock. If you don't have much, you'll be chasing that rabbit for a long time.

 

[Willie X]

A beat amp, left in play over several hours, can locate small problems that cause poor rate keeping.

Every tick should be the same, variations in beat or volume (soft ticks) tell you there is a problem somewhere. Finding the fault can be very hard to do ...

Note, don't waste your time on a 'run of the mill' common clock movement. They may be very dependable and may have stood the test if time but unfortunately just aren't cut out to be good timekeepers. Five or ten minutes a week might be the best it can do. On the other hand, most high grade clocks or just about any 8-Day weight driven clock, is capable of keeping a rate inside one minute in a week.

My 2, Willie X

 

[Elliott Wolin]

Thanks for all the advice. I checked the swing and it looks good, the same as it's always been. This clock has never stopped on me and it's been running continuously for over a year. The beat looks good as well. And I'll check the overswing.

I'm wondering if Willie X is correct that it's just a cheap movement and I'm getting about what one would expect (I paid $5 for it, it wasn't running and needed work). I'll check how far off the clock gets over a week, I compulsively adjust it whenever I see it's off so I don't really know (I sometimes adjust my other clocks by a small amount when I wind them on Sundays, occasionally during the week...as I said, I'm compulsive!).

 

[Willie X]

Yep, OCD folks and those claiming to be "mechanical engineers" seem to have a really tough time with regulating their clocks ... Willie X

 

[Elliott Wolin]

Actually, I'm a (retired) research physicist, but I've done a lot of engineering (various types of mechanical/civil/cryo and some electrical).

I have two bathroom scales in my house and they disagree by about one pound. My wife thinks this is irrelevant, but I told her it really bothers my inner need for accuracy and precision in measuring devices. She just doesn't understand...

 

[Uhralt]

Thanks for all the advice. I checked the swing and it looks good, the same as it's always been. This clock has never stopped on me and it's been running continuously for over a year. The beat looks good as well. And I'll check the overswing.

I'm wondering if Willie X is correct that it's just a cheap movement and I'm getting about what one would expect (I paid $5 for it, it wasn't running and needed work). I'll check how far off the clock gets over a week, I compulsively adjust it whenever I see it's off so I don't really know (I sometimes adjust my other clocks by a small amount when I wind them on Sundays, occasionally during the week...as I said, I'm compulsive!).

Regulating at different time points during a week may be the problem. As Willie said, 5 minutes deviation a week is quite common. These simple movements often run a bit fast when fully wound and run slower as the spring runs down during the week. The best you can do is to regulate only at the same time during the week, for example on a Wednesday. Then the clock may be only a couple of minutes ahead during the first two days of the week and a couple of minutes late at the end of the week.

Uhralt

 

[Simon Holt]

I agree that trying to regulate a spring-driven clock on a day-by-day basis will lead to frustration.

Below is a graph showing how late a clock I have on test struck each hour over the course of a full week, starting off fully wound. Strangely, it started running fast on day 3...


Simon

 

[shutterbug]

I have two bathroom scales in my house and they disagree by about one pound. My wife thinks this is irrelevant, but I told her it really bothers my inner need for accuracy and precision in measuring devices. She just doesn't understand...

Put one foot on each scale and add the resultant weight. You'll only be 1/2 lb. off that way

 

[Schatznut]

Actually, I'm a (retired) research physicist, but I've done a lot of engineering (various types of mechanical/civil/cryo and some electrical).

I have two bathroom scales in my house and they disagree by about one pound. My wife thinks this is irrelevant, but I told her it really bothers my inner need for accuracy and precision in measuring devices. She just doesn't understand...

"A man with one bathroom scale knows how much he weighs. A man with two scales can never be sure." Averaging the two doesn't solve the conundrum either...

Below shows the difference between accuracy and precision. Elliott, you're fighting a lack of precision, whatever its cause, in your quest for accuracy. Anyone that has ever fought a torsion pendulum clock, trying to get it into regulation, wishes for much greater precision in the regulator on the pendulum.

 

[Elliott Wolin]

Elliott, you're fighting a lack of precision

I think I would put it differently:

Both scales are fairly precise. I.e. they both measure small changes in the weight put on them, and do so consistently and reproductively. To my mind, this implies that they are both precise measuring devices.

But one or both is wrong. This implies a lack of accuracy in one or both.

 

[Schatznut]

My observation regarding precision was aimed at the lack of precision in the regulator nut assembly, Elliott. If there was a vernier adjustment on that rating nut, you would be able to dial it in much more easily. Unfortunately, I've never seen such a thing. I wish the threads on the rating nuts on 400-day clocks were about five times finer. I chase the correct rate back and forth even making the most minute (my-noot) adjustments possible. Every once in a while i get lucky and hit the sweet spot.

I concur with your comments regarding the repeatability and accuracy of the bathroom scales. But repeatability is neither precision nor accuracy. One of them could be bang-on or they both could be off. Unless you have a known standard against which to compare one or both of them, you don't know about their accuracy, which is what matters.

 

[Willie X]

Fine adjustment of all pendulum clocks is done slow to fast.

400-Day clocks are the opposite. Fine adjustment is done fast to slow.

Willie X

 

[Elliott Wolin]

If there was a vernier adjustment on that rating nut, you would be able to dial it in much more easily

You are correct, although the rating nut threads are fairly fine, it likely isn't fine enough, I probably keep on over-adjusting or under-adjusting the pendulum length, never getting it right on. Just to make matters worse, the pendulum bob is not loose on the rod, so I have to be careful the bob moves when the nut is adjusted, especially when trying to make it longer. In fact, when I first got the clock running I kept turning the rating nut to no effect until I realized the bob wasn't moving.

Well, in the end, it's an inexpensive clock, maybe I'm getting all I can hope for. Perhaps one day I'll take it apart and give it a better overhaul than I did way back when I was just starting out.

 

[Dave Diel]

I had a problem of inconsistent timekeeping with one of the first clocks that I rebuilt. I tore that thing apart a total of 5 times before I finally figured out what the problem was:

The clock would run fast, I would lower the pendulum and reset the time. Then the clock would run faster and I would adjust and reset. Then it would run slow......very slow. Then, by some miracle it would keep time for a couple weeks and for seemingly no reason would start to run either fast or slow.

Finally figured out that Hermle didn't bother to put any chamfer on the pivot hole at the back of the handshaft. No matter how square the shoulder on the pivot is, there is some radius. The handshaft would move in or out slightly when I adjusted the time, which changed the amount of drag between the edge of the pivot hole and radius on the pivot. After I chamfered the pivot hole the timekeeping was rock solid.

I will never assemble another clock without checking that there is a small chamfer on the inside of all pivot holes. Likewise, all bushings get a chamfer.