trouble regulating a NYSWco. clock.

[spxer]

I have had this NYSWco. electric clock running for months now. It runs well, but the time keeping is questionable. It will run to the second for a week or three. Then it suddenly starts gaining or losing. I adjust it with the micro adjustment at the bottom of the pendulum. After some adjustment it runs accurately again for 1-3 weeks, then it starts the gain or lose cycle over. Am I expecting too much accuracy from the clock? BTW, I am using Atomic Time app for reference.

 

[Jmeechie]

Hi spxer,
My recommendation would be to maintain a log of what is occurring and what you did to correct. A running log book with dates, inaccuracy, correction and weather conditions. Mechanical clocks, and I know I’ll take some heat on this!, are at best excellent time keepers but are greatly affected by temperature, humidity and other outside influences besides their own friction changes. Hipp-Toggles are noted for battery consumption so I would have to ask how you are powering the clock as battery drop off will greatly effect motive force and pendulum swing?
Also, you note gaining or loosing bit by how much? For example, most master clocks SETco, IBM, Brillie, Synchronome all advance the slave every 30 seconds so they could be off by as much as 30 seconds and that would technically be acceptable time keeping.
BTW I love these big NYStd’s and yours is a great looking clock!
Cheers,
James

 

[spxer]

James, I have the clock powered by 8 D-cells wired to provide 3 volts. I have been keeping a record of sorts. I write down the date and seconds when the clock deviates from the Atomic Time app. I am wondering if I am not allowing the clock as much leeway as it needs. I start correcting it at 3 seconds off. I have not noted weather conditions.

 

[TQ60]

Do not touch the clock...

Does the clock have a beat scale, if not, make one, just print something and tape it behind the bottom of the pendulum.

Get a calendar And a small thermometer that can be placed inside the clock and viewable without disturbing clock.

You can do a schedule or by happenstance, note Time of day, temperature, error and pendulum swing on your calendar.

Over time you can review your data to see what is changing.

You may find the rate related to temperature as an example.

In our case, we have an Atmos 526 that had given us some trouble over some time.

Finally got it going well, the rotation would vary depending on temperature.

It is outside, upstairs in the not climate controlled shop.

During winter at 40 degrees it would be at 400 or so degrees and this last weekend at 90 degrees it was 600 degrees.

Everything moves with temperature, sometimes it will surprise you.

 

[Jmeechie]

James, I have the clock powered by 8 D-cells wired to provide 3 volts. I have been keeping a record of sorts. I write down the date and seconds when the clock deviates from the Atomic Time app. I am wondering if I am not allowing the clock as much leeway as it needs. I start correcting it at 3 seconds off. I have not noted weather conditions.

Spxer, my first question is how long (days, weeks….) before it has gained or lost 3 seconds? If it’s more than a day I’d say IMHO you aren’t going to get it closer. It sounds like you have a good power supply and are at the correct voltage.
Temperature, humidity and barometric pressure all have an effect on the pendulum swing and friction in the movement. The wood pendulum rod is not ideal for precision accuracy. Both the pendulum rod and bob were the attention of many a precision clock maker using Invar steel rods and mercury compensated bobs to name just a couple of variations on achieving accurate time keeping. Wear also plays a factor and these movements do suffer from wear at the armature pivots. This will cause erratic impulse to the pendulum from increased or decreased switch engagement also effecting the pendulum arc and accuracy.
Again, keep a log book and track the daily weather, note time keeping errors and only correct on a weekly (not daily) basis, noting amount of correction at rating nut. After a month or so let us know what you find.
Cheers,
James

 

[spxer]

I suppose I am expecting more accuracy than I should because it does run dead on for weeks at a time. My question now is how far off do I let it go before I make a correction?

 

[Jmeechie]

Ah, ok, a couple weeks before you gain a minutes drift is pretty darn good!
I’d still be inclined to keep a daily log of accuracy and weather conditions. I’d probably consider any minor adjustments once a month and not any sooner as you’ll want to see if there is any gaining followed by loosing without touching it and making any adjustments.
Cheers,
James

 

[spxer]

So I should think in terms of minutes and not seconds.
I understand about the weather/temperature, but there is very little I can do about it. Even if I see a correlation, what good does that do, except to say there it is?
I will try and contain myself and let the clock run to 1 minute out.

 

[Jmeechie]

Yes. Only reason I was suggesting keeping track and not messing with it for a month was in the event it gained a minute after several weeks and then lost 30 seconds going into the fourth week.
Certainly these clocks never were atomically correct to the second but once rated were fairly close after a month ms running. You stopping the clock (pendulum) to regulate and then restarting will take a short period of time for the swing to settle down to where it wants to be. This was one of the reasons I was recommending not to touch the clock for a month.
Cheers,
James

 

[spxer]

Actually with the micro adjuster at the bottom of the pendulum one does not stop the swing. Just a touch of a pin as the pendulum swings by. There is a slight disturbance of course.

 

[IDPomfret]

Actually with the micro adjuster at the bottom of the pendulum one does not stop the swing. Just a touch of a pin as the pendulum swings by. There is a slight disturbance of course.

View attachment 764103

 

[IDPomfret]

I cut my electrical horology teeth on one of those 25 years ago and found them very difficult. The problem is the impulse is controlled by a Hipp toggle which relies on a decrease in the amplitude before re-impulsing the pendulum.
Alas the mechanically controlled time circuit, relies on a very fine ratchet system to work and it just occasionally gets too big or to small an impulse and will gather an extra tooth or lose one.
Best wishes with this project.
Ian

 

[spxer]

Would anyone have a write-up of how to set the various adjustment of the electrical switch? I adjusted it until I got it running good.
Some seem to think a long span between impulses is desirable. I am not so sure, in line with what problems Ian described. The clock goes 30 seconds or so between impulses as set.