Rate Variation

[bbodnyk]

I received a TYMG MTG-1000 Timergraph the other day and have been putting my pocket watches on it to check their rates. Several of my watches exhibit a varying rate which would look like a sine wave if plotted. Generally with maybe a minute or two from peak to peak. What could be causing this type of variation?

Also the timer lets me see the amplitude. What amplitude threshold should I use to determine that something needs to be done to improve it?

Thanks!
Bruce

[doug sinclair]

These modern timers tend to assess the rate and charcteristics of the performance of mechanical watches, almost microscopically! Can any quality watch in good condition, rated and timed to a high rate of accuracy actually pass the close scrutiny of these machines? I'll be interested in the answers to your question! My Rolex Date-Just which I have recently serviced, looks perfect on my Vibrograph. And on the wrist, it performs to within 5 seconds a week! But it sounds as though it wouldn't pass according to these modern instruments.

[LarFure]

Bruce,
There are two reasons,that I can think of,that would cause a line like you describe. One being magnetism,and the other would be a loose roller jewel.

Doug,
I have an old watchmaster timing machine and a new solid state machine. The new machine will put you to task in making a watch run to close tolerances,but the old machine makes for a better diagnostic tool. I find that I'll use the old machine when I'm trying to diagnose a problem,and do the fine tuning with the new machine when the movement is in good running order. BTW,I just finished an Elgin grade 509 and here are my results for positional rate. DU - 10 sec.,DD - 10 sec.,PU - 10 sec.,PL - 10 sec.,and PR - 15 sec.. I'm going to call this one good enough and not try to hunt down that 5 second discrepancy.

[Smudgy]

The variation is likely an out of round wheel. It'sm probably a small variation in the escapewheel, but the period of the variation will identify which wheel it actually is. I haven't used the machine you are talking about, but the newer machines have the ability to change the scales on the screen (zoom out/zoom in). If you look at the scale over a larger range of numbers (zzom out) the sinusoidal wave will flatten out. Keep an eye on the scales when you are diagnosing the watch. Often, changing the scale will help clear up some of the data (as well as averaging and grouping data).

[bbodnyk]

Searching through earlier postings someone mentioned the book "Better Watch Repairing Faster" by Max Lee as a excellent book on helping interpret readings from timing machines. Amazon had two and I bought one, a bit pricey but well worth the money if it helps diagnose problems. Clearly reoccuring patterns in the dots being displayed on the timing machine display mean something and hopefully Max Lee's book which reportedly has many illustrations in it will help explain them.

An out of round wheel as Smudgy suggested makes the most sense to me; I've already demagnitized the watch. It would seem to me a loose roller jewel would cause a more random variation than the regular variation I'm seeing.

Thanks to All!
Bruce

[LarFure]

Most timing machines only give a thirty second sample of actual run time. I doubt that an out of round wheel fault will show up in that time period. Think about it. The fourth wheel would only make a half revolution in that time frame,and an escape wheel fault would give a different line than what you are seeing now.

[Don Dahlberg]

An out of round wheel will do that, but so will too much sideshake in any pivot. As the a tooth and leaf come in contact, they push away from each other if the pivots allow them to. The energy transfered from the tooth to the leaf changes through the interaction. The period of the oscillating rate is a strong hint. For example, if the period is about 5 minutes, it is probably a worn mainspring arbor hole, usually in the barrel cap. It takes about five minutes betweet one tooth and the next interacting with the center pinion.

If the period is shorter, then the problem is with a faster moving wheel down the train with one of the faster moving wheels.

If there is a wheel out of round, then the period would be the time it takes for a complete revolution of the wheel. For example, the fourth wheel makes a revolution every minute. The center wheel would take a whole hour.

By the way, this is why some of us keep our B200 timing machines. They make a paper trace. We can put a watch on the machine and come back ten minutes later and measure the period of the sine wave.

Don