There are many other effects that can produce MicroSet readings that aren't stable. One important thing to consider is the LEVEL control on MicroSet. If it's set too low, you may miss some beats and you'll get results that are too slow. If it's set too high, you may pick up extra sounds as aditional beats and get readings that are too fast.
MicroSet incorporates a "Blanking Window" to help eliminate additional sounds that might get counted as beats. Be sure to read about the Blanking Window in the instruction book. It's very effective. You can also use the SETUP MODE to specify a minimum Blanking Window. If you make the Blanking Window as long as possible for the timepieces you measure, you'll have less trouble with extraneous sounds. If you will be measuring very fast (36,000 BPH) watches, the minimum value for BLANKING must be as small as 4. If you don't work on fast beat watches, you should set the minimum BLANKING interval to 8.
The most significant way to improve the accuracy of your readings on pendulum clocks is to use the optical sensor rather than the acoustic sensor. People tend to favor the acoustic sensor because it's easy to clip on. But it's always succeptible to noise and false triggering. The optical sensor does not have these problems. In addition, the optical sensor measures at the pendulum rather than at the pallets on the escape wheel. The pallets and escape wheel are very low mass components and are succeptible to irregular movement. The pendulum is heavy and much more consistent in motion. Always use the optical sensor if you can.
You may still set an accurate rate on a clock yet find it to be fast or slow the next day. There are several reasons this can happen, and they're not because MicroSet is innacurate. For one thing, clocks respond to temperature. If you rate a clock on a warm afternoon, it may run substantially faster after a cold night. The graph below shows the same Hermle we've been working with. It runs about one minute per day faster in the morning than it did the previous afternoon.
Many clocks slow down a bit when they strike. You may find that striking clocks average a slower rate than the rate you measure when they're not striking.
Clocks with very delicate movements run slower when the minute hand is being raised from six to twelve and faster as the minute hand falls from twelve to six. You can easily measure this on 400 day clocks. More accurate results will be obtained if you measure when the minute hand is near the twelve or six rather than near the three or nine.
The bottom line is this: MicroSet is very accurate. If you get irregular results, your first thought should be that this is really what the clock is doing in the interval you have measured it. Think about the eccentric escape wheel and the LEVEL control. Make adjustments to get more accurate readings. Use the optical sensor on pendulums.
When MicroSet is hooked up to a precision reference, it will measure the same reading within one part per million for hour after hour. People sometimes return a timer after a year or two to have a new feature added. I always recalibrate the timer in these cases. They're never off by more than one or two parts per million, an error of just one second per week.
If you have questions feel free to give us a call or send an email. We're happy to help you understand what MicroSet is telling you.
