For best results we now recommend the use of the Peterson Vibration pick-up.
It is a piezo-acoustic contact sensor, with clip mount and requires no additional adapters to connect to the iPhone! It provides best vibration pick-up and is not affected by room noise.
Link to Amazon store.
The internal iPhone microphone will work just fine with the app, however, an external microphone will always work better than the internal phone microphone. These will isolate the ticks better and eliminate extraneous sounds. The mic on the iPhone earbuds is a great improvement as its a noise canceling microphone and helps with the background noises that may occur. This also will allow placement of the microphone inside the case and then the iPhone can then be laid down for longer periods of testing.
A further improvement has been found to be the Griffin SmartTalk, which is an external phone microphone that allows you to add your own earbuds with the microphone. The advantage of this is that you do not have earbuds with it hanging down from your clock case in the way - just the microphone on a 3 foot cable. We have found this to be a much better microphone than the iPhone earbuds mic. It cancels external noise much better and allows for a lower sensitivity level setting.
The next best microphone, would be a dedicated external microphone like those used for video recording. These, however, require the use of an iPhone microphone adapter cable to convert these mics two pin 1/8th in plug into a 3 pin plug. (The Apple earbuds and the Griffin SmartTalk are already 3 pin plugs.)
Truly the best microphone would be a piezo-acoustic sensor attached to the clock. They can pick very slight vibrations, and are not affected by ambient noises. These would also require the use of the adapter cable cable though.
|iPhone earbud mic||Griffin SmartTalk||Piezo-acoustic sensor|
It is imperative to mention that the longer your test sample is, the more stable and more accurate the Error Rate will be. A test of at least ten minutes or more should be conducted for a Rate of any value. 30 minutes or more will allow for more complete analyzation of the data should you wish to go more in depth for troubleshooting purposes.
You will not be able to achieve perfect zero due to the inherent errors of the clocks escapement teeth. Adjust as close as possible. For example, on a mantel clock you can level the clock by raising one side or the other of the clock by small amounts and listen for even ticks. Watch the Δ for finer adjustment. If the clock is on the wall, then move the bottom of the case left or right by a small amount and listen for evenness in the ticks. Again use the Δ for finer adjustment. For gross adjustments, the pendulum crutch needs to be adjusted. This should only be attempted by a qualified clockmaker.
After a stable Error Rate has been determined, the weight on the pendulum needs to be raised or lowered by turned the nut or screw holding it onto the pendulum. Small amounts are all that are necessary. Half a turn is a good starting point for a minute of error. This of course will vary greatly depending on the type of clock and length of the pendulum. Let the pendulum settle its swing before retesting. Some mantel or bracket clocks are adjusted from the front of the dial with a key. This adjusts the length of the pendulum and has the same effect. Again small adjustments.
The raw data measured time between ticks is saved in a .txt file. These times are measured to the 15th decimal place. An example of the data file is below:
These times can then simply be imported to your favourite spreadsheet for further use. For example, the times can then be charted or used to figure the BPM for each tick, or the running error average at each tick and then charted. Each chart would then have its own information to give. A specific pattern can be found, an instability figured or a bad tooth on a wheel can be found.