Or, a broader question is “How do the data from the WDSS and Motion Detector compare? Both will display acceleration values.”

The sync between the WDSS and any other interface is approximate, although it can be quite good. The WDSS has its clock; the LabPro or other interface has another. Each clock has a characteristic delay in starting, plus some jitter (random variations in that start delay). Part of that jitter is from Windows not being a real-time operating system, and part is from the interface. Logger Pro can account for the average characteristic delay, but not for the jitter, since that is different from run to run.

Of the various interfaces (WDSS, Go Link, Go Motion, Go Temp, LabPro) the LabPro has by far the largest jitter.

A common combination of sensors to use is the motion detector and the force sensor or accelerometer of the WDSS. This invites a comparison between the acceleration data of the motion detector and the WDSS acceleration data.

If you are going to do such a comparison, it’s important to use a very simple experiment so we’re not troubleshooting a complex experiment at the same time. I used a WDSS clipped to a clipboard, moved up and down with an up-pause-down-pause-up-pause motion, which generated nice short acceleration peaks. See the attached Logger Pro files for the example.

You might look for a time shift between the two data streams, and you can also compare the magnitude of the readings.

In comparing the magnitude it is very important to set the number of points used for derivative calculations to 3, the minimum value. Logger Pro then uses three position values to find one acceleration value. That means that there will be some smearing out in time of the acceleration graph, but not too much. If a larger number of points is used, the acceleration graph is much quieter, but the magnitude of short-lasting accelerations (the peaks) will be reduced due the time-smearing. You can easily see this by taking some data and changing the number of points setting. See Settings for (file name) in the File menu to change the number of points used in derivative calculations. Setting it to three will make the acceleration data noisier.

In comparing the time values of the two devices (eg, when do peaks occur?) you need simple data with peaks. I set up a Go!Motion and a WDSS taking acceleration data to do the above-mentioned clipboard motion experiment. Data rate was 250 Hz, which means the WDSS gives 250 points/second, and the motion sensor was at 50 points/second.

The sync between the wdss acceleration stream and the motion acc stream was essentially perfect, meaning that the peaks were within one motion detector collection interval, or 20 ms. It can’t get better than that, given the relatively slow data rate of the motion sensor.

Next I did the same experiment with a LabPro, Motion Detector and WDSS. This time the relative delay was apparent; the LabPro data were 38 ms early, meaning that the LabPro started 38 ms late. That’s about two data points in the motion detector data stream.

The difference is apparent in an experiment set up to reveal the difference. Is a 38 ms difference good or bad? For most motion experiments done in a classroom, that’s not a problem at all–but there is a difference.

Now, it is possible that a different computer would exhibit different average delays in the start of data collection when two devices are used.

Related TILs:
How good is synchronization between two WDSS? Good enough for impulse experiments?