We changed our strategy today after yesterday's tests (that involved moving the Faro around & measuring the same points over & over) produced unacceptable results. As a last resort before declaring this arm unusable, we decided to keep the Faro in one place & move the measurement points around to see how that works...
We employed a steel bar with 4 measurement points drilled along its length for this purpose. The bar can be clamped to the Faro test plate in practically any position, allowing the arm to measure the same lengths (between the 4 points on the bar) in a variety of ways. This lets us exercise all the arm's encoders during the measurements, thereby averaging out the encoder errors.
This produced remarkably good results so although the AMNH Faro Arm's out of spec, the mean values obtained from the various sets of 1000-point measurements are in good enough agreement for our purposes. When using the arm in anger we'll need to take 1000-point averages & make sure we exercise the encoders as much as possible.
No comments:
Post a Comment