Googling around, I could not find any information on how to do this with simple tools in the home shop. Some commercial methods which come to mind are using an optical comparator or 3d laser scanner. While taking myself and the dogs for a walk this morning at 5:30, I realised, like many times before; the old school machinists did not have all the fancy measuring equipment we have noways, yet managed to build very accurate and precise machines. The idea of tolerance measurement similar to how snap gauges work came to mind.
To do this one has to measure the ball diameter. There are a few methods I can come up with to do
this. The intention is to measure diameter as accurately as possible, not just variations in the Lot or precision.
| Tool | Method | Accuracy | Repetability | Issues | |
| Micrometer | place the ball between the anvil and measure | Good | Fair | Measuring Force | |
| Dial indicator | Capture ball on stand, compare height of balls. | Relative | Good | Allignment | |
| Slip Gauge | Build gate on surface plate with
high and then low tolerance, roll balls through the gate |
Best | Best | Cleanliness |
In a previous post I have described the calculations and selection of ball diameter. For reference, I am working with Grade 10 Balls, 13/32" or 10.31875mm
| Grade | Sphericity [mm] | Lot diameter variation [mm] | Nominal ball diameter tolerance [mm] +/- | Maximum surface roughness (Ra) [µm] |
| 10 | 0.00025 | 0.00025 | 0.0013 | 0.025 |
My method will firstly select balls which roll under the gauge block stack bridge setup for the high tolerance level of 10.32005mm, and then from this selection set, select the balls which do not pass under the low tolerance stack of 10.31745 mm. This is the simplest, and most accurate and precise method I could think of for the home shop machinist.
Great resource with lots of useful information:
http://www.precisionballs.com/Ball_diameter_Calibration.php
http://www.precisionballs.com/Ball_Diameter_Errors.php
