Vertical holes - Test 1

The set up is as shown opposite. To minimised the sideway play in the drill, I have added a washer to take up some of the slack and must always push the lever away from me as the drill descends. In order to better measure the angle of the drill (from the vertical) and also to lessen the possibility of drill wander due to wood grain, I will use a thick sandwich of polystyrene and thin MDF.  The cross (top right corner) is equivalent to the blue block on the previous page diagrams. The thickness of the sandwich block varies from 55.08mm to 55.22mm (0.25%) on one side and 55.60mm to 55.79mm (0.34%) on the opposite one - but a total spread of 0.71mm in 150mm fore/aft (z axis).  However down the 1st testing sideways edge (x axis), the variations is only 0.40mm in 150mm - less than 0.2o slope (if it is uniform) - not bad! For reference, one degree is 0.96mm at a 55mm depth. The dowels, I shall insert in the holes to measure, have a maximum variation of 0.18mm in their 6mm diameter - attempting to sand one down, by placing it in the drill and holding sandpaper against it, did not remove the oval (guess there was too much spring in it). But by marking them (so the orientation is consistent), I can reduce the variation to 0.04mm. The opposite diagram shows two sets of holes (A & C of previous page) and for the right block which measurements will be taken.  By comparing the top and bottom measurements I should be able to deduce the drill wobble (black lines) as well as its angle (blue lines).  It is important to keep the block parallel to the drill in order to be able to adjust the working surface along the x & z axes. Anyway that's the theory - so let's try and see!

1st drillings Well it was an interesting test.  I found I could not line up the Brad drill very well (need glasses & some light). I also noticed that sometimes the top MDF bent down as it was drilled.  That is because the holes were on the edge. The results are entirely inconclusive.  I measured everything twice with an average variation of 0.07mm (maximum of 0.25) The dowel diameters varied by up to 0.17mm (top and bottom measures).  My measurements between the top and bottom of the drill holes were mostly around 0.15mm with 3 reading greater than 0.50mm (bold below). DRILL Measure  (top-bot) Average       HOLES one two variation   DOWELS    variations (top-bot) wobble           'inline' '90 degree' A-12 -0.17 -0.23 0.06 -0.20   0.03 0.03 A-14 0.16 0.17 0.01 0.16   -0.17 -0.16 C-56 0.08 0.12 0.04 0.10   0.04 0.00 C-57 0.25 0.15 0.10 0.20   0.03 -0.06 D-12 -0.01 -0.01 0.00 -0.01   averages   D-14 -0.06 -0.05 0.01 -0.05   -0.02 -0.05 B-56 0.19 0.18 0.01 0.19       B-57 0.83 0.92 0.09 0.88       angle               A-C 28 -0.35 -0.22 0.13 -0.28       A-C 48 -0.44 -0.69 0.25 -0.57       A-C 36 -0.26 -0.19 0.07 -0.23       A-C 37 0.00 0.05 0.05 0.03       D-B 28 -0.21 -0.06 0.15 -0.13       D-B 48 -0.72 -0.67 0.05 -0.69       D-B 36 0.25 0.26 0.01 0.26       D-B 37 0.09 0.10 0.01 0.10         average 0.07           min = 0, max = 0.25       All measurements are the difference between the outside edge of the pegs in the top and the bottom holes (not their distance apart). So my variation in measurement were half of the top/bottom differences (.07 to .15).  I will try again in the middle of the board. 2nd drilling I tried again, using a non-brad drill - amazingly the resultant hole size was different and my pegs were quite loose compared to the above.  This made measurements not very reliable.  Lesson learnt - never assume drills bits make the same size holes!! I also decided from now on to measure the outside of the pegs and their inside distance as (outside + inside) / 2 should give me the centre distances of the holes and remove the oval peg errors. 3rd drilling I used the Brad drill again, through 1/2" chipboard, for 3 holes (A1, 2 & 4), just to measure the wobble.  The measuring pegs were a tight fit.   chipboard test   pegs turned in last 2 rows of measurements   wobble measure1       measure2       top bot top-bot   top bot top-bot A-12 15.97 16.01 -0.04   15.94 16.02 -0.08 A-12 15.92 15.99 -0.07   15.965 16 -0.04 A-12 15.975 15.95 0.03   15.955 16 -0.04 A-12 15.93 16.005 -0.08   15.965 16.055 -0.09 Average 15.95 15.99 -0.04   15.96 16.02 -0.06 Spread 0.05 0.06 0.11   0.03 0.06 0.05                 14 12.31 12.37 -0.06   12.28 12.37 -0.09 14 12.295 12.3 0.00   12.32 12.32 0.00 14 12.185 12.315 -0.13   12.325 12.26 0.07 14 12.29 12.345 -0.05   12.25 12.305 -0.05 Average 12.27 12.33 -0.06   12.29 12.31 -0.02 Spread 0.11 0.06 0.06   0.07 0.11 0.16 There appears to be a consistent difference between the top and the bottom measurements, with the bottom always being larger than the top by an average of 0.045mm.  But the variation in measurement error makes this inconclusive as the spread of measurements averages 0.04mm. I wonder if I can improve on this experiment or can I only measure to an accuracy of around 0.05mm?

Next page Content Menu

.