Pretty close. What happens to the runout when you shift from
light to tight tightening of the chuck?
BTW Drill shanks are soft, and the chuck's jaws can bite into them
so they won't slip. End mill shanks are hardened and ground
smooth so they work well in collets or end mill holders, and are
*not* best used in a drill chuck. (The same applies to the
center drills, FWIW.)
The soft shanks on drill bits also means that it is possible for
the shank to bend, so the tip will describe a circle even if the chuck
And -- you are unlikely to find a long drill bit which has a
shank long enough to allow runout measurements. (Aircraft drills are
the exception -- but they may be too long for your machine.)
O.K. So R8 is not in the game. It is possible that the arbor
(MT-3 to whatever Jacobs taper the chuck uses) is bent -- but less
likely with that 0.0005" TIR. This then suggests that the jaws in the
chuck either have burrs in them, or there is wear in the body of the
chuck. Different brands of chucks have different quality levels. The
Albrecht keyless chucks are very accurate and can handle a long service
before they start showing problems. (But they are larger, so they leave
less space between the spindle and the table, so check for that too.)
What happens if you use collets for the center drill and the
plain drills? (Assuming that the plain drill is of a size which matches
a collet which you have.)
They are hardened and ground -- but fairly short usually. The
best thing is drill blanks, which are hardened and ground to precise
diameters -- but much longer. (And quite expensive, which is why I
don't have an index full of the standard sizes. :-) But that length
makes it easier to check for the drill chuck holding the drill at an
angle, thus causing the circle. Measure runout just below the chuck,
and close to the end (but before the diameter starts to fall).
[ ... ]
Aha! This is what I was answering somewhere above. You are
unlikely to find a drill whose smooth shank is long enough. Usually the
extra length is in the flutes.
[ ... ]
So -- use magnification -- and a small LED flashlight for
And if the axis of the chuck is not parallel to that of the
spindle, you will still have the point wandering in a circle. One of
the reasons for the sliding offset is so you can push it while spinning
slowly until the point stops moving in a circle. That is the center of
the spindle axis -- but perhaps not the center of the chuck's own axis.
But both of those are longer than the center drill, so you will
have had to move the head up to get them in the space. And, BTW, it is
possible that your column is truly vertical, but the head is slightly
off so its axis is at a tiny angle to the column. Another reason for
measuring with a machinist's square and a drill blank.
Yes -- as someone else suggested, hardened steel rods from dead
printers (or some older dead disk drives) can be used in place of the
drill blanks -- if not badly worn.
Give in the chuck somewhere -- a jaw with a chip under it, or
something similar. Try another chuck. Or check with collets.
Your No. 1 is not fully deep in the chuck, so you are only
holding it with the tip of the jaws -- and they may be flexing in their
ways. Try another chuck. Yes, more money for a chuck *does* mean a
more accurate one.
Yes -- flex of the jaws in the body when you don't have a shank
contacting the jaws the full length into the chuck is a possibility.
Again -- try collets.
So, here is the result of 3 hours' work:
1) I found a dowel pin. At least I am pretty sure it is a dowel pin.
It is pretty and shiny, 0.5" x 6", it has those black caps either end
and on it it says USA made, 1/2". I do not know how I got it.
2) In the chuck, using a different indicator from last time, the TIR
<3/8" under the jaws, 2" and 4" respectively was 0.001", 0.003" and
0.006". Re-tightening made no difference.
3) In a collet the figures were <0.001", 0.0015" and 0.002"
4) Repeat test with the same 3/8" drill I used before the TIR under
jaws was 0.001"
5) I tried my 3/8" reamer. Under the jaws the TIR was the same but
2-1/4" down the shank 0.0115" !
6) The No. 3 center drill runs out at 0.008"
7) The No. 4 at 0.007"
8) I repeated the center finding tests. the two instruments varied
sometimes by as much as 0.02"
9) As the punch mark have definite dimensions I tried just locating
crossed lines. This is quite difficult: Just because you can see the
lines on the bench does not mean you will see them clearly under the
spindle, paint etc. notwithstanding. The results, confirmed with
center drilling , were considerably worse than center-finding a punch
mark, however small.
10) In view of the discussion re: spotting drills vs. center drills I
thought it made more sense to abandon using No.1 and No.2 and I used
No. 3 instead to make a small dimple with the pilot only. This should
approximate a starter hole produced by a spotting drill. This was a
definite improvement and the difference between the center drill hole
and subsequent 1/8" twist drill position was only 0.005" on the Y axis
It should be noted that I used my machine screw length drills so the
actual distance from the chuck jaws to the point was pretty much the
same for both the center drill and the twist drill.
11) I cleared the table and returned to the dowel pin in the collet. I
used the machinist square and could not see any deviation at all. The
two squares I have pretty much agreed with each other. Note I did this
along the X axis with the spindle in 4 different positions. The table
is too small to do this effectively along the Y axis.
a) The machine was out of tram to the tune of 0.008" over the X axis.
This was quickly corrected to <0.0005" with the head low. With the
head high this increased to 0.0035".
b) Moving the table along the Y axis produced no significant change. I
interpret this as the table being flat.
c) A spindle sweep in an arc from back to front, however, showed a
change of 0.005" over the 2" available in the Y axis. I interpret this
as the axis of the spindle being off in the Y axis plane.
1) Do not use digital indicators for this sort of work.
2) Change the center drilling procedure as outlined above.
3) Grind a spotting drill?
4) To locate cross lines on a work piece it seems preferable to use
optical punch first and then locate the center on the machine. This
probably makes center finder and spotting drill unnecessary and can be
done with a twist drill directly.
5) Consider acquiring better center drills.
6) Use collets whenever possible.
7) Ideally one should try to correct the spindle axis. From what I
read of others' experience this is very difficult with this machine.
It is what it is. Like they say on South Park "I have learned
something today". I should be able to minimize the errors by being
aware of the pitfalls.
Thanks again for the help.
Campbell River, BC