The collet is a new SB, sk-polishing offers these frequently on eBay
and I took advantage when one didn't seem to draw many bids. I have
accepted it as a standard in testing my collet chuck, at least so
far...
Your comment on working to tenths is very interesting. Having no
experience or training in machining I've been developing my own
methods on an ad hoc basis. I've run into the problem you mentioned
of "noise" on the DTI due to surface finish; my crude solution has
been to use fine carbide paper to polish the surface a bit prior to
taking the reading. I then wipe the surface with a paper towel and
add a little oil. My measuring technique is to run the lathe at its
lowest speed, about 12 rpm, and watch the DTI. The needle is still a
little shivery but watching several cycles allows me to get a reading
by averaging the maximum and minimum readings. I am open to any
improvements in technique since I am very much a beginner at this.
Your point on accepting a couple of mils of runout is one I have
slowly arrived at in using the 3 jaw. Lately, I try to leave the
diameter on the large side if a part will need to be removed and re-
installed in the chuck, going to the finish size as the last step
when possible.
Thanks for the info.
John
--- In 7x12minilathe@..., "roylowenthal"
<roylowenthal@y...> wrote:
I wonder if the brass collet is part of the problem. Coming
from
e-bay, there's no telling what sort of abuse it may have suffered.
Steel collets are "spring hard" to avoid distortion during use;
they
eventually wear eccentric. (Long eventual, except in high volume
production.)
Trying to work to tenths is sort of frustrating - surface finish
of mating parts has a disproportionate effect on readings.
Actually,
surface finish on the part being indicated can "drag" the indicator
to a slightly false reading. Presumably, your measuring technique
gives the same reading on return to the same point. There's enough
flex in these machines to get a measurable deflection from resting
your forearm on the headstock.
Except for really critical work, ignoring any runout under .001"
or .002" saves a lot of set-up time.
Roy
--- In 7x12minilathe@..., "John" <moran03@e...> wrote:
Hi Roy,
Thanks for the link. I looked at the site and wrote to the
owner,
Dave Audette, who provided a considerable amount of information
in
a
detailed reply. One key bit of data is that his runout is about
4
tenths, similar to mine when the collet settles in the chuck
properly. I also found a CNC site which gave collet runout as 5
tenths max so apparently my collet chuck's performance is in the
ballpark.
However, it is interesting that Dave gets 4 tenths runout with a
SB
collet holder which I assume is probably concentric to the
headstock
MT3 socket. Makes me think my headstock may be poorer than most
since I find the following when indicating various points on the
backplate:
Outer rim = 5 tenths
Land for chuck registry = 2
Inside of MT3 socket = 6+
Inside of 3C chuck = 2
Piece in collet = 2-4 typical, sometimes 10 (tapping with a block
of
wood causes it to pop into alignment yielding 2-4)
Logically, a perfectly concentric collet chuck's runout could be
no
better than the socket it is mounted in.I use a witness mark to
align
my collet chuck to the backplate/MT3 socket so the socket runout
should be cancelled because the chuck was machined in that socket
and
replaced in the same orientation. That is, my collet chuck is
slightly eccentric but in a way which compensates for the MT3
socket's runout.
The above figures indicate (no pun intended) that the runout of
the
MT3 socket in my backplate is larger than desirable and that
Dave's
machine seems to run truer by quite a bit since he gets the same
runout without having machined his chuck in place.
Based on the above I would guess that during manufacture the
backplate is turned to size on a production machine with the
chuck
land left slightly larger than final size. The backplate is then
mounted on the lathe and the chuck land is machined to final size
in
place to minimize runout.
I intend to polish the inside of the collet chuck in an attempt
to
cause it to register the collet the same way each time so I don't
sometimes have to tap the work with a block of wood.
John
--- In 7x12minilathe@..., "roylowenthal"
<roylowenthal@y...> wrote:
See if there's anything here that helps:
Roy
