The job went well. It would have been a breeze with a CNC mill! I used a
3/4" center cutting endmill plunged into the center then did a "X" and "Y"
feed close to the OD and a square pattern using the DRO. After doing this
to all 35 pockets I finished the ODs with a boring bar in a boring head.
This job would have been hell without the DRO.
It would be nice to have a CNC mill but I do so very few jobs that a CNC
would shine that I just can't justify it. Besides, doesn't a CNC setup make
it difficult to use the mill in manual mode?
I have run into this while looking for a mill, I found one that has a tracer
attachment to it, but the hydraulic controllers replaced the handles and could
never find any information if I could get handles to work with it as well.
I did see one cinci knee mill that had computer controller powerfeed 3 axis,
with dro's that still had the handles on all the axis, that was when I first
started looking and I did not ask questions as it was out of my price range at
5k. looking back,
I wish I had dug into it now that I have a CNC router/plasma table, it is
awesome, heck I used it just to cut rectangle backing plates that were 10-16 *
7-10" that I would had easily cut with the smaller plasma cutter I have square
enough to do that little job, I cut 4 plates at a time I needed 12, and I did
them all in half the time I could had done it by hand using a straight edge
If I had the money, I would have a cnc mill or some sort! the hardest part
would be learning to use the software to do 3d work.
I am about to setup to use this table to cut pockets myself for foam and will
also setup to tray some aluminum and see how that works out, I'm putting a 3
hp router on it to start, I looked into single motors but to learn they are
far out of price range for now
Not necessarily. I set my CNC up with handwheels. you can just turn
the servo off and turning by hand, the servo encoder still works so
you have a DRO.
I must say for several years i haven't used this feature. I have a
great manual mill that uses the same tooling. Its even more rigid than
my CNC with power feed on all axis. There are a lot of jobs that are
easier on a manual mill.
It depends. The first CNC mill which I encountered -- a
Taiwanese clone of a Bridgeport retrofitted by Anilim to CNC had Dished
handwheels with spring-loaded folding crank handles whose balance was
such that the handle would stay folded in at high speeds. (This was a
servo conversion, not a stepper one, FWIW, and the rapid motion was
scary. :-) Anyway -- stop the CNC control, and you have just enough
current through the servos to avoid self-feeding of the ball screws, and
you can easily use the handwheels to feed it -- with proper dials still
However, my Bridgeport BOSS-3 (being converted to servos) is
The X-axis ball leadscrew is rigidly mounted to the right end of
the table, and the ball nut is mounted between a pair of opposed tapered
bearings with the proper preload to minimize backlash). That ball nut
is rotated via a stepper (or servo) mounted well below the table on the
right side of the knee. No access by hand to the timing belts used to
The Y-axis ball leadscrew is rotated, but by a timing belt from
a stepper below it which is recessed into the knee, and it is all
covered. (There is a Veeder-Root counter to display the X and Y axis
positions.) But there is no access for a handwheel.
The Z-axis is actually a hollow ball screw which surrounds the
quill, so the feed is truly on axis (the Anilam conversion above used a
ball screw in place of the feed stop screw, so it is pressing off center
and would hasten wear of the quill in the headstock.) However, again
the ball nut is mounted in bearings and turned via a timing belt and a
motor off to the left of the headstock. The original stop screw is used
to actuate the feed limit switches instead, and can't be used to read
the position. There is a dial on the bottom of the pulley so you can
tell where in a rotation the motor (and ball screw) are, but not which
The control, when the CNC is stopped, has jog modes which allow
motion of 0.001", 0.010" 0.100" and 1.000" -- rotate a switch to point
to the desired size of the jog, and then push in the switch (which is
also a pushbutton) to cause the motion. I've actually done a little
machining with this, but it was a serious pain. :-)
So -- all in all, I would have to say that this one was not
suited for manual machining, and that the Anilam conversion (or some
other custom conversion of a manual mill) would be the better choice,
with the folding handle handwheels, so you don't get whacked by the
handle when you stand too close just when it is going into a rapid
Most any CNC mill can be used manually if it has an MPG. Just select the
axis and a suitable step rate and crank away. I can't really see why you
would want to though when you can just "manually" machine one line of G
code at a time in MDI mode.
By MPG I presume you mean Manual Pulse Generator. The
Bridgeport did not have that, and while there were complete schematics,
the CPU was a DEC LSI-11, and there was no source code for it to allow
adding such things to it. (Not to mention that it was limited to 64K
bytes (32K words) of memory -- both the G-code program memory, and the
machine language code (in ROMs) which translated the G-codes into actual
motion. This was in the early days of CNC. Bridgeport's BOSS-3 was the
first model to escape form the factory (nobody seems to know what
happened to the BOSS-1 and BOSS-2 -- presumably died of too many bugs to
be released. :-) They stayed with the stepper motors up through the
BOSS-6, and I'm pretty sure that the BOSS-8 was servos, but I don't know
whether the BOSS-7 was similarly invisible or not. :-)
Nobody ever released a version of EMC/LinuxCNC for the LSI-11,
and I doubt that anyone would bother with the limitations of memory.
(Back then, they even measured the memory in "feet" (feet of punched
paper tape, at 10 bytes per inch. That whole address space would only
hold 546.1333 feet of tape -- even if no space were being used for the
code in ROM. :-) This is one reason why they came fitted with a punched
tape reader. If you set your program up right, you could load almost as
much as memory was available for, then run that part, and when it was
done, it would automatically read the next chunk of tape to proceed with
the program. Then, you were only limited by how much tape you could get
on a reel. (Or maybe even more, if you could swap tapes while it was
running the last chuck which would fit on one tape -- but that required
being a bit more alert than CNC made common.
Why not three MPGs -- one per axis, so you could crank two axes
at once -- or even three if you put one in reach of your knee. :-)
Obviously, as I convert it to servos, a MPG is a nice feature to
add for positioning and for testing layouts with the system serving as a
It's kinda fun sitting on the table of a large gantry machining center,
MPG pod in hand, dialing around the table you're sitting on (X) as well
as the gantry (Y) and head (Z) as you reinstall the spindle cartridge
after rebuilding it.
FWIW, even the latest Fanuc CNCs drip-feed large programs to the
control's memory from a remote DNC computer. It only has 8 Mb, I
think, and it's all a special flash memory that's designed to last for
Fanuc is obsessive about reliability and will not put a hard drive in
their controls -- they say hard drives have no place on a factory
I've used a Bridgeport ProtoTRAK mostly in manual mode with the jog
dial. It's a little less convenient in some ways than turning the
handwheels, but I didn't have to reach as far. It was good enough for
one-off prototype jobs that often included milling a casting to a
scribed line or drilling at transfer punch marks.
Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here.
All logos and trade names are the property of their respective owners.