Getting the feel for the Tormach 440; an interesting machine! Not the most powerful spindle and it doesn't reverse so it's no good for tapping, but it's a sweet design. Hoping to add ATC when it becomes available, hopefully end of yr.
It doesn't use Mach 3, (which I'd finally become used to) but Path Pilot instead. Still a bit of struggle to get that one working right. Fortunately I've got a pal who writes code for a living and he's being my 'brains' while I figure it all out.
"Path Pilot" is Tormach's "skin" for LinuxCNC - or maybe Machinekit, which
is largely the same. Any particular problems you are having with it?
You might want to join the LinuxCNC user's group at
Also, check out the web site at
There's a wiki, forums and a place to download documentation.
I've been using EMC since 1998, and stuck with it from EMC to EMC2 to
Reverse? Reverse what?
Reverse the toolpath, as in back out several moves in a wire EDM manuever?
There is a very recent version that has this feature. it may be only in
Machinekit, and not in the regular LinuxCNC, I don't quite remember. But,
this feature has been developed. It only will reverse about 20 G-code
blocks, I think.
If reverse toolpath is not what you are thinking about, then please clarify.
Well, you need a spindle drive that is reversible, and a connection to the
control to command that. If the Tormach machine just has two spindle
contactors, that is not going to reverse smoothly enough with a tap deep in
I just rigid tapped 100 holes in 2-56 and 6-32 sizes on my Bridgeport with
LinuxCNC. I have an encoder rigged into the bull gear in the head
(Bridgeport heads do not allow a standard encoder to be easily fitted).
And, I have a VFD with forward and reverse commmands rigged to two digital
outputs on my digital I/O board. To make sure the reversal is smooth, I
have a software filter in line with the speed command, so the VFD ramps down
to low speed before the command to reverse it.
I make some production parts that have 4-40 holes in them, and have done
thousands of them with rigid tapping, using combo drill-taps, which make it
extremely fast and easy.
I don;t know the Tormach machines very well, although I've attended a couple
meetings at their shop. So, I don't know how they deal with spindle
And, I'm sure a number of their machines have exactly that.
Well, I rolled my own:
On the 1J Bridgeport, the bull gear doesn't move. On the 2J, the bull gear
moves up and down on a spline to engage the back gear, so this gets more
I use machine taps and combo drill-taps. I do 4-40 and smaller at 1000 RPM.
I've done THOUSANDS of 4-40 holes using the same tap. I break a tap once in
a while when I forget some step in the setup, otherwise I keep using the
same tap until I make a mistake. If the first hole is good, the rest of the
job goes fine. I use alum-tap from Wibro, fantastic tapping fluid.
I used a Procunier "CNC" tapping head for a while, it has the reversing
gears and clutch that engage for push and pull on the tap holder. It worked
quite well, but was a VERY long extension from the spindle, and had to be
set just right in the collet so it didn't wobble.
I have about a 3" section of hex bolt on the end of my drawbar. When the
quill is extended, that completely disappears down into the head. On the
1J, there is a bearing for the cone pulley, that moves up and down about a
quarter inch to engage/disengage the direct drive clutch. This always runs
at input pulley speed, so when in back gear, it is NOT running at spindle
speed, but about 6X spindle speed. So, you can't take the encoder off that.
The outer spline for the spindle is WAY down in the hole.
I'm having trouble imagining how such a thing could be fitted to a 1J. Two
ways - you could have a long tume that somehow attaches to the bull
gear/outer spline, passes through the whole cone pulley assembly and allows
the drawbar to pass inside it.
Or, attach a tube to the top of the spindle, with a flat so the drawbar
clamps it in place. The problem is if the drawbar is loosened, the tube
will not be centered. This would slide up and down with the quill, so
that's another problem.
My solution works great for the 1J without the mickey-mouse mechanical
issues, and doesn't interfere with the drawbar at all.
I thought it was a cool way to solve the problem, so that's why I published
it on the web.
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.