Rigid tapping and thread milling on a PM25 LinuxCNC mill


Hi everyone. I’m gonna describe the
rigid tapping setup I’ve made for the PM 25 CNC conversion. Rigid tapping holds the
tap securely in a tool holder like an end mill would be and if you compare
that to a tapping compression head it’s easy to see why rigid tapping gives
better control over depth especially as the tool begins to wear. And depth
control is especially important if tapping into a blind hole. To perform
rigid tapping there needs to be synchronization between the tool height
and the angular position of the tool and that’s measured by a rotary encoder on
the spindle. In this case, I attached the rotary encoder to the drawbar. The rotary
encoder is attached to a differential receiver which is attached to a
Pico Systems Universal Stepper Controller which LinuxCNC uses to
coordinate all of this. The other thing that’s needed is sufficient reserve
capacity in this spindle motor and the axis motor. Ideally, during tapping the
spindle would maintain velocity but any change in velocity has to be met
with acceleration of the axes motor. The rest of the video is the construction of
this assembly, testing using a spiral point and forming tap, some
examples of problems that occur with the system if it’s run too fast and then a
comparison to thread mill. Now, rigid tapping is just another tapping tool and
how I plan to use this tool is when I have parts that have a lot of small
threaded holes that I can tread with a forming tap, because it is kind of
inconvenient to take this assembly off to change tools, otherwise, especially
with external threads, I’ll use single point treadmills, and with popular
threads like m6-1 which is the most popular thread
that I use, that I make, is to use a multi-point thread mill. If you have
any questions let me know otherwise let’s get started.
I previously made this drawbar and I’m taking this chance to modify the hex to
be compatible with the 17 millimeter wrench that I use for my clamping kit.
The hex is then turned down to receive a flex coupler which will attach the
drawbar to the rotary encoder. This isn’t terribly convenient especially when
wanting to change tools but it gives a nice easy way to demonstrate rigid
tapping. I considered also using a magnetic base encoder on the top of the spindle
motor and also using a timing belt to move the rotary encoder to the side of
the head but I’m planning to change a spindle motor and until that’s done I
don’t want to do anything more complicated. The rotary encoder gives
position feedback to LinuxCNC which I use to calculate velocity and position of the C
axis. This might be useful for operations like broaching where you’d want to know
the angular position of the spindle precisely but you would have to come up with a
means to lock the spindle at that position. Rigid tapping feels pretty sketchy the
first few times and before recording this threading test in aluminum I did
several threading tests using wood. Takes a little bit of a leap of faith I guess.
I think this is so cool that even on do-it-yourself CNC machines that two
axes can be synchronized and controlled so precisely that rigid tapping is
possible. From the side you can see how far the spiral point tap needs to come
down in order to fully tap this through-hole. Besides a spiral point tap
we’ll also try a forming tap. The forming taps require a different size drill bit
and they deform the metal instead of cutting the metal. More of the tap is also
available for threading as well. To compensate for the threads that are not
usable you have to change the breakthrough amount in Fusion 360 and
that can be parameterized by the thread pitch. The threads created by a forming tap
should be stronger than those created by a cutting tap because the forming tap
deforms the metal to create the threads. Also with a forming tap there are no
chips created so tapping to a bottom of a blind hole is more practical.
Unfortunately a forming tap requires two to three times a torque than a spiral
point or spiral flute tap does and in a high speed belt setting the PM25 just
does not have sufficient torque at the low end speeds. Since the height of the
head is synchronized to the angular position of the tool in the event of a
spindle stall the head will stop moving and they’ll be minimal or no damage to
the part but unfortunately a recovery is quite a pain. Peck tapping may not be available in
LinuxCNC and if generating that code in Fusion 360 there are no warnings
shown unless the NC editor is enabled. Other than demanding too much torque
during tapping the other type of problem that can come up is if tapping too fast
there’s considerable overshoot because the head has to remain synchronized with the
spindle motor as it free spins to slowdown. Also, with a 2,500 pulse per
revolution encoder, in quadrature mode, at 800 RPM, you need a very fast sampling rate.
So if the spindle is run too fast, even the high megahertz clock of the Pico
Systems controller board isn’t able to fully sample the rotary encoder. 400 RPM
seems to work pretty well, especially for fine pitch threads. Another gotcha is
since the final upward motion after leaving the hole is uncoordinated, it
isn’t possible to re-thread a hole. So you only get one chance to thread a hole
and if you try to go back and thread it again it’s going to destroy the thread.
We can also a thread with thread mills. Here are two single point and one
multi-point thread mill. Thread mills use coordinated motion of the
XYZ axes and don’t require a rotary encoder on the spindle. To use a thread
mill, a hole is first created to the minor diameter of the thread. Then some
thread relief is created with a chamfer tool to help the head of the screw set
properly. In Fusion 360 I use a 2d contour to create that chamfer and this
small Python script I wrote to calculate the chamfer width.
And this is what the treadmill looks like in Fusion 360. Probably every time an expensive
treadmill drops down into a hole is stressful but these tools are so cool. This hole didn’t get threaded quite all
the way through because Fusion 360 is looking at the tool length and bringing
the bottom of that tool to the bottom of the hole. For this through-hole, we need
to add a slight amount of breakout and that can be parameterized by the tool
pitch. A huge advantage of thread mills is that
a single thread mill can cut a variety of internal and external threads of
different diameters, and if using single point thread mills a variety of thread
pitches. But the interpolation required for thread milling is going to always
take more time than using rigid tapping with a specialized tool.
I forgot the chamfer on these holes but I hope otherwise this demo was clear.
Okay, recapping we looked at adding a spindle encoder to a CNC conversion of
a PM25 milling machine, some successes and some failures of performing
rigid tapping using the stock motor, and then compared rigid tapping to thread
milling. I hope you find this as interesting as I do. Coordinated motion
between multiple axes is really interesting subject. If you have any
suggestions or questions please let me know otherwise thank you for watching
and I’ll see you in the next video.

Leave a Reply

Your email address will not be published. Required fields are marked *