Making a Custom Die on the Proxxon PD 250/e Lathe


Welcome to Adventures with a Very Small Lathe The carriage of my G. Boley watchmakers lathe
needs a lot of restoration work to get it back into working order. The lead screws have been badly damaged, and
need to be repaired by chasing them with a die. The thread geometry is very unusual, so commercial
dies aren’t available. The first step to making a die for a thread
this small is making a tap, which I did in a recent video. In this video I’m going to use that tap
to make a matching die. The most suitable standard die size has an
outer diameter of 13/16”, which is just over 20mm, so I chose this 20mm ground silver
steel bar to make it from. Most dies this size are ¼” thick, which
is a little over 6mm. To face the stock flat I used a parallel behind
the work to get it reasonably square. The stock is already about the right size,
and the thickness isn’t critical, so all I have to do is face the two sides roughly
parallel, and knock off the corners. This would have gone a lot better if I didn’t
have the tool way too high. With the tool height fixed it cleans up a
lot better. To face the other side I used the parallels
once again. This should be easily parallel enough for a simple thread die. After cleaning up the part is almost exactly
a quarter of an inch thick, which is mostly luck. I spotted the hole with a starter drill, as
it’s better suited to drilling small accurate holes than a centre drill. The thread is a very unusual geometry. 0.6mm
pitch by 4mm diameter, with a 55 degree angle and a left hand direction. For full details
of how I worked this out, watch the tap build video via the link at the top right now. The thread geometry requires a 3.1mm centre
hole, but my good quality drills are in coarse sizes. I used a good quality 2.5mm coated twist drill
for the first hole, then brought it to diameter with a very cheap drill of the right size.
The cheap drills break really easily if I use them from scratch on tool steel. Now to try and tap the thread. The shop made tap has never been used on steel
before, so I don’t really know how well it’ll work. It seems to cut ok, but it’s not easy going. I backed it out a couple of turns to break
the chips, and added more cutting oil. The tap felt tighter as the thread progressed, and after a few turns I could feel something slipping. This turned out to be for two reasons. The first issue was that the shaft of the tap was soft, and started to deform in the tap wrench. The second issue was the chuck wasn’t quite
tight enough, and the part started slipping. The second issue was easily fixed by tightening the chuck, but the tap wrench was clearly not working. I hadn’t bothered to harden anywhere other
than the cutting parts, so the tap wrench had mashed the square shaft out of shape.
In hindsight it would have been desirable to harden the whole tool to make it more wear
resistant. It was also a bad idea to drill a centre hole in this end of the tap, which has made it much weaker. The square was too damaged for the tap wrench
to work, so I switched to using an adjustable wrench which could clamp with two flat surfaces
over a wider area. Before pushing any further I wanted to check
in detail how the tool was doing, so I backed the tap out, and cleaned everything down for
inspection. Looking at the die, the thread appears to
be forming correctly, and there’s no sign of cross-threading. The tap itself looks okay at this point. All
the cutting points look in good condition. You can clearly see the messy job I made of
grinding the relief though. I kept working at it, clearing chips regularly. Using the tap follower makes it easier to
apply torque to the tap without worrying about pulling it away to one side, and damaging
the straightness of the thread. Off camera I switched to a toolmakers clamp,
which allowed me to get a really firm grip on the tap shaft. Eventually I managed to finish the thread,
and from a casual inspection it looks reasonably well formed. It’ll be more obvious how well
formed they are once the cutting edges have been machined. The lathe work for this tool is now complete,
and the next step is to cut holes around the die to form the cutting edges. Commercially made dies around this size typically
have three round holes in them. They basically define the cutting edges, but their shape
and size is also important for allowing split dies to be opened up, adjusting the depth
of cut. This M4 die has three 6mm holes in, with their
centres 4.25mm from the dead centre of the die. I based my own design on the same dimensions,
and modelled it up in OnShape to make sure the geometry looked sane. Check the link in
the description for access to my OnShape model. The holes needed to be precisely located relative
to the centre, and I planned to locate them around the centre using the Proxxon simple
dividing head. The dividing head table is exactly the same as the lathe spindle, allowing
me to move the chuck from the lathe to the mill with the part still held in the jaws. To set this up I borrowed a trick from Joe
Pie. Before I started the project, while the chuck
was still empty, I used it to set up everything I needed on the mill. To locate everything correctly, I first needed
to align the axis of rotation of the chuck on the dividing head with the mill spindle. I put a broken end mill shank into the spindle,
and tightened the collet firmly. I clamped the chuck jaws around the same shank,
with the dividing head free to move on the table. This meant the chuck was exactly centred
on the axis of the mill spindle. I fixed the dividing attachment in place,
and locked the axis, so everything was set up in advance for later. Fast forward to where we were, and the chuck
can now be removed from the lathe, and fitted to the dividing head. The centre of the dividing head is already
directly inline with the axis, so it’ll be easy to move it to the correct offset. As we saw on the CAD model, the holes edges
are interrupted as they overlap with the threaded centre hole. Twist drills generally won’t cut straight
if the hole is interrupted, so for the final diameter I would need to use an end mill. End mills which plunge straight down are available,
but on my small wobbly mill they chatter a lot and cut very poorly near the centre. I needed to find another way of cutting straight
down and still deal with the interrupted cut at the edge. I used the dials to move the table so the
spindle axis was 4.2mm from the die centre, the correct distance for the location of a
hole. I then spotted the location with a starter
drill. Then I unlocked the dividing head, turned
it 120 degrees to the next hole location, and locked it firmly again. The spindle was now directly at the correct
location for the next hole, and I spotted the position again. I then rotated another 120 degrees for the
third hole. For the first pass drilling the holes I chose
the largest twist drill size that I could be certain would stay clear of the threads
in the centre. If a twist drill cut is interrupted on one side it’s very likely veer in the
direction of the interruption, as the cutting edges are less loaded on that side. This would
really screw up the hole. I chose a 3mm drill just to make sure I had
half a millimetre of safety margin. The twist drill starts very smoothly at the
spotted position because the starter drill point has an angle of 120 degrees to match
the point of my twist drill set. I used fairly gently drilling pressure, as
the material wasn’t fully supported underneath. Turning the dividing head 120 degrees guaranteed
that I would be back at the correct location for each hole. The drill cut pretty smoothly, and I wasn’t
sure if any lubrication was necessary, but it doesn’t seem to hurt, and should help
to keep the twist drill cool. I planned to use a 6mm end mill to bring the
holes to size, and the 3mm holes I already drilled made sure the centre of the bottom face
of the end mill didn’t need to do any cutting. This reduces a lot of the load on the end mill,
and also makes it more stable while it’s cutting. End mills are always held in a collet rather
than a drill chuck, as the collet is designed to keep the end mill rigid under side load,
which is exactly why I was using an end mill for this hole. A great tip I learned recently is to keep
the plastic protector on the end of the end mill while installing it on the machine. It
protects the fine points on the end mill in case I fumble or drop it, and it helps to
prevent me from stabbing myself in the hand if I’m too clumsy. The end mill sounded very noisy at first,
but I discovered it made a lot less noise if I kept a fairly consistent vertical feed
pressure on the cutter. The squealing seemed to be worse when the cutter was moving too
slowly, and is probably due to the cutting edges rubbing against the surface. Fundamentally this milling machine is just
not very rigid, so it’s very difficult to get rid of chatter entirely. It would probably have been better to use
compressed air to clear the chips, but I didn’t think of that until later. The surface finish inside the holes isn’t
perfect, but the cutting edges are the most important thing. They look well defined, and there are no obvious serious defects on the threads. Commercially made dies are often split, so
it’s possible to open them up to vary the depth of cut. I intended to split this die before using
it, so I can control how close it cuts to the target size, but I won’t do that until
after heat treat, in case it makes the die more susceptible to deforming during hardening.
Commercially made dies which are not yet split typically have a v-groove on the edge, aligned
with the side of a hole. This can be used to align the die correctly in a holder or
wrench, but it also marks the place where the split should be made. The die will be opened by tightening a screw
between the sides of this groove, so it needs to be a V shape. I placed the die in a vice, and lined it up
so the right position was at roughly 45 degrees. I have no reason to believe the angle of the
sides of the groove are in any way critical, so eyeballing it seems good enough. I cut the groove with the same 6mm end mill. I advanced the x-axis and the z-axis for each
pass, to keep the centre of the groove in the same place. Dies also need two detents cut either side
of the groove, for retaining screws that hold the die in place while cutting. To quickly mark the right location I coloured the outer surface with a sharpie, put the die in a holder,
and tightened the retaining screws to make a mark at the correct position. The detent only needs to be deep enough for
the tip of a retaining screw to align with, so I used the starter drill again. Before heat treating the die, I wrapped it
in a paste of boric acid and denatured alcohol. Boric acid is only slightly toxic, but it’s
important not to get it in your eyes, or breath in the dust. I had no guidelines on the proportions, so
I just added denatured alcohol until it was enough to bind the crystals into a paste. The paste behaves like a flux preventing the
surface of the steel from being exposed during hardening, and helps prevent the scale from
forming. To keep the paste in place, I used a mesh
of fine wire wool strands. Both boric acid and denatured alcohol are
nasty enough that I don’t want to touch them with my bare hands, so I wore nitrile
gloves. The most important area to protect is the
cutting edges, so I made sure there was plenty of paste in the centre. That turned out to be way too much wire wool,
so I cut most of it away. I pre-warmed the hearth while I was preparing
the die for hardening, to try and get more even heating. The burner uses MAP/Pro gas,
which burns in air a few hundred degrees hotter than Propane. For very small parts, the silver steel data
sheet recommends quenching in clean water, which has the advantage of being less messy
than oil. Some metalworkers recommend brine to reduce bubbles, but I’m going to wait
until I have something that is easier to re-make before experimenting too much. The wire wool strands burn away quite quickly,
but the boric acid crystals have already melted into a sticky layer which clings to the steel
pretty well. It seems to work just as well as making a cage from thicker iron wire. I
prefer not to use thick wire, as I’m concerned it will prevent the part from being quenched
properly. The copper wire looped through the die quite
predictably melted through almost immediately. I have no choice but to do the heat treatment
outside, which means on days like this the wind causes quite a few problems when gusts
divert the flame, and the part cools down again. Silver steel needs to soak for a short while
after reaching the right temperature, to make sure it’s heated through. With the wind gusting it was tricky to find
a moment when all parts of the die were heated to red fairly evenly, but eventually I got
the right moment, and quenched it. Tempering by eye requires a clean metal surface
so I stoned one side of the hardened part to a bright finish. This bright surface should
change to a straw yellow colour as the steel starts to temper, and for this tool that is
where I’ll remove the heat. Temper heat needs to be fairly gentle and
even to ensure the entire part reaches the right temperature at the same time, and a
traditional way to achieve this is to temper in a bed of brass chips. I rescued the chips from some brass drilling
and lathe work I’d done, and degreased them with acetone. I set the cup on these bricks so I could easily
control the level of heat. Fairly quickly it became obvious that the
acetone hadn’t degreased the chips enough, and the heat started to produce oily smoke. The smoke was thick enough to make it hard
to see the bright surface, and it looked as though it was also starting to discolour the
surface of the steel. I ended up having to guess the right moment,
and remove it from the heat. After the temper I went back to the stone
to remove the surface discolouration, and bring all faces to a clean bright finish. I cleaned up the groove, the detents, and the interior surfaces using a rubberized abrasive
bit in a dremel style tool. The final operation was to cut a slot to split
the die. Now that it’s hardened, conventional machine
cutters won’t work, so chose a fine cutoff wheel and fitted it to my Proxxon dremel-style
tool. I clamped the die to this door hinge, and
clamped the hinge to the end of the mill cross table, so I could advance it towards the cutoff
wheel while keeping it pretty straight. This isn’t the most professional fixture
in the world, but it worked. Thin cutoff discs are very brittle, so I advanced
the feed pretty slowly. There didn’t seem to be any problem with
the die heating up, probably because I was taking it so slow, and the fixture allowed
the heat to conduct away quite easily. Now that the die has been split, I can try
it out. This is the thread I need to repair. Under the microscope it’s fairly easy to
see how the thread is damaged. Another part pressing on the thread points has flattened
the tops, and mushroomed the steel out to the side. The end of the screw is relatively
undamaged, and shows the contrast. I fitted it to the tailstock die holder, as this was the easiest die holder to grip
in the way I needed to for this kind of work. I opened the die up as far as I could be reasonably certain was safe,
and ran it down a few mm of the lead screw. I’d opened the die to make a very shallow
cut for the first pass. Any mistake I’d made while making the die would only damage
the thread even more. Back under the microscope the improvement
is pretty clear. The point where I stopped is around here, and the thread is clearly
in better condition than it was. I’m going to spend plenty of time carefully
adjusting the die to get the best possible result, but that’s a story for another video. The carriage for the watchmakers lathe is
getting quite close to the point where I can reassemble it again, and I’m really looking
forward to making my first chips on the lathe. I have a few other projects already in production
which are going to come first, but there isn’t too long to wait now. My next video will be a race between some
work on the Chinese lathe, and a couple of small projects I have in the works, so check
back soon.

47 Comments

Add a Comment

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