Duane's CNC Lathe Conversion

I've had a CNC retrofitted Sherline vertical mill for few years, and I really wanted to be able to turn some parts on a lathe.  I first tried using the mill as a vertical lathe, chucking the material to be turned in a spindle-mounted chuck, and making some cutting tools to cut from the side.  It worked fairly well in easy to cut materials, and having the extra Y axis allowed putting multiple tools on the bed for automatic tool switching.  I was never fully satisfied and wouldn't have dreamed of trying to turn steel or steel alloys.

I hunted around and found one of those mini lathes for sale.  They're good little pieces of equipment, made in China, but this one was retrofitted with real English feed screws.  It is a MicroLux True-Inch Mini Lathe from www.micromark.com.  The regular import mini lathes have drive screws that are one-millimeter pitch thread.  One millimeter is exactly 0.03937 inches, and so they put a 40-division knob on it and pretend that each division is 0.001 inch.  That's fine if you want to cut a part that is 0.984 inches long instead of 1.000 inches long.  The MicroLux unit has a real 1/20th-inch pitch threaded drive screws and 50-division knobs on it.  This lathe is also a 7x14 lathe with a fairly long bed for holding chucks and drills in the tailstock.  I obtained the lathe and used it for about 6 months with some accessories shipped to get the feel of how stiff it was and what it could and could not do.

Once I finished my planning and obtained the highest torque stepper motors in a NEMA 23-size frame I could find (270 oz-in), I set about stripping off everything I didn't need and retrofitting the X and Z axis for computer numeric controlled operation.  Here is a picture of the final converted lathe.



I built the ten-drawer cabinet specifically for housing the controls and circulating oil system.  The ten drawers provides plenty of storage space for tools, cutters, fixtures, small pieces of raw stock, and spare parts.  The cabinet is on casters and can be easily rolled out from its designated holding spot in the shop.

The plan was to eliminate all the built-in screw-cutting capability, remove the entire apron assemble normally used for quickly slewing the carriage left and right, and provide a controller to allow manually positioning and running the lathe.  Here is the converted X axis.



The brass block attaches to the carriage using the original mounting screws.  The two aluminum interposer blocks house a bearing and a connecting coupler that attach the motor to the modified X-axis drivescrew.  Note that the unmodified compound slide is still in place.  I didn't need it for its manual capability, but it holds the toolpost.  I tightened down the dovetail gibs so it wouldn't move.  Notice that there is no backside shaft on the motor.  (So how do I preposition it before running parts?  Do I HAVE to have it hooked to a PC?  You'll see in a minute.)

I had originally envisioned using the 16-threads-per-inch (TPI) drivescrew that came with the lathe to drive the Z axis.  However, it was not a standard thread, and the half nuts that it engaged were of cast iron and were subject to a lot of wear.  I removed the old drive screw, and used a new piece of standard National Coarse 3/8-16 threaded rod for the Z axis drive screw.  As shown in the above picture, the rod drives a brass traveler mounted where the apron and hand wheel used to be.  On each side of the brass traveler is a plastic two-thread wiper.  These two plastic pieces engage a short portion of the thread as it enters the traveler, and remove excess oil and swarf (chips) from the screw before it goes into the brass piece, thus eliminating the potentially damaging wear of steel chips against brass.  The wipers are designed as near mirror images of each other.  Looking at the above picture, envision the drive screw rotating clockwise and advancing into the traveler.  The debris is wiped off on the BOTTOM edge of the wiper, and drips off into the oil pan below.  On the other side, the wiper is located on the opposite side of the screw, and wipes off and drips off as the screw advances into the other side (withdraws from this shown side).  Thus the debris is actively removed and drips off by itself.  Here is a picture of the Z axis and the manual controller, called Knobby.



The same type of 270 ounce-inch motor is used.  The main mounting plate and two other aluminum pieces house two bearings to allow thrust in both directions.  When the motor shaft is pushed or pulled, it tends to move slightly into and out of the motor housing, rendering them less usable for moving a cutter through metal.  That is why I put bearings into the assemblies.  A coupler and pretensioning coupler are also enclosed in the aluminum assembly.  The Z-axis motor mount attaches to the end of the bed through three tapped 1/4-28 holes.

The Knobby controller was designed to allow controlling the two motors with familiar digital-inch readouts.  The knobs are spring-loaded center-return potentiometers on analog-to-digital converters.  A microcontroller inside reads the ADC values, translates their values into speeds and directions using a parabolic speed curve, steps the motors at the appropriate rates, keeps track of the displacements, and shows the position in real inches (EG  1.234) on a 5 digit by 2 line LED display.  Each axis has two separate "odometers" or trip meters for doing master and incremental positioning operations.  A "cutoff" button on the right side advances the X axis into the workpiece at a steady, user-programmable speed for smooth easy cutoff operations.  A Motors On/Off button toggles the enabling of the motor phases.   The eight motor-control phases come out of a DB9 connector in the back and goes to a two-ported motor driver power supply.  You can completely and accurately control each axis manually with Knobby right down to the fraction of a milliinch.  Since I spring loaded the potentiometers for center return, as soon as you release the knobs, the motion stops (there is a small amount of dead-zone centered around the springy center position of the knobs).  And since a parabolic (squared) curve is used to translate the linear knob position into a speed, you can rapidly move the carriage at the extremes of the knob positions, and slowly, 0.0005 inch per second, move the carriage into final accurate position.



These motors are current-hungry, taking 4 amps per phase.  I drive two adjacent phases at a time in unipolar style, giving 190 oz-inches of torque 270/sqroot(2), which turns out to be plenty of torque for both axes.  The motor driver box, shown in the first photo to the right of the oil tank, houses two independent 8-Amp power supplies and phase driver circuitry.  The digital phase control comes into two parallel-connected ports, and the high current phase signals drive the motors.  I provided two input ports so that I can connect Knobby to one, and the personal computer (PC) to the other, without having to swap them.  Only one agent may control the driver at the time, hence the Motors On/Off button on Knobby.  Sometimes I forget to turn off Knobby control before starting a program on the PC.  Fortunately, when two agents are both trying to run the motors, the motors simply don't turn, and you abort the CNC program, disable  Knobby, and restart the program without losing any positional preset.

Turning steels and stainless steels requires coolant.  The tank holds 2 gallons of oil.  A ten-dollar, 12-Volt, submersible marine bilge pump pumps the oil from the tank, through a control valve, into the Loc-Line nozzle, and over the cutting edge.  The oil drips into the pan under the lathe.  I drilled a patch of draining holes into the pan to prevent larger chips and parts from falling into the tank.  A telescoping drain tube brazed to the drain pan bottom guides the oil back into a hole in the top of the tank, completing the cycle.  The drain tube was made telescopic so that I can collapse the tube upwards to allow removing the tank if necessary.  A box with a power switch and 12-Volt power supply feeds the bilge pump on demand.  I have not yet put a relay inside to allow CNC coolant-on coolant-off control yet.  Maybe someday.  I also want to change the oil nozzle mounting so it travels with the carriage so that the oil is always pointing right at the cutting edge.

I replaced the stock toolpost, a four-point turret type of toolpost, with a new Phase-II Micro quick change tool post.  The Phase-II  Micro tool holders, although adjustable, did not have any way of locking the adjustment screw into place.  I quickly made some flanged height-adjusting nuts with jam nuts to facilitate locking my height adjustment for each tool holder.


To run the lathe from the PC, I modified the CAM software I'd written for the milling machine, allowing for up to nine different tool offsets, stripping out the Y axis control, and adding a few more lathe oriented features.

I'm pleased with the performance of the retrofitted lathe, and it can consistently produce parts.  It's really nice to be able to do tapers, spherical arcs, and just about any type of turning not easily performed on a strictly manual lathe.  I did not tap into the spindle speed encoder to allow screw cutting.  The best Z-axis speed I can get is about nine inches-per-minute, which would require a quite low RPM on the spindle for turning screws.  I've decided that for the low quantity of parts I produce, and they're mostly for my own use, I'll run a tapping die over any part that needs a screw lead.  The digital Knobby manual controller allows accurate manual turning without the use of the PC, and moving the cutter and carriage around for presetting the cutter starting position for a CNC program.

Back to Homepage
Send me an email at snowleop@sover.net