Why build it when you can buy it?

Because.
1: There are many good, popular machines on the market. In order for the companies who make them to stay in business, and pay their employees, their G&A, and their overhead, and maybe even make a profit, a good portion of the $2,000-$3,000 price simply cannot be shipped with the product.

2: If one builds it, one must face the Customer every day. Therefore, the motivation to do a good job is high.

3: The homebuilder chooses the components, and lives with them. Some cost-cutting accountant is not dictating the use of nylon bearings.

Why Not.

1: You need a good metal lathe (Or a friendly machinist who REALLY likes jewelry.)

2: You must have some mechanical ability, but since you are already cutting stones, well..

3: Time and effort. It is much easier to write a check....if you will settle for someone else setting the rules in your shop.

Essentially a lapwheel with a dividing head which indexes the stone around a central axis to provide symmetrical placement of the facets. The head is adjustable for angle as well, since different stones require various angles due to differences in refractive index. This machine is built on 1" thick 6061-T6 aluminum tooling plate, The permanent magnet DC motor is reversible and the speed is controllable from 0-2,000 RPM with constant torque. The mast is 1" diameter type 316 Stainless Steel, and employs a linear ball bushing , and all needle and ball-bearing construction to allow Zero Lash. Fine height adjustment is accomplished with a micrometer spindle. The dop chuck accepts 1/2" dops, because they can be readily made from aluminum screw machine stock, and can easily be machined to modify them for any size or shape stone. Why 1/2" Dops? You can bore the 1/2" quill (index) shaft 0.2505 X 3/4" deep and drill and tap a setscrew if you want to use standard 1/4" dops. However, consider these. They will not bend, flex, or vibrate, and are much easier to line up. The larger radius makes a big difference in easy locking into the quill without using pressures on the setscrew that will deflect them from center or score them. I use a 1/2" collet in the lathe, and make them from 6061-T6 aluminum. A lifetime supply can be made in an hour. I ground a 42 degree countersink to make the conical ones. Make them from brass, if you'd rather. Or stainless. Or titanium. Or whatever, as long as the diameter is accurate and is a sliding fit into the quill coupling. Naturally, if they are undersized, the setscrew will jack them off-center to the opposing side of the coupling bore, and this is a serious fault. You cannot cut a perfectly centered stone if the dop is acting like a cam, and transferring the stones will be inaccurate.

Absolute Requirements:

1: A source of industrial supplies such as ball and needle bearings, motors, controllers, hardware, etc., such as McMaster-Carr, Grainger's, etc., and a source of some precision components, such as PIC design and Berg.

2: A metal lathe whose spindle and chuck or collets having runout of 0.0005" T.I.R. or less.

Access to a milling machine whose table has been indicated to 0.001"/Foot or less.

A bandsaw with an appropriate blade for 1" thick aluminum.

3: An operator for the above.

4: Rudimentary electrical skills.

5: Some money, checkbook, or other barterable medium of exchange.

6: About 40-100 hours.

1: Use sealed (not sheilded) ball bearings. What happens to a precision bearing when rock dust and diamond powder gets into them has to be seen to be believed.

2: Laps must be at least statically balanced. No runout or "hop" can be tolerated, regardless of effort, because gemcutting is not an impact process.

3: Sleeve bearings/bushings of Nylon, porous bronze, etc. should be avoided. When abrasives get into them (not if), they make good laps, and proceed to ruin the shaft. (The expensive part.) They are cheap, but otherwise, there is no excuse for using them.

4: Avoid V-Belt drives. There are hysteresis losses, and they are unbalanced. If spatial constraints demand a belt drive and arbor, use timing belts. I use direct drives because motor ball bearings are usually sealed Conrad Type R bearings, will support a thrust load, and can be quickly and cheaply replaced.

5: House Rule here :-) If it can be heard running ten feet away, fix it.

6: Be messy on the lap assembly if you must, as long as the spindle/arbor assembly runs perfectly true, and whose axis is perpendicular to the base to the limits of your best efforts at measurement.. but on the faceting head, any runout or error will be transferred to every stone you ever cut. Zero runout or slop means "Zero", or the resolution limit of the best dial indicator or other device you can buy, beg or borrow. (Includes laser interferometers. "Zero" means Zero. If you make the attempt, you will have a better machine than can be bought.)

7: Motors used are DC, either Permanent Magnet or wound field types, having a shaft diameter of at least 1/2" (12.7 mm), and using sealed ball bearings, rated at least 1/5 HP (~150 Watts).

---

Introduction:

This is not a Beginners' Project. By its very end use, the machine is intended to produce faceted gemstones of high precision, and it cannot produce something more accurate than itself. Familiarity with mechanical principles is needed to complete this project. A proficiency with mechanical assembly practices and the use of common shop tools is a must.

For those unfamiliar with the operation of machine tools such as lathes and milling machines, it is best to get help from a machinist, toolmaker, or other skilled craftsperson. Often, a local trade school can help greatly..sometimes they are grateful for a "special" challenging project. Of course, you will have to pay for materials and supplies at the least. Another possibility is to enroll in a basic machine shop/ metalworking course. The skill will not be wasted (No skill ever is..), as later on, you can make your own laps and other tooling, saving many times the cost of the course.

Design Principles: The baseplate can be aluminum tooling plate ("Jig Plate"), 5/8" thick, or Plexiglass, 3/4" thick. If the latter is chosen, it should be drilled and tapped so extruded aluminum channel can be bolted to the back surface as stiffeners. In either case, the materials must end up being better than 0.005" flat per foot. Plexiglass is very easy to work with common woodworking tools, so many would prefer it. It is cast to sufficient flatness so it can be used. It does scratch easily, of course. Both materials are illustrated later on.

A DC motor with sufficient torque is used so it does not slow down under load. A solid-state controller with feedback may be used for controlling the motor speed. However, faceting does not put great demands on the motor, so a simpler speed control, such as a Variac, a rectifier, and a capacitor may be used with good results. Very inexpensive controllers are available based on Triac or SCR switching, like a lamp dimmer. The output of these devices may be wired to the appropriate transformer for the motor's voltage, rectified, and connected to the motor. For example, a 24 Volt permanent magnet motor can be run very well by buying a 24 volt transformer of sufficient current rating, connecting the output of a "dimmer" to the input, and connecting the output to a bridge rectifier and a capacitor,(Rated 30-50 VDC, ~10,000 MFd.) through a reversing switch, to the motor.

Ball-bearings are used everywhere, of the sealed type. Should any diamond dust or grit get into them, they can be changed quickly and cheaply. The use of ball bearings eliminates most "slop", and the accuracy of the unit does not deteriorate over the years. Also, all the close tolerance work is done by the bearing manufacturer...all for less than ten dollars.

The faceting head and quill assembly is all ball bearing. The bearings used in the trunnion assembly should last forever, as they are not exposed to spray. The gears are off-the-shelf precision stainless steel, and are assembled to an accurately machined shaft adapter, so they can be changed in seconds. Because the gear is silver brazed to the adapter, there will never be any indexing errors caused by slippage.

The mast assembly consists on a stainless steel centerless ground 1" rod, chilled into an aluminum baseplate. A Thompson or Barden ball bushing with Zero "slop" is used on the head assembly for smooth and error-free adjustment and travel. The depth is set by a micrometer spindle which rests against a coarse adjustable aluminum clamp, so a wide range of depths is possible whether cutting a table or girdle. The Base is straightforeward, and will be built first. If you decide to stop when you are finished with the base, you can always build the faceting head and mast assembly later on if you want, and buy Wykoff's "Calibrated Jam Peg", and with skill and practice, probably turn out better stones than I do, and still be far ahead in cost.

Materials Needed: 1 Sheet 3/4" plexiglass, or 1 sheet 5/8" aluminum Jig Plate 18" X 14", or whatever dimensions you might prefer. These are a judgement call by the builder. Some might want to build it into and existing cabinet or countertop, others may wish to build a base for it so it will be "portable", and will sit on a counter.

Construction details

Revision "H"!...In Stainless Steel.

Motors used are DC, either Permanent Magnet or wound field types, having a shaft diameter of at least 1/2" (12.7 mm), and using sealed ball bearings, rated at least 1/5 HP (~150 Watts).

Here is the c-face motor mounted to the underside of the acrylic base. Note the bolt pattern, and the machined splash and dust guard (Made from an old wire spool). The face of the motor was machined dead perpendicular to the shaft, to remove errors caused by the manufacturer's paint. To do so, the motor shaft was mounted to a live center on the lathe's tailstock, and the motor was chucked, after its wires were carefully taped to the body of the motor. This step may NOT be neccesary in most cases. I just did not like the looks of the paint application, and I like to run the lathe, anyway. A "bump" of paint will affect the perpendicularity, just as a shim would.

Here is a simple fixture, accurately bored, which mounts on the spindle, and measures wobble relative to the top surface. Since the faceting mast assembly mounts to this surface, and the motor is mounted to the opposite side, this is the datum plane from which measurements are derived, and the plane which is the angular reference for every stone which will ever be cut. The parallelism specs for cast acrylic sheet are very tightly controlled.

The faceting head assembly, plopped onto the base. It is secured with a single 3/8-24 bolt and clamping nut and washer, so it can slide in and out towards or away from the centerline of the spindle. You will find that if the perpendicularity is accurate (See section on the mast assembly below), it can also be swung through an arc, allowing unlimited positions for operator comfort.

---

Mast assembly

The adjustable Thompson ball bushing in the faceting head assembly rides on this massive 1" stainless rod. By using an adjustable bearing, the tracking of the faceting head assembly can be trimmed perfectly, and the bearing adjusted to Zero Lash. This allows perfectly repeatable settings. Suffice to say, the mast does not bend, vibrate, or deflect in use!

The bore for the 1" mast MUST be perpendicular! Bore it on a lathe, milling machine, or a really accurate drill press which has been dial indicated. Otherwise, different height settings will affect angle settings, and it will drive you crazy if you are trying to cut a perfect stone...and don't we all? Try to keep the tolerance in seconds of arc. It's not easy, but it's worth the effort. You are only doing this once, right?

Faceting Head.

This is the challenging part. This section of the document will be growing over the months, as more sections appear in the Eclectic Lapidary. Warning: Graphics intensive. If you have a slow server or connection, go get a cup of coffee or something.

A Word About Gears.

This design uses gears with a Pitch Diameter of 2". You will notice in some of the preceeding photographs that I do not number the index gear teeth, except by a binary pattern of dots. I think it's faster and more visual, but of course, you can number the teeth if you feel you need to. For translating cutting data for those who need it, I use this

Worksheet and design form.

Variations between gears of different pitches needed to supply different numbers of teeth for a given design size are accomodated by the spring-loaded indexing pawl. The "V" design of the indexing tooth accomodates different pitches (Tooth sizes). Here are some examples:

96 tooth, 48 pitch. PD 2.000, OD 2.041

64 tooth, 32 pitch, PD 2.000, OD 2.062

48 tooth, 24 pitch, PD 2.000, OD 2.083

32 tooth, 16 pitch, PD 2.000, OD 2.125

128 tooth, 64 Pitch, PD 2.000, OD 2.031

These gears are all available from W. M. Berg, (516) 599-5010, and other sources. Start with the most commonly used number, the 96 tooth one, and you may as well get them in stainless steel. It's not that much more, and it won't rust. This design does expose the gear to cutting spray and mist. The gear is actually used as a splashguard to protect the spindle bearings. A sleeve coupling (Their stock number CT-6) in 303 Stainless Steel, is required for each gear desired. This is a precision sleeve coupling with a bore of 0.4998. Please..You CANNOT use the zinc-plated cheap hardware store versions of this. Your dops will be mounted in this. It is a critical part. The gear hub(s) are bored to 1" (Light press fit) to the OD of the above coupling, and then silver soldered in place. In use, the setscrew on the coupling can be loosened, and the gear/hub assembly changed instantly, should a different cut require a gear other than the 96t unit. Of course, because of the silver soldering, there will never be be slippage between the gear and the coupling.

The drawings shown in this document may differ from the photos of Cate's unit. That unit was an earlier revision, with an aluminum mast follower housing. These drawings and dimensions are for the latest "Revision H", which I built from Stainless Steel. It is larger and even more rigid. You CAN abbreviate the outside dimensions and scale it down if you wish, but from what I have seen, and heard from people with commercial machines, flimsiness and susceptibility to harmonic vibrations is undesirable.

This view is of the side usually away from the right-handed operator, showing how the dowel pin is installed. The pin is a rest upon which the angle stop rests. The angle stop rotates with the trunnion shaft and locks to accept any angle setting.

That's it for the big pieces. Now all that are needed are two pieces of 1/2" shafting with a little milling and turning, a fairly simple angle stop clamp, a stolen protractor, and a few little parts like the indexing pawl , a spring and stuff and then we can put it together and calibrate the protractor. When these parts are done, you can start worrying about buying rough.

Last Revised 12-01-97. Date of first publication: December 17, 1996: Single use copyright granted to Eclectic Lapidary Magazine for online article. IP licence granted to end user for the manufacture of one unit for personal use. This article was produced without compensation, to contribute to an enjoyable pastime.