design required end-facing a 5 ft. long 6.625" OD tube. But how is this
done on a 19" lathe? The tail stock was removed and a temporary "tail
stock" was made from wood, a 1/2" steel rod, and mounted on the work bench.
Two casters are needed to support the tube at the chuck end because
there is not enough clearance to move the carriage under the tube. An
"end buttress" (also used in the final assembly) is clamped on a
long threaded rod and is used to drive the tube from the chuck.|
The lathe is set to its lowest speed, and only slow feeds are possible with this arrangement.
|Axial and radial alignment must be carefully done for this arrangement to work. The two casters and the "tail stock" must hold the tube at the correct height and angle. Paper shims can be used to raise the height of the caster plate if it is off slightly. The lathe's hollow chuck can be used as a "bore sight" to get the "tail stock" close to its correct position, and correct angle. If anything is out of alignment, the buttress cylinders will emit creaking noises and will gradually work their way out of the tube. Torque requirements will also increase.|
|A special fly cutter had to be made to cut an opening
the 11" serving bowl bottom. The hub was made from a length of 2.5"
steel bar stock, and the transverse tool holder was made
from 3/4" steel bar stock. The bit holder
itself was made by drilling an 11/32" hole in the bar and then pressing
a 1/4" lathe cutting tool bit through the hole. Three 1/4-20
screws anchor the bar and the tool bit.
A wooden plate is bolted to the drill press table. A central hole is drilled in the plate to facilitate alignment when the table has to be moved up or down (it can swing sideways). A circular groove was cut to anchor the edge of the serving bowl.
|Another 3/4" wooden plate (MDF) has a large circular
hole cut into it to firmly hold the bowl against the bottom
Here the 11" metal serving bowl is clamped in the fixture and a circular section (6.75" ID) has been cut out of the bowl.
This scheme requires rigid fixturing, a sharp cutting bit, the lowest drill press speed, slack drive belts, and a tool bit setup to lean slightly backwards so that it drags while cutting. This prevents the tool bit from digging into the thin, easily deformed metal, but results in a noisy, chattering, cutting operation. The hole is still a lot better than what could be produced by hand. It is finished by filing and sanding.
caps couple the curve of the bowl to the flat end-buttress. The
bowl is 11" in diameter (rim to rim) but the bowl curve itself is
actually about 11.5". The depth of each cut was calculated on a
drawing and then the carriage feed dial was used to make each cut to
the correct depth.|
Both the end caps and the end buttresses were made from three plies of 3/4" thick Medium Density Fiber board (MDF) . The plies were glued together and then carefully drilled in a drill press with a 15/32" drill to create a slightly undersized perpendicular center hole. The glued disks are mounted on a 1/2" threaded rod, and then the surfaces are end-faced in the lathe so that they are both perpendicular to the rod. This precedes any other machining operations
The lathe tail stock, not shown, is used in this operation.
plateaus were evened out with the compound feed and then final
smoothing was done with a coarse file. The profile is matched to that
on a drawing mounted on a piece of cardboard.|
The nut, washers, and tail stock must be removed to cut a flat for the flat portion of the bowl, which protrudes inward about 1/16".
During final assembly the end-cap is mated with the buttress by a 1/2" wooden dowel.
end bolster and curved end cap are joined together by a 1/2" wooden
dowel, and plenty of acrylic spray (these were later made as one
A hole is drilled in the dowel for a coupling nut which is then epoxied into the hole. A 10-24 screw is used to attach the spherical electrode. Note that there are no electrical connections to the center screw.
A shallow strip of MDF is chisled out to accommodate a brass electrode.
|The brass electrode strip is soldered to a multistrand wire and then glued to the endcap.|
is what the inside of the buttress+endcap looks like. The cap holds the
4" ID tube and the 6" ID tube in alignment and serves as a mounting
surface for the spherical electrodes. The black nylon 3/8" rod goes
through transverse diametric holes and anchors the cap to both tubes
and resists longitudinal slippage.|
The wire is terminated with an ordinary spade connector.
|This is a side view of the buttress+endcap. The ends of the brass electrode are bent up slightly to insure contact with the spherical electrodes..|
|Strips of aluminum (10" x 1" x 0.032" ) on one hemisphere serve to mate the two hemispheres into one sphere. The strips have 1/8" holes drilled 3/8" apart and 5/8" from the top edge to speed the setting of the silicone I sealant. The stainless steel surface must be roughened with 60 grit sandpaper, cleaned with household scouring powder, rinsed with tap water, and then with distilled water. The strip is likewise cleaned, then preformed to the curve of the bowl, coated with silicone, and clamped into place with document clips.||The two hemispheres are pushed together to make a sphere with a cutout that allows it to slide onto the 6" PVC pipe. They can be permanently glued together or held together securely with packaging tape spanning the seam on the inside.|
|The coil assembly is mounted on two vertical 4" ID PVC tubes, which are in turn mounted on a plywood base plate with casters. Base studs go inside the tubes to hold them in a vertical position. They are made from 5 plies of 3/4" MDF (Medium Density Fiber board).||Bolt holes have to be bored into the studs so they can be mounted on the plywood base. Bolts are made from 1/2" x 7.5" threaded rod.|
fly cutter (homemade) is used to cut a contour on the end of a 4.5" OD
vertical PVC tube to mate smoothly with the 6.625" OD horizontal
tube. A ring previously trimmed from the 6.625 OD tube is used to check the profile. |
Note how the far end of the tube is held up by a loop of parachute cord, and the near end is held in an undersized vise by hose clamps.
is a picture of a vertical support tube and one of the end spheres.. The tubes slip over the base
studs which are bolted to the plywood base. The horizontal coil housing
rests on top of the two tubes.|
The hole at the top of the vertical tube is for a parachute cord tie-down of the horizontal tube which contains the secondary winding).
For storage, the spheres are removed. Everything else easily disassembles.
|About 2000 feet of #24 AWG copper magnet wire has to be wound on the 4.5" OD tube. Doing this entirely by hand would be tedious and time consuming. Hence, an arrangement was made to drive the tube from a lathe chuck by using a slip clutch made from a spring and some washers. The lathe is run with reverse rotation and at its lowest speed.||Turns
are kept adjacent and tight with fingernails and finger pressure. The
wire is tensioned by letting it slide between folds of paper towel that are pinched together by the other hand (not shown).
After about 4 inches of turns are laid, the lathe is stopped and the
turns are sprayed with clear acrylic lacquer to anchor them in place. |
The coil winding took about a half-hour and was by far the easiest of all the fabrication operations .
|This is the finished secondary coil. Its 52 ohm resistance is equivalent to about 2000 feet of wire.|
Two MDF centering rings protect the coil when it is slid into the outer 6" tube.
|finished primary under test|
|This is a 'scope trace showing the self-resonant frequency of the secondary
with both end spheres attached. The primay consisted of 5 turns of #16
wire and was "pinged" with 1.6 volts from a small battery. The result
captured on the single sweep display was:|
Freq = 122kHZ
Vpp = 48 volts
The turns ratio in this test is therefore about 48*5/1.6 or 1:150
Text book calculations for inductance, capacitance, and resonance were:
L = 12.6*(10^-7)*(1)*(2000^2)*(0.003806)/(0.8636) (MS Word field formula)
C for two partial 14" spheres = 20 picofarads (est.)
The distributed capacitance of the coil was not included.
freq = (10^6)/(6.28*(22000*20)^0.5)