The Ta Shing lathe has a centre height of about 5" (125mm) &
24" (600mm) between centres. It has a 2-speed primary belt drive,
3-speed secondary belt drive & backgearing - making 12-speeds in all,
covering 45rpm to 1336rpm. Fitted with a classic style of screwcutting
gearbox (40 Imperial threads). Included was a robust stand/cabinet.
Originally a 3-phase motor, now a 1-phase motor, re-wired to be capable of forward &
reverse. I'd guess it weighs about 350kg without the
|Ta Shing lathe during rebuilding.
(Visual distortion - caused by
wide angle lens on camera).
A screwcutting gearbox - what a luxury - auto feed on the saddle & the cross-slide - even more
old lathe, there's lots of work to do to bring it back to operational status -
at least it's now clean enough I can touch it without
using grease cleaner on my hands afterwards.
This will be a 2-stage
process: Firstly, I will bring it back to a level where it is
functional. Secondly, at a later time, it will be mechanically restored -
I have other projects I will complete in the meantime.
Problems - Repairs & Restoring to functionality:
|General clean - Done|
|Many of the washers installed were of the wrong types (or absent) - Done
- Each washer style has its own purpose. Used a mix of flat, spring,
& tooth washers, as appropriate.|
|Change the motor - Done -From 3-phase to 1-1/3HP single phase (motor came off an
industrial floor polisher, a framework was welded together, to convert it
from flange to foot mount).|
Make a new dual pulley for the motor - Done
|Wear on the carriage & bed - up to 6.5 thou on each - temporarily
packed with shim to compensate - Done -
Down the track I will come back to this, regrinding & refitting the
|Half-nuts stripped - I made a new threaded half-nut sleeve from a piece of
phosphor bronze - Done
|Worm to drive carriage feed & cross-slide has worn bearings - Done
- Bronze thrush washers/bushes made.|
|Idler used for change-gear drive, 2 teeth missing - Done
- See Gear
|Tailstock has 0.04 mm wear on front edge - Body rocks on its base - Tailstock 1.5 mm low - Done
|Cross-slide taper gib is too small (somebody ground the S*** out of it) - Done
- Shimmed at the back of the gib. All working surfaces blued &
scraped to fit.|
|Cross-slide gib adjustment screws (front & rear) missing - Done
- Replaced with large headed machine screws|
|Cross-slide bronze nut & the saddle in contact & scraping against
the body - Done
- 0.5mm removed from both sides of the bronze nut for clearance.|
|Compound slide bronze nut mounting has stripped thread - Done
- Thread re-cut & a slightly larger mounting screw used. Shimmed
at the back of the gib. All working surfaces blued & scraped to
|4-tool toolpost has 5 square head machine screws missing - the 3/8"
ones were easy to replace, but the one that was 7/16" had to be
fabricated. The 7/16", originally a 3/8"W had obviously been
stripped in the toolholder body and re-tapped for a larger size. I
used grade-8 bolts, machined to length, a square head was milled into the
hex head using a carbide router bit (a woodwork one, using high speed &
low feed - to prove it can be done with a cheap router bit). The square head was made to match the 3/8" bolt heads,
thus all needing the one wrench - indexing was done using the homemade 'Piston'
simple indexing head - Done
- See: Milling a square on a shaft for proportions
of the round blank size to final square size.|
Making a dual pulley from 2 single pulleys.
I needed a dual 'B' type pulley of 2-1/2" & 6" for the
motor. I found that here in Perth dual pulleys are not available 'off the
shelf' in anything but 'A' type.
I could have bought a large block of aluminium & machined one from
scratch, but my solution to this problem was to buy 2 pulleys of the correct
sizes & join them at the hub.
My idea: Machine one hub to fit inside the other hub (see
below). The spacing for 'B' pulleys is 0.750", the pulleys were
machined to push together & seat at this spacing.
When setting up for machining in the lathe, the pulleys must run true
both on the outer edge of the pulley & on its front face - it must be
axially true with no run-out (or wobble) or it won't run true when installed.
I elected to bore the shaft hole in the middle last. The pulleys came
with a 5/8" bore, which fitted an arbor I already had. Whereas, the
final hole size had to be machined to a non-standard size (neither metric nor
imperial) to fit the motor shaft.
|After measuring both hubs I decided on a 1.25" Dia joint.
This was roughly mid-diameter of the hubs between the keyway & the
outside surface , minimising them being weakened.
The grub screw was moved from the hub & re-installed into the
middle of the smaller pulley.
The arrows at the bottom show the location of lock screws - 'Scotch
||Another (simplified) sketch showing how it goes together, the 2 pulleys,
2 scotch keys, woodruff key, grubscrew & shaft.
||The 2 hubs after machining. The hub on the smaller pulley was made
2-thou larger than the hole in the larger pulley - hence one is pressed
into the other - an interference fit.
||Here they have been pressed together after coating the joint surfaces
While pressing them together, I made sure that the keyways lined up, by
slipping a small piece of keystock & a 5/8" arbour into the
||I then had to machine down one side of the larger pulley by a small
amount, allowing clearance for the drive belt not to chafe when on the
smaller pulley. Notice one side of the large pulley is thinner
compared to the other.
||To lock the pulleys I used 'Scotch keying'. In the two recesses in
the hub are #10 x 1" socket head machine screws. They're
positioned so that the central axis of each screw runs straight down the
joint between the 2 pulleys (see arrows in top sketch), well away from the
It was returned to the lathe & trued up in both axes. The
shaft hole was bored 2-thou undersize for an interference fit on
the motor shaft. The keyway was filed to it's correct depth
Repairing Stripped Half-Nuts:
The half-nuts were stripped, I'd guess because the engaging lever had lost
its detent & would not remain in the up (disengaged) position.
With the lever falling down of its own accord, the nuts must have been 'riding'
the leadscrew much of the time.
||Half-nuts seen together, as they would be when mounted behind the apron,
wrapped around the leadscrew. (Rotated 90-degrees though).
The 3/4" x 8 tpi ACME thread is so badly worn, it can no longer engage with the
||The half-nuts separated. One side is stripped bare of thread, the
other side has about 20th of thread remaining. Because the nut was
originally made slightly off centre, it has worn this way & does not
engage the leadscrew.
||With the apron removed, working on an assumption that the worn area was
central on the leadscrew, the first job was to super glue a piece of
3/4" OD pipe into the hole, this holds the nuts in position
relative to each other when removed. Next a bit of strong angle was
machined square. The half-nuts mounted to this with 4 small bolts.
Next, mounted on the faceplate, centred (using the pipe) and bored out to
1.000", ready for a replacement threaded sleeve. A small area
of the outside of the half-nuts was machined, for set-up purposes, in case I needed to
return it to the faceplate later. This formed a reference surface
for the dial gauge or dial test indicator.
Shown here after machining.
||The acme threading tool was made of silver steel, machined, ground &
filed to shape. Hardened & tempered. Finally, honed with a
||An oversize piece of phosphor bronze was mounted, turned down to 5 to 10
thou under 1" (allows for braze to flow into the joint). Parted
off at 2" long. Re-mounted, centred carefully, step drilled
& bored to the required minor diameter. Threads were cut
(3/4" x 8tpi ACME), until reaching the major
diameter. This is the threaded sleeve finished.
||The sleeve is brazed into the half-nuts, making it one complete &
'fused together' unit.
When the brazed unit is slit, a small gap
will be left between the
half nuts. This way, as the nuts bed in, there is allowance made for
the half-nuts to move closer together. Otherwise 'slop' would
develop in the thread. I allowed 0.50mm = 20thou = 22% of the total thread depth.
||Finally the half-nuts are separated using a thin saw blade & burrs
removed. At this stage it is ready to re-mount behind the apron, restoring the
lathes' thread cutting ability - at least as far as the saddle is
When setting up a tailstock, manufacturers often make the tailstock 1 to 2
thou high, this allows for future wear - the tailstock needing to be
re-fitted when it gets to greater than 2 to 5 thou low - this standard is
what I aimed for.
||At 1.5mm low, I can only assume there was originally some packing
material in the tailstock, that's been lost before I
acquired this lathe.
The tailstock base & body needed machining to correct the centre's
being 1.5mm low & contacting surfaces being worn. The base-to-body
contact surfaces were milled flat & even, on the mill-drill.
The right amount taken off to allow placement of 1.6mm gauge plate between
them. The centre is now just a couple of thou high (as it should have
been when manufactured).
||The gauge plate is clamped to the base & holes drilled through the
plate, into the base, to suit the locating pin's diameter. These
de-burred. The pins are inserted into the base using Loctite to retain
A scrap of similar gauge plate is drilled, placed over the pin (once
the Loctite is set) as a guide. The top of the pin is lightly ground or filed down
to the 1.6mm height. The guide plate's removed, the pin then has a few
extra thou filed off the top for clearance & de-burred. The
tailstock, assembled, is now ready to use.