The standard gear train (40T, 95T, 44T & 88T - for a SAE lead screw) of the lathe covers ~85% of all threads / feeds. The 56T gear is used to produce the rest (it replaces the 44T in the gear train). It was missing when I got the machine. So it was time to make one. The members who attended the meet-up last Saturday saw the partially finished gear on the Deckel.
The gear specs are as follows:
56 tooth spur gear, 20* pressure angle, module 1.5, steel, 1/2 inch thick, internal 6-spline (approximate SAE 1 inch x 6; Colchester likes to modify things!).
I used a piece of A36 (44W), 1/2” mild steel plate roughly cut to size with a grinder. Drilled and bored a hole in the middle for an arbour. Took a chunk of round bar and turned the OD, a step and shoulder and a tight fitting seat for the previously bored blank. Single point threaded the end for 3/4”x10 TPI, all in one set-up. Then I mounted the rough blank on the arbour and turned it to size (OD equals # of teeth + 2 times the module ==> 87.00mm).
Then I partied the arbour off, leaving the gear blank on it for the next operation.
My universal dividing head is very precise and came with a really tight Hardinge 3-J chuck. I set everything up on the Deckel FP3L and decided to use the horizontal spindle for the job (the vertical spindle could also have been used, but it adds one extra source of error (still would have been just fine as that mill is tight all around). After indicating in all planes / axis to < 1 thou TIR, it was time to do the first pass.
2mm depth of cut, 315 rpm, 40 mm/min feed rate. Indirect indexing was 35 holes on the 49 hole circle. The second pass was 1.375mm, same speed and feed.
Using an online gear calculator to determine the measurement over wires, the pitch circle diameter after the second pass was 0.197mm too large. Since that measurement is based on a true involute gear tooth profile for 56T (I am using an approximation when cutting gears with involute cutters), I left it alone. The gear meshes well with a sample from the gear train and a banjo is used to set the gear back lash.
The third operation was on the Bridgeport milling machine: cutting the internal spline. Dimensions were copied off another gear. I had to use the rotary table and a 4-J chuck to do the indexing as the little 3-J does not open wide enough for the gear (no, I do not have outside jaws for it; seems to be the story of my life when it comes to my 3-J chucks…no outside jaws; they are on the list of things to make).
Once everything was set-up concentric under the spindle, I used a 1/4” endmill (actual diameter 0.2475”) with the appropriate offset to hog out most of the material by plunge cutting.
Then I switched to my broaching set-up. I was able to use the same tool I had ground when I made the key way in the carriage hand wheel for the CMT Ursus 250 lathe. Three passes were required: one in the center, then one each on the side offset to get the width. I had to do another 2 thou pass in the end on one side as the spline off the banjo is a little tighter than the output / input splines.
If I had an overall more rigid broaching set-up and an exact width tool, it probably would have been done in one pass. I used gauge blocks to monitor progress on the width.
It fits!
I will use the surface grinder to clear off the mill scale on the back side and the pantograph to engrave the # teeth on the front as I don’t have any stamps.
The gear will then be mounted to the inside of the change gear cover so that it won’t get lost.
The gear specs are as follows:
56 tooth spur gear, 20* pressure angle, module 1.5, steel, 1/2 inch thick, internal 6-spline (approximate SAE 1 inch x 6; Colchester likes to modify things!).
I used a piece of A36 (44W), 1/2” mild steel plate roughly cut to size with a grinder. Drilled and bored a hole in the middle for an arbour. Took a chunk of round bar and turned the OD, a step and shoulder and a tight fitting seat for the previously bored blank. Single point threaded the end for 3/4”x10 TPI, all in one set-up. Then I mounted the rough blank on the arbour and turned it to size (OD equals # of teeth + 2 times the module ==> 87.00mm).
Then I partied the arbour off, leaving the gear blank on it for the next operation.
My universal dividing head is very precise and came with a really tight Hardinge 3-J chuck. I set everything up on the Deckel FP3L and decided to use the horizontal spindle for the job (the vertical spindle could also have been used, but it adds one extra source of error (still would have been just fine as that mill is tight all around). After indicating in all planes / axis to < 1 thou TIR, it was time to do the first pass.
2mm depth of cut, 315 rpm, 40 mm/min feed rate. Indirect indexing was 35 holes on the 49 hole circle. The second pass was 1.375mm, same speed and feed.
Using an online gear calculator to determine the measurement over wires, the pitch circle diameter after the second pass was 0.197mm too large. Since that measurement is based on a true involute gear tooth profile for 56T (I am using an approximation when cutting gears with involute cutters), I left it alone. The gear meshes well with a sample from the gear train and a banjo is used to set the gear back lash.
The third operation was on the Bridgeport milling machine: cutting the internal spline. Dimensions were copied off another gear. I had to use the rotary table and a 4-J chuck to do the indexing as the little 3-J does not open wide enough for the gear (no, I do not have outside jaws for it; seems to be the story of my life when it comes to my 3-J chucks…no outside jaws; they are on the list of things to make).
Once everything was set-up concentric under the spindle, I used a 1/4” endmill (actual diameter 0.2475”) with the appropriate offset to hog out most of the material by plunge cutting.
Then I switched to my broaching set-up. I was able to use the same tool I had ground when I made the key way in the carriage hand wheel for the CMT Ursus 250 lathe. Three passes were required: one in the center, then one each on the side offset to get the width. I had to do another 2 thou pass in the end on one side as the spline off the banjo is a little tighter than the output / input splines.
If I had an overall more rigid broaching set-up and an exact width tool, it probably would have been done in one pass. I used gauge blocks to monitor progress on the width.
It fits!
I will use the surface grinder to clear off the mill scale on the back side and the pantograph to engrave the # teeth on the front as I don’t have any stamps.
The gear will then be mounted to the inside of the change gear cover so that it won’t get lost.
Last edited:
