Part one by Graham Meek

I took delivery of my Emco Maximat Super 11 in 1986. At that time this was only one of two lathes in the Model Engineering world that came with a handwheel dial as standard. The other being the Myford 254 lathe. Having this luxury was something I had grown up with in industry, but the 19 mm per revolution graduation was a bit of a disappointment. Nonetheless, I soon became used to some quick mental arithmetic to get all those desired dimensions above 19 mm that I wanted. There had always been an intention to remedy this shortfall but somehow I never seemed to get around to it. Matters came to a head recently some 30 years later when a series of mental arithmetic mistakes forced my hand. With the resultant mistakes ending in several scrapped components.

Given how the mistakes had happened things would have been a lot simpler if this was a whole number, or even 25 mm per revolution. Preliminary sketches soon showed that the 25 mm dial version, along the lines of the Myford S7 design. Was going to need a dial nearly as big as the Emco handwheel, which looked hideous, so no more work was done on this design. This was a pity because this would have made an ideal base for a 1” dial version for those readers with an Imperial versions of this lathe. The next design route was to make the new dial read 20 mm. This would mean a simple reduction with-in the dial which had a 19 to 20 ratio, or numbers thereof.

However, that was until I decided to check the actual carriage movement for one complete turn of the dial. I was a little surprised to find that the actual distance travelled was 18.97 mm. Given the countless thousands of jobs completed, many of them on a commercial basis. This error had never once scrapped a component. Even when the dial had been used to measure off lengths involving multiple turns. It did somewhat put a “stone in the ashes” as regards my simple 19 to 20 ratio though. Of course I could have just left it at that and used the 19 to 20 ratio, but the opportunity to get it right was beckoning, plus I do so love a challenge.

The existing Emco dial sits directly on the handwheel boss, with the face of the dial in close proximity to the apron wall. This can just be made out in Photo 1 (above). Although this photograph was originally taken to show my version of Jacques Maurel’s quick retracting screw cutting attachment. In the Emco design the handwheel is used to transmit the drive via a woodruff key set into the pinion shaft and any endfloat is eliminated via a Symonds, or Nyloc nut on the exposed end of that shaft. The dial is free to rotate on the handwheel boss, but a semi-circular spring set into an undercut in the handwheel boss, provides friction on the dial. This means it is possible to zero the dial at any point and the dial will stay put. That is provided attention is paid to the endfloat. Slight up and down movements of the handwheel during the carriage travel can cause the dial to rub the apron wall, if the endfloat is too great. This rubbing can alter the dial as I had found out once to my cost. It is in this space that I also wanted to incorporate a radial ball bearing race to eliminate this slight up and down movement brought on by 30 years of wear.

There was never going to be an easy way to incorporate all that I wanted to in the existing envelope. This was beginning to feel as though I was trying to “Square the Circle”, ie trying to do a very difficult or impossible task. Finally I did manage to square the circle as can be seen in Photo 2 (below).

Eventually after many permutations I found that a 37 tooth gear driving a 39 tooth gear gave me a near perfect ratio, to convert the 18.97 mm to the desired 20 mm. The theoretical distance moved for one revolution of the dial would give 19.995 mm of travel, somewhat nearer than the standard set-up. The only problem with this ratio was the need to mesh both gears using a common idler gear. The reader will see from the General Assembly (GA), that these gears are co-axial. Two standard gears could never mesh with one plain idler gear. The only solution is to increase or reduce the PCD of one of the gears. Given the design constraints, reducing the 39 tooth gear was not really an option, the only alternative was to increase the 37 tooth gear.The problem is that this is only usually done in industry with the use of a gear hobbing machine, or unless you own a Jacobs gear hobber. As I do not possess any form of gear hobbing machine, this needed a different approach. I have always been comfortable drawing gear tooth profiles for standard gears and making my own single point gear cutters from these drawings. However where to start with an increased PCD gear was something I was not familiar with. The time was ripe for a perusal through Machinery’s Handbook, (or the Engineers Bible). Sure enough in the “Gearing” chapter, under a section headed, “Circular thickness of tooth when outside diameter has been enlarged”, page 677 of the Fifteenth Edition. I found what I was looking for.

This section gives the formula:

t = p/2 + e tan

Where, t = Tooth thickness, (at new PCD)

             p = Pitch diameter of standard gear

             e = Amount the outside diameter is increased over standard gear.

             = Pressure angle

Using the above formula it is possible to draw the required gear and those with AutoCad or similar will find it dead easy. These gears as they are drawn have no backlash. In order to create backlash, or play in any gear train. It is just a simple matter of cutting the teeth slightly deeper. As a very rough rule of thumb, every extra 0.025 mm in depth, gives 0.08 mm of backlash. In years gone by, the commercial gear cutters used to come with the depth of cut etched or stamped on the side, in the form of “D+f”. Where “f” is the extra increment to give the desired backlash. Sadly these days, even given we have the luxury of laser printing this information has been omitted on those new cutters I have purchased. One other point to make about the designed gear form; is that this is the correct tooth form for that particular gear. This is as close to a generated form, ie, hobbed form, that is is possible to get with a form tool cutter. The result of making these gears this way is that they are very smooth running.

However for this application it would be advantageous to have no, or as little backlash in the system as possible. To this end I worked to the basic Whole depth of the gear tooth, and just to make sure I cut one Delrin Idler Gear shallow in depth by 0.03 mm, while the other was cut to full depth. As it worked out the full depth gear gave no backlash, but I still have the larger one should any wear take place.

The next problem to raise its head was that there was no provision on the Emco or my own design of Dividing Head, Photo 3 (below), to cut 37 teeth. This was overcome by using AutoCad and the Array command to draw a new 37 hole indexing plate. Measuring off the coordinates on the drawing for each hole with reference to the indexing plate centre, a table of coordinates was prepared to drill the circle of holes on the FB 2. I hasten to add that my FB2 is not equipped with a Digital Readout. Some may think this method is fraught with errors but the resulting indexing plate did not show up any pitch errors and the Dividing Head indexing pin entered every hole as it should.



the only free and the only weekly magazine for model engineers. 

Editor: David Carpenter