PDA

View Full Version : worm wheels

rmsdist
04-23-2005, 06:44 AM
This question goes out to all of you who enjoy gear cutting. Have you ever tried to generate a 127 tooth worm wheel? If so what were the results. It seems that this is the number that keeps us all mistified. But if you can generate a worm wheel with 127 teeth
you can then use it to cut the much sort after 127 tooth gear, (aka 100/127 metric magic). Am I thinking straight or is my brain still in the winter mode????

John Stevenson
04-23-2005, 07:11 AM
Anything with 127 divisions will work to generate the elusive translation wheel.

It may be hard to get 127 teeth by the free hobbing method where you feed a tap into a blank of a given size.
I must admit I have never done this but from reports I have read it's as easy to get 126 and 128 as 127 by this method.
Many pregash before hobbing but if you can index to gash 127 you can index to cut 127.
The chicken and egg situation.

Other methods put forward by many over the years are using a length of bandsaw blade wrapped round a wooden disk, a length of hose clip material around a disk, a CAD printed "division plate" glued to a disk and many similar devices.

There are also division plates on Ebay that do have the elusive 127 hole circles.

Is it possible to borrow a 127 gear from someone even if it's not correct for your machine to use as template.

John S.

JCD
04-23-2005, 10:26 AM
Havenâ€™t cut a 127T worm, sounds like a good project.
I just completed cutting a 127 tooth gear using compound indexing on a 40:1 dividing head using the 39/49 hole index plate.
The error in the division is neglishible.
JCD

3 Phase Lightbulb
04-23-2005, 11:53 AM
worms have rights too....

Paul Alciatore
04-23-2005, 06:02 PM
I have heard of the difficulties in producing odd numbers of divisions for some time now. I have a rotary table which is set up to be rotated in degrees, minutes, and seconds. The Vernier scale has a resolution of 10 seconds and the the manufacuruer claims an accuracy of +/- 30 seconds. If you do the math, a 10 second error at a radius of 5" (10 inch diameter, which is the size of my table) is 0.00024" or about 1/4 of a thousanth. The figure for 30 seconds is three times that or 0.00072" or about 3/4 of a thousanth. At smaller radai, the numbers will be proportionately smaller.

Now a 127 tooth at 16 DP will be about 7.9" in diameter. So the best I could expect on my table is about +/- 0.00057". That's about as good as I would ever expect.

Now, why have I detailed the above? With about 3 minutes of effort in an Excel spreadsheet, I was able to list the exact angles for each division of a 127 division. Starting at 0, the next division is at 2* 50' 4.72". The one after that is at 5* 40' 9.45". With very little effort, I have all the divisions. And there is no cumulative error because each one is based on the exact (to 7 or 8 decimal places anyway) angle for that division. The only rounding is in the seconds and since I can only set to the nearest 10 seconds, two decimal places there is gross overkill.

My conclusion is that I do not need any setting circles. I can easily make and print a table of values for any number of divisions that is far more accurate than my table (or any table for that matter). Then I just have to dial in each individual setting.

I use this technique on a table divided into conventional degrees, minutes, and seconds but it can be used for any number of divisions with any Vernier. It will produce results that are as good as the table used.

Paul A.

Jim Hubbell
04-25-2005, 02:48 AM
Paul
I used the system you describe with my rotary table and have a 'transposing gear' of 127 teeth. My change gears are 16DP and the 8in. radius would leave little room for the rest of the gear train. My 127 tooth gear is 32DP and about the same size as a 64 tooth, 16DP gear. Two more 32DP gears ( 40 and 80 tooth) allow me to cut most any Metric pitch I wish.

alwynoak
04-25-2005, 07:33 PM
i could be mistaken here,but i dont see how a 127 worm wheel can be used to make a 127 spur gear.1-1 ratio,is this what you mean?

rmsdist
04-27-2005, 06:11 AM
My thinking is if it takes 127 turns of the worm to make one full revolution of the worm wheel then one turn of the worm would give me 1/127 of the worm wheel. My intentions were to make an index plate from this setup.
If I am wrong please show me the error of my ways.

Indexer
04-27-2005, 10:44 PM
Obtaining 127 divisions is easily done with compound indexing using 2-23/39 + 12/49 in 9 revolutions of the work. Granted, it is an approximate solution, but it will yield 127.00018 . . . which is better than the machining process you are likely to be using.

------------------
Rich Kuzmack

Pi = 355/113 . . . to
&lt;85 parts per billion