Dividing Heads

[dbb-the-bruce]

So you would have to make a second plate with the same hole layout for the dividing head?

Any errors in dividing plate hole placement that is introduced by calculating and moving in X&Y on a mill will be reduced by the ratio of the worm/gear when in use. Obviously, we are talking about an angular error on the dividing plate - any radial errors won't affect the results of using it, other than making it hard to use the pin locator.

The point is, make any plates you don't have, make them by layout and drill/mill X&Y. You will then have an acceptable set of plates which presumably would allow you to make another set on your dividing head. (without doing a lot of math and plate hole counting, I can't tell you which plates can be used to make more plates, but that's the concept).

The second set that you make on your dividing head using any of the "sloppy" plates made on the mill will be 40 times more accurate (based on your worm ratio). If the sloppy plate hole is off by 1/2 deg, the hole it makes will be off by 1/80 of a deg. It's really that simple.

You can do this as many times as you like, however, once is probably enough for most work, twice is definitely enough - error is reduced to 1/6400.

Regarding the layout of the holes. There are a lot of them, and keeping track manually (cranking and watching dials or a DRO) will be challenging. I would lay it out using any reasonably good drawing program - or maybe CAD depending on what you have. You can then just print it out and stick it on the plate to be drilled (spray adhesive, stick glue etc. anything that will come off with solvent).

Most printers are very accurate in the relative repeatable placement of ink. What you need to account for is differences in X&Y scale. Print out a large square or just four points and verify what the printing "scale" is and also that X & Y are the same.

For instance, if you layout dots 6" apart, you will find that when printed they are not exactly 6" and X & Y may be off by different amounts. Calculate the correction factor in X & Y, apply it as scale in your drawing program and print & check again. This can all be done easily in Inkscape.

- Dave

 

[BROCKWOOD]

For repeatable accuracy the pin & hole sizes should match. You don't want the pin to be able to move inside the holes on the plate.

 

[MrWhoopee]

For repeatable accuracy the pin & hole sizes should match. You don't want the pin to be able to move inside the holes on the plate.

A tapered pin and tapered hole make it easier to get positive location without critical fits.

 

[martik777]

I use this printable divider wheel fixed to a round plate for indexing: https://www.cgtk.co.uk/metalwork/reference/divider

It could also be used as a template to layout the holes

There is a linear scale on that site as wel,l which I use for making calibrated hand wheels ie: 200 division cross feed on the lathe for direct reading

 

[whitmore]

A tapered pin and tapered hole make it easier to get positive location without critical fits.

A mill can finish a hole with single-point boring, but can it make a tapered hole that way? Tapered chucking
reamer, is that a real thing?

 

[markba633csi]

I don't think a tapered hole is really necessary; a tapered pin is sufficient

 

[BROCKWOOD]

Tapered end mills in 1 to 6 degrees are available.
I was just pointing out that not every index plate you can mount on your dividing head (or modify to fit) will be ideal. As an example: there is plenty of room on a Van Norman 7 1/2 DH to adapt the index plates from a Van Norman 10 DH, but the pin holes are twice as big.

 

[tcarrington]

With care and a big gob of math and I am talking about sines and cosines, a good two axis mill can make a dividing plate. I really recommend a 2 axis DRO and use the PCD or BCD function. This will guide to within the accuracy of your mill. You can also do it manually and demonstrate your ability to manage any backlash in your XY and spindle. I would be using the DRO.
Doing the rough math, a 4.5 inch plate results in about 40 thousands per degree of rotation on the dividing plate. Of course a six inch plate would have more. That means 0.001 is a very small number of minutes as input to the dividing head. that works down to the resolution and accuracy of most of the small rotary tables. Taking care with the DRO and you might get down into the tenths.

 

[MarkM]

For a dividing head in my person opinion towards this laying out isn accurate enough. Every error will be compounded.
Do the math and be dead on with your x and y coordinates.
X is radius times cosine of the angle. Y is radius times sine of the angle. Example. A six bolt circle with a four inch diameter. Unless the holes need to be indexed I usually just start my first hole off of the radius then move to the next hole the amount of degrees needed. 2( radius)times the cosine of 60 degrees for x. 2 times the sine of 60 degrees for y. Next hole is 120 degrees and so on.
A dro is nice to not have to worry about backlash but once you know how to work with dials it really isn t an issue.