Phase II 6" Rotary Table Indexing - Dividing Plate Set

Would anyone here know what the actual increments/hole numbers are correct for a 90:1 ratio Rotary Table?
I've read that there are a specific number of holes in plates, depending
upon the RT's turns ratio, but I don't have a good understanding of the relationships of those numbers.
I know that there are folks posting here that know more about numbers than I probably ever will, and maybe they could shed some light for someone in the dark, about how the numbers on plates correspond to RT positions. Is this sort of information in Machinerys Handbook?
I bought a set of 3 plates and associated hardware, that were said to be for a 6" RT, but the actual correct make/model of RT was unknown. PYH is the brand on the label, and the plates' center holes are too small for the Phase II RT.
So, since the plates may, or may not be the correct number of holes for a 90:1 RT, I thought I'd take a chance on them.
If this plate set isn't the correct one for a 90:1 RT, they may work on a dividing head or other indexing accessory. Since the plate center holes and bolt circle don't match the 6" Phase II RT, they might be adaptable to something else I may get in the future.
PYH TDM1F/6 6" Rotary Table Dividing Plate Set Made in Taiwan
TDM1F/6 Cat # 51-225-1 Model # 05082466
Stock/SKU label number 87405082466 070 (looks like an Enco or maybe MSC label)
These plates look about the same as the Enco/MSC import brand SPI 6" RT plate set.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
I don't know who wrote this, probably someone here. If you recognize your work, please claim it.
I found this to be more help than anything else I ran across in trying to decipher my indexer. -------------------------------
Using the indexer
48:1 It's just a simple division using 48 as the main figure and then work in fractions.
Taking your example of 30 you divide 48 by 30 to get 1 and 18/30th remaining.
This means that for each 1/30 of a tun you need to take the handle round one full turn and 18 holes on a 30 hole circle. As there isn't a 30 hole circle you need to reduce to the lowest denominator so 18/30 = 9/15 = 3/5. You have 15, 16, 17, 18 and 19 so out of these you need to move 9 holes on the 15 circle. --------------------- 40:1 It's just a simple division using 40 as the main figure and then work in fractions.
Taking your example of 30 you divide 40 by 30 to get 1 and 10/30th remaining.
This means that for each 1/30 of a tun you need to take the handle round one full turn and 10 holes on a 30 hole circle. As there isn't a 30 hole circle you need to reduce to the lowest denominator so 10/30 = 5/15 = 1/3. You have 15, 16, 17, 18 and 19 so out of these you need to move 5 holes on the 15 circle.
That plate you have is only part of a set of plates as with just these 5 numbers you won't be able to do all numbers. It's also an odd one as there are usually six rows per plate and three plates per set. -----------------------------------
Each full turn of the handle will rotate the head 1/40 of a full rev. If you want it to rotate 1/30 of a rev instead you need to turn the handle 1 and 1/3 turn each time (40 divided by 30 = 1 1/3). Since your plate has a 15 hole circle one third of 15 equals 5. You would turn the handle one full turn plus five holes. That location becomes the starting point for the next turn of the handle which would be another full turn plus five holes and so on. --------------------------------- Suppose you wanted to do 72 divisions. You'd need to turn the crank 40/72 of a turn (and do that 72 times). So to get 40/72, or 20/36, or 10/18, or 5/9 of a turn, you'd need some plate with a row of circles divisible by 9 in it. Using J Tier's plates, you could do it with the 27-hole circle on his #2 plate, for example.
------------------------- You set up the distance to turn with the two arms that should be in front of the plate. If you need to move 10 holes, you set to show 11 holes, and lock them together (they now turn as a unit).
That is the hole its in, plus the 10 to move.
When you have moved it, you push the arms along to mark the next move.
Whenyou work with it, always do the same order of operations, such as: turn crank, lock spindle, set arms ahead.
Otherwise you will get discombobulated as to where you are....
Once you have the hang of the system it will only take you a minute or so to figure the number of holes to move.
normal set per B&S, L-W etc : Plate #1 is 15, 16, 17, 18, 19, 20. plate #2 is 21, 23, 27, 29, 31, 33. Plate #3 is 37, 39, 41, 43, 47, 49.
Wild_Bill wrote:

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Thanks for posting this info, RB. Some of the authors used to frequent this newsgroup, but not anymore.
My experience with RCM leads me to believe it's because of all the non-metalworking bullshit and worthless gossip getting posted here day after day.
Here is the origin:
http://bbs.homeshopmachinist.net/archive/index.php/t-7790.html
I know that at least a couple of these guys are a wealth of practical information.
John Stevenson and Marv Klotz, just to mention a couple of them. Looking up articles or websites of these guys will reveal lots of great info. I believe John could write volumes about any number of ways to fabricate almost any part, and Marv has a site with free software for solving lots of metalworking problems/puzzles.
--
WB
.........
metalworking projects
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Wild_Bill wrote:

Bill, thanks for adding that attribution. I know of Marv and his site, need to revisit since my interests have changed since the last visit. I don't know that I've seen John's site, but I'll look for it.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Marv is still monitoring from time to time. He has piped up when I question or made a comment refering to him.
Not only does Marv have a nice site, he shares his source code.
I always pay attention to John Stevenson except when locked in verbal combat with Evan.
Wes
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

    Do you have a copy of _Machinery's Handbook_? There are tables in there describing the circles needed to get each number of divisions and such.
    Most are for the 40:1 B&S dividing/indexing heads, which have three plates of six circles each.
    However there is also a table for the Hardinge dividing heads which have a 20:1 ratio (IIRC), and need a lot more plates.
    Also, there are formulas to allow you to calculate what number of holes is needed for each number of index points, which might be where you need to start, since you are using a rotary table instead of a dividing/index head.

    Multi-page tables, including differential indexing (which requires stacks of gears rotating the index plate as the crank is turned to get some of the figures which don't work out well for a normal 40:1 ratio.
    The 90:1 ratio is not chosen for good division ratios, but rather to get an even number of degrees on one turn of the crank (4 for a 90:1 ratio). The 40:1 comes out at 9 degrees on the dial which is more awkward.

    Well ... they *will* work on 90:1 for some numbers of holes. Let's see -- 4 degrees per turn of a 90:1 crank, and divide that by 37 holes (one of the choices on the 40:1 plates) and you get 0.2432 degrees, or 0 degrees 14' 36"

    Right -- or adaptable to the RT, with a collar to ensure centering and various other bits. The main thing is to get the formulas out of _Machinery's Handbook_ so you can calculate how many holes will be needed -- and remember that any integer multiple of the number of holes needed can be used -- for example if you only need four holes, you can skip over 9 holes per on a 36 hole circle, or 10 holes on a 40 hole circle and so on.

    How different do you expect a plate set to look (other than overall size)?
    Good Luck,         DoN.
--
Email: < snipped-for-privacy@d-and-d.com> | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Thanks for the additional info DoN. I thought that I'd remembered from earlier rotary table and dividing head discussions, that there were details in Machinerys Handbook. I've known that I should get one (one of the great procrastinators, ya know).
I haven't used indexing plate sets, and the only differences that I've seen is that generally, there may be 2 or 3 plates in a set. I wasn't aware that there were a specific number of sets of holes (six per plate), or other standards, but it stands to reason since these aspects have been figured out before the manufacturing of plate sets began. I'd read that some plates have different divisions on two sides of each plate. Other differences would be center hole diameter and number of hole for attaching the plates (3 or 4).
For simple indexing, I have a 5C spin index, and that's about the extent of my indexing experience.
What I think would be best for me is a digital readout that displays x number of degrees with a resolution of .001, to be sure about the settings before cutting.
The minutes and seconds method is beyond my comprehension (or desire to), and it seems to be an unneccesarily more complex way of thinking about locations in a single rotation. It may work well for others that have experience with it, but it just seems to be way more complex than other positioning methods.
The feeble grey matter doesn't function at the levels of a CNC computer. I would see fitting a DRO to a indexing device to be an easier solution. Encoders have become smaller, and digital counters aren't too complex to set up, so having a display with 0-360.001 or 0-3600.00 seems to be a good way to achieve positioning locations.
For any method of indexing, I wouldn't have a clue as to how much error might be introduced by accumulation, at say, the last tooth on a 100T (or 113T) gear blank, for example. Hopefully, one could make other adjustments in a design so a 113T gear wouldn't be required. For cutting 6 splines on a shaft, I would assume this would be a fairly simple task with indexing plates.
--
WB
.........
metalworking projects
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Wed, 7 Jan 2009 17:53:19 -0500, "Wild_Bill"

------------ For hobby use the older editions may prove more helpful. http://www.alibris.com /
For just the information on indexing see http://lindsaybks.com/bks/mmindx/index.html There are some other reprints available that are shown in their catalog but not on their web page. Indexing     869        $4.00
If you are exceptional ambitious you can even build your own dividing head/indexer and/or custom plates from the Gingery plans. see bottom of http://lindsaybks.com/dgjp/djgbk/series/index.html
Several of the other Lindsay books also have information on dividing/indexing and gear manufacture.
Unka' George [George McDuffee] ------------------------------------------- He that will not apply new remedies, must expect new evils: for Time is the greatest innovator: and if Time, of course, alter things to the worse, and wisdom and counsel shall not alter them to the better, what shall be the end?
Francis Bacon (1561-1626), English philosopher, essayist, statesman. Essays, "Of Innovations" (1597-1625).
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

    Grab one at a used bookstore -- or an eBay auction. I would say that for a non-CNC hobby metalworker, anything from about the 15th edition on would be quite useful -- though if you want specs of standard collets, you need the 25th edition or later.

    The specific number of circles per plate is six for the B&S ones, which come with a set of three plates.
    You could have more circles (thus fewer plates) with larger diameter plates and longer indexing arms.

    Thicker plates (to keep the holes from going through and interfering with each other) and fewer plates needed for the a given division ratio of the gears.

    And whether the holes are countersunk or counterbored. IIRC, the B&S ones are counterbored, and fairly thick plates which can be mounted from either side, though you have the same number of holes on each side.

    A resolution of one degree on those.

    Hmm ... 0.001 degrees is 0 degrees 0' 4" which is probably fine enough for most things.

    It comes from an earlier culture -- the Egyptian, IIRC -- who liked to base things on 60s -- and it is what our timekeeping comes from as well.
    But it is not that much more difficult than miles, feet, and inches, or feet, inches, and fractions -- you are just accustomed to those.

    However -- if you get a plan which specifies angles in degrees minutes and seconds, you'll have to deal with it -- at least long enough to convert it to decimal degrees. (This is where having an old HP 15C scientific calculator is nice. It includes keys which will convert either way -- useful for both time and angles.

    Actually -- a CNC computer is running at a rather dumb level, so you are probably better than it. It is just more patient at following precise directions -- which *you* need to give it -- or a program (or set of programs) on another computer needs to give it.

    Hmm ... if it is reading the absolute angle in binary, it would need at least 19 circles of encoding bars, which would make it fairly large. For reading it incrementally, you would need a single circle with 360000 bars, plus a single index indicator hole.
    You *could* read the angle the crank turns, and count full turns, but then you would not get a true reading to 0.001 degree out of 360 degrees. The backlash in the worm, and perhaps eccentricity of the wheel could introduce a lot more error than that. You really want to measure the angle at the rim of the table, which means that you would need a scale wrapped around the rim of the table, and given the resolution of the typical digital calipers (ignoring the 0.0005" step), you would need a scale 360" long, or a table with a diameter of 114.59 inches, or 9.5493 feet.

    That's what the worm gear and the plates are used for. You look up in the table the number of teeth (divisions), it will suggest one or more circles (so you may not have to change discs), how many full turns plus how many holes to move from one to the next, and the angle to set the arms to so they will guide you to the right hole. Then you shift the pair of arms so the starting arm comes to rest against the pin were the finish arm was at the end of your previous cranking.
    When you complete your circle, you are *right* on by the nature of the worm gear.
    Note that you did not have to know either degrees and decimal fraction, or the degrees, minutes and seconds. All you needed to know was the number of teeth or divisions you needed to make, and where to find the lookup table (either in Machinery's handbook, or on a metal plate which came with your dividing head when new.
    However -- cranking in degrees and decimal fraction or degrees and minutes (I don't think that seconds are marked on the dial on a rotary table) does lead to accumulated error by the time you finish your circle.
    Be warned that some numbers of divisions are very difficult on a dividing head (unless you go to a differential head perhaps). 127 teeth is an example of that which is important to lathe users.

    Very much so. With a 90:1 ratio, you use a single hole on any circle of the plate, and crank the table through 15 turns of the crank and end up back at the same hole.
    With my 40:1 B&S index head, it is 6-2/3 turns, so you would need a circle with a integer multiple of three holes -- such as 18, or 36 to mention a couple of common ones. You would also need to set the arms to stop you at the proper number of holes past the starting point.
    Enjoy,         DoN.
--
Email: < snipped-for-privacy@d-and-d.com> | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
For a 90:1 ratio, you index 90 increments of the number of divisions you want. For example, if you need 12 divisions, you index 90/12 that can be halved to 45/6 and further reduced to 15/2. So, you need 7-1/2 turns of the handle to make 12 equal divisions on a 90:1 rotary table. So, in this case you can use any even number of holes in the dividing plate, 7 turns plus 1/2, 2/4, 3/6, ...
Another example, if you need 25 divisions, you'll need 90 increments of 25 90/25 reduces to (2*3*3*5)/(5*5) 18/5, that's 3-3/5 so you could use 3/5, 6/10, 9/15... So if you have a plate with 15 holes, you could use 3 turns plus 9 holes.
Another example, if you need 100 divisions, you would have to turn 90/100 that reduces to 9/10, you could increment 9/10 or 18/20 ... depending on the plates you have.
The bottom line is that if the plates you have fit your table then you can reduce the ratio to minimum and raise it in minimum increments to see which plate you need to use. If you need a number of holes that is a prime number, you'll need that number of holes in the plate if it's a 40:1 or 90:1 ratio.
RogerN

<snip>
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Thanks for including those examples Roger.
--
WB
.........
metalworking projects
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I just made up the index tables for 40 and 90 using OpenOffice Calc. Rather than posting it I'll describe how so you can make your own for whatever number of divisions you need, or a different type of indexer. I leave space at the top for headers, comments and the file name. Don't type the quotes, only what's between them.
Enter headers "N" in cell A3, "40/N" in B3, "90/N" in C3. Enter "1" in cell A4, "=1+A4" in A5. It should show 2. Select and copy A5, then select a block from A6 downward. Paste. The pasted cells should read 2 3 4 5 6 etc Continue until you have all the sequential numbers you want in column A. I went to 100.
Now type in the indexing formulas. Enter "=40/A4" in B4, "=90/A4" in C4. Select and copy B4 & C4, paste them in below as far as the end of the numbers in A. Select columns B and C and format their cells as Fraction. Change the format code to # ??/??
You should see: N 40/N 90/N 1 40 90 2 20 45 3 13 1/3 30 4 10 22 1/2
Columns B and C give the full and partial handle turns to divide by N in column A. For instance, 3 on a 40:1 indexer requites 13 full turns plus 1/3rd more, like 5 spaces on the 15 circle. The denominator of the fractions is always the number of divisions N or a factor of it. It may help to figure out and type in the factors of N in column D, such as 63 = 3 X 3 X 7. Then you can see that 3, 7, 9 and 21 are all factors of 63, and match them to the available hole circles. The sector gear I recently made used 20/34 for 68 teeth.
You could create a short one-page table and then enter whatever value of N you need as text to see which index circle(s) will work for it.
Jim Wilkins
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.