Announcement

Collapse
No announcement yet.

Vernier for a Rotary Table

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Originally posted by Bented View Post
    7 minutes of arc is .1166 Deg.
    Tan = .002
    .0122 X 6 = .0122
    .025 included, this is close enough
    .
    .0122 X 6 = .0122 ...huh?? I think you mean the tangent of the angle X 6" .
    Lynn (Huntsville, AL)

    Comment


    • #17
      Originally posted by Jim Stewart View Post

      I calculate it to be around .0003 over 6".

      -js
      6 * tan (7/60) = 0.0122... is the radial difference. Multiply that by 2 to get the diameter difference and you have 0.0244... which is very close to 0.025.
      Last edited by mklotz; 02-12-2021, 03:41 PM.
      Regards, Marv

      Home Shop Freeware - Tools for People Who Build Things
      http://www.myvirtualnetwork.com/mklotz

      Location: LA, CA, USA

      Comment


      • #18
        Originally posted by Baz View Post
        It is kind of pointless going too far as the wormwheel won't be all that accurate. Simplest available already thing to start with is to make an adaptor that holds your changewheels, and a marker or detent.
        I'm curious about why you'd say this? What you're suggesting is the sort of accuracy we expect from existing rotary tables and indexers. And which would not work well with the divider plates if they were as bad as you're suggesting. Or are you suggesting that home shop made worms and crown wheels would be that way?
        Chilliwack BC, Canada

        Comment


        • #19
          Not necessarily. My RT has a 90::1 worm and the scale on the worm reads in arc minutes. It has a six part Vernier so that gives 10 arc second as it's smallest readable division. The manufacturer's stated accuracy is +/- 30 arc seconds. I have no way to verify that but I also have no evidence that it is not true.

          If it is accurate to that 1/2 arc minute spec, then having the 10 arc second Vernier is a worthwhile feature because it removes any doubt or inaccuracy in setting it to the stated accuracy.

          Most worms and worm wheels are not made in a home shop: they are usually purchased when a home shop owner wants one. And the accuracy of a purchased worm and worm wheel can be very good. And even if they are made in a home shop, there are ways of making them with excellent accuracy. So, with the careful selection of a supplier or with care in their manufacture, excellent accuracy is possible. Is it needed in a home shop? Well, the answer to that depends on what the home shop owner is trying to do.



          Originally posted by mklotz View Post

          Too right. There are 21600 arcminutes in a circle. If you think you can obtain meaningful, repeatable subdivisions of an arcminute with a DIY mechanical device, you're deluding yourself.

          Not to mention the fact that it's highly unlikely that anything built in a home machine shop is going to need that sort of accuracy.
          Paul A.
          SE Texas

          And if you look REAL close at an analog signal,
          You will find that it has discrete steps.

          Comment


          • #20
            Originally posted by Bented View Post
            7 minutes of arc is .1166 Deg.
            Tan = .002
            .0122 X 6 = .0122
            .025 included, this is close enough
            .
            I’m curious as to why you and your boss thought the customer wouldn’t be able to measure this?

            Comment


            • #21
              Originally posted by MikeL46 View Post
              What kind of accuracy are you looking for? a 90:1 gives 1 degree every 1/4 turn. Attach a simple spindex to divide that by 360 and it should satisfy most needs.

              Mike
              This!

              Dont get overly complicated with the ratios. There are only two . The 90:1 is the better of them. JR
              My old yahoo group. Bridgeport Mill Group

              https://groups.yahoo.com/neo/groups/...port_mill/info

              Comment


              • #22
                Originally posted by oxford View Post

                I’m curious as to why you and your boss thought the customer wouldn’t be able to measure this?
                The customer does not have the equipment nor personnel to do so.



                Comment


                • #23
                  Originally posted by Paul Alciatore View Post
                  Well, the answer to that depends on what the home shop owner is trying to do.
                  Too right.
                  On another forum last month we went through something very similar. The OP asked for a specific angle to accuracy of seconds. Eventually after a few days it transpired he just wanted to cut a course gear and had even got the sums wrong so the real angle he needed was quite easy but a dividing arrangement would also have been easier still to avoid adding up all the angle offsets.
                  In a similar vein on a lathe specific group some unfortunate soul had bought a special custom changewheel to get an odd imperial thread to approximate a metric thread to match a miss-measured imperial thread. In the end his every calculation was up the creek and he could have used his existing gears to cut both the 'wrong' metric and the actual thread he needed.

                  It always helps to be more up front with what you are actually trying to do.

                  Comment


                  • #24
                    Lacking a knowledge of circles leads many astray.
                    Simply put.

                    Circles are divided by the orbital position of certain planets, planet X is never used because it is hypothetical.
                    When dividing gear teeth only use Venus to Saturn unless there is a Mars/Venus alignment at the time, this will cause problems.

                    Comment


                    • #25
                      Originally posted by mklotz View Post
                      With a 90:1 worm gear, one rotation of the cra........

                      Provide 4 divisions on the crank disk for 4 degrees. Subdivide each of those one deg segments into N divisions. Then each division will correspond to 60/N arcminutes.

                      With N = 20 a division will correspond to 3 arcminutes.

                      3 arcminutes = 0.00087 radians (call it 1 milliradian)

                      A milliradian is very small; if you misalign your laser by 1 milliradian at a target 100 yards (3600 in) away, you'll miss the bullseye by only 3.6 in.

                      This level of precision is better than almost any DIY project could require.
                      MKLOTZ
                      Thanks, I like the idea but I need detail explanation on the implementation to build the thing. I get lost or cannot figure out how to build the gadget after I got the 90 divisions and dial each division to give me 4 degrees. Do i need to build a "vernier" of division plates? Any way you could point to somewhere this was implemented?
                      I understand the math, cannot figure out the Building of the "thing".
                      Thanks a lot

                      Comment


                      • #26
                        Originally posted by Paul Alciatore View Post
                        Not necessarily. My RT has a 90::1 worm and the scale on the worm reads in arc minutes. It has a six part Vernier so that gives 10 arc second as it's smallest readable division. The manufacturer's stated accuracy is +/- 30 arc seconds. I have no way to verify that but I also have no evidence that it is not true.

                        If it is accurate to that 1/2 arc minute spec, then having the 10 arc second Vernier is a worthwhile feature because it removes any doubt or inaccuracy in setting it to the stated accuracy.

                        Most worms and worm wheels are not made in a home shop: they are usually purchased when a home shop owner wants one. And the accuracy of a purchased worm and worm wheel can be very good. And even if they are made in a home shop, there are ways of making them with excellent accuracy. So, with the careful selection of a supplier or with care in their manufacture, excellent accuracy is possible. Is it needed in a home shop? Well, the answer to that depends on what the home shop owner is trying to do.




                        Paul: Any chance you could send me a close up picture of your table to see how the "vernier" is incorporated/built?
                        Thanks

                        Comment


                        • #27
                          Originally posted by Bented View Post
                          Lacking a knowledge of circles leads many astray.
                          Simply put.

                          Circles are divided by the orbital position of certain planets, planet X is never used because it is hypothetical.
                          When dividing gear teeth only use Venus to Saturn unless there is a Mars/Venus alignment at the time, this will cause problems.
                          Bented: I love the picture with the circle divided so evenly. Any pointers on how could I implement that and build myself for example a collar of 8-12" diameter and scribe the divisions?
                          I was thinking of using an old saw blade of let's say 1/2" pitch, but I only have 3/4" PITCH from my LumberMate 2000 Sawmill.
                          Thanks

                          Comment


                          • #28
                            Originally posted by ElectronMini View Post
                            MKLOTZ
                            Thanks, I like the idea but I need detail explanation on the implementation to build the thing. I get lost or cannot figure out how to build the gadget after I got the 90 divisions and dial each division to give me 4 degrees. Do i need to build a "vernier" of division plates? Any way you could point to somewhere this was implemented?
                            I understand the math, cannot figure out the Building of the "thing".
                            Thanks a lot
                            There are many ways to build it so I'll just try to outline the principle which can then be mechanically implemented in a number of ways.

                            Imagine that you had a circular disk (the 'disk' could be a circular sleeve) that was marked with 4 * 20 = 80 divisions. You can attach this disk to the shaft that drives the worm that turns the table. By providing a pointer fixed to the table, you would have your four degrees per worm revolution as well as the three arcminute subdivisions of each of the four degrees.

                            So the problem becomes one of coming up with a scheme to subdivide a circle into 80 parts. If you could find an 80 tooth gear you could use it directly or attach a disk to it and use it as a reference to index the disk and mark the 80 divisions.

                            Another approach is to find a friend or machinist club member who can be bribed with beer (that ought to be easy) and have him use his rotary table to divide the disk for you.

                            Another approach is to make yourself a cardstock master and use that to mark the disk. If you draw a large circle of diameter D, and then set your dividers to a length given by:

                            L = D * sin(180/N)

                            you can use them to strike off N divisions around the circle.

                            As an example, take D to be 255 mm, then with N = 80 we have L = 10.01 mm. (Yes, I adjusted the numbers to make it come out to about a centimeter).

                            Yes, using cardboard will incur some error but remember that any error occurred here will be reduced by a factor of 90 when translated into table movement. We can use that reduction iteratively to further reduce the error. Divide one disk using the cardstock model, mount it and use it to divide another disk which will be 90 times better. Rinse and repeat until you perish from boredom or accept the fact that you will never make anything that requires sub-arcminute accuracy.

                            Once you decide how you will approach the task, keep me informed. I'll be glad to provide whatever help I can.

                            Regards, Marv

                            Home Shop Freeware - Tools for People Who Build Things
                            http://www.myvirtualnetwork.com/mklotz

                            Location: LA, CA, USA

                            Comment


                            • #29
                              Originally posted by ElectronMini View Post

                              Bented: I love the picture with the circle divided so evenly.
                              Any simple free CAD program can do so with a few key strokes.

                              Moving from a simple drawing to machine instructions is a different matter that is often called CAM. A very accurate printer will do this.

                              Comment


                              • #30
                                Originally posted by ElectronMini View Post

                                Bented: I love the picture with the circle divided so evenly. Any pointers on how could I implement that and build myself for example a collar of 8-12" diameter and scribe the divisions?
                                I was thinking of using an old saw blade of let's say 1/2" pitch, but I only have 3/4" PITCH from my LumberMate 2000 Sawmill.
                                Thanks
                                Odd as it might sound doing a layout like Bented's circle in CAD is only the work of a few minutes. The trick once you have the drawing is to print it accurately. Many printers compress or stretch the image slightly. I've run into that on occasion where it was important to get a really good proper size. Like even the roughly 1% distortion in the one direction was too much for what I wanted to do. So I copied my original CAD drawing to a new document and stretched or compressed (many years ago but I think it needed stretching by some decimal amount like 1.3 or 0.85 or some such nonsense) the drawing and saved it as a new file which I then printed.

                                I found the error by drawing a simple box on a sheet of paper that was near the maximum printable size and measured it. Then I scaled the one direction.

                                Bottom line is that if you would like to work with a printed page drawing it would not be hard to send you a file with a test box which you print and measure. Then I'd scale the box as needed and send it back to you. When we determine the scaling factor needed I could send you a PDF file that would print out in Acrobat on two legal size pages which you would trim in one case and tape it over the other to get a nice big indexing circle with 360 marks. And a big one for every 10th mark and half big at each 5th mark between.

                                You didn't mention yet if you're making a simple rotary table or if you're doing the whole worm gear setup. It sort of sounds like you're thinking of the whole full meal deal. that's a LOT of pretty darn precise work. And for doing that sort of work you would generally need an indexing head or at least a place to borrow one for such a job. Not to mention some pretty fancy welded assemblies or castings to make the base and the wheel and such.

                                Bet that as it may I found this picture online which I think illustrates how most 90:1 tables with 4° handwheels are set up. It's certainly how my 6" table is marked. The file is pretty big so click on the pic below to see the full size version.

                                To set the stage what you're seeing is the 0 index on the 60-0-60 vernier scale is sitting on the 40 minute marker that is resting between the 2° and 3° marks which can be seen just outside the shadow bar on the lower side for the "2" and just before the shadow line near the top for the "3".

                                Each mark from the 30m mark to the next bigger mark which is the 40M mark is two minutes of arc. The vernier is also two minutes of range being bascially -60sec to 0 to +60sec in 20 sec increments. So technically this table would be good for + or - 20 seconds. But I'm sure that with a magnifier you could fudge it to closer to + or - 10sec of arc. Whether or not the table will respond to such subtle shifts is another thing. But all else being equal it should.

                                To use this particular style of vernier when it is between two marks on the minutes index you would find which minute marker the 0 line is closer to. Then find the vernier mark to that side which is closest to one of the other marks. You then either add to the lower minutes mark or subtract from the higher minutes mark as needed to correctly split the 2min division. Clear as mud? I know it took me a fair few minutes of head scratching when I got the table and it's still a little bit of stop and ponder moment when I use it even now.



                                Click image for larger version  Name:	rotarytablevernier.jpg Views:	0 Size:	155.5 KB ID:	1928032
                                If you want to make a bandsaw blade indexing wheel perhaps invest in a shorter and cheaper 4 or 6 TPI blade from a 14" saw?

                                For example, if you were to gash out your own 90T worm driven wheel. I think I'd want to start with something like bandsaw stock which is 4TPI. Then I'd want to make the wheel pretty large to minimize any creeping errors. If I used 4TPI stock and went with 180 teeth and used every second one to produce the driven side of the worm set the band would be 45 inches long. And that would fit around a disc that was roughly 45/3.14= 14.3 inch diameter. I say "roughly" because you'd want to join the band with some care to ensure that the teeth at the joint are spaced accurately as you can manage and that the disk would be a hair smaller to account for half the thickness of the blade. At that point using every other tooth would also reduce any build up of errors. Or another option might be some 6TPI stock and go for 360 teeth. That would give you an indexing disc to work with that was 19.1" in diameter. Some extra care would want to be taken if you went with this idea so that the center hole is more accurately located within the outside diameter than is seen in most wood working. Perhaps a flanged bushing or shaft for the center? Then skim the OD for a snug fit of the blade?
                                Chilliwack BC, Canada

                                Comment

                                Working...
                                X