Page 2 of 3 FirstFirst 123 LastLast
Results 11 to 20 of 22

Thread: I made a Repeat O Meter

  1. #11
    Join Date
    May 2015
    Location
    Coral Springs, FL
    Posts
    133

    Default

    Quote Originally Posted by Dan_the_Chemist View Post
    It's sort of interesting that you can see the ghost of a note I made on the metal near the bottom foot. I had written the measured depth of each hole with Sharpie, but then I removed it with acetone before pressing in the feet. You can clearly see that the rust didn't affect that area. Odd.
    Hmmm, you might have discovered something... Rust prevention by Sharpie. There are a couple more of those areas visible on your bottom, too. It's possible that the moisture came from your hands.

  2. #12
    Join Date
    Aug 2009
    Posts
    6,663

    Default

    Quote Originally Posted by Dan_the_Chemist View Post
    I went with a flexible strip milled out of the bar.



    Top Lipton had tried an inset piece of flexible bronze spring stock, but he had to drill and tap 14 holes in the steel and the bronze spring stock. That seemed like a lot of bother... It's interesting to note that while he proposed it in an early video and showed it being built, he never mentioned it again in later videos in collaboration with John Saunders. They noted the problems of having a stress riser at the base of a long cut, so they devised a method based on a drilled hole and a saw cut.



    I liked this, but I was a little afraid of trying to predict the necessary thickness of the web that formed the flexure. After all, it was a piece of "gawds only knows what" steel. It could be cold rolled or 4140 or who knows.. So I did some modeling with Fusion 360. My goal was to figure out a flexure that could be thinned to make it more flexible without making it susceptable to easy failure due to stress concentrations.

    I modeled it using various types of steel and chose a value that would give 0.020 flexure under the weight of gravity without coming close to the elastic limits. First thing I noticed was that wider cuts gave greater safety margains... a narrow cut quickly exceeded the safety limits and could lead to failure, but a wide cut spread the stress out over a larger area and it worked better. So I decided on a 0.550 inch wide cut to leave a 0.060 web. This would probably be too stiff, but if I had a good jig set up I could remount it in the mill and thin it.

    In practice 0.060 was too thick. I ended up thinning it to 0.035 to get the desired degree of flex.

    The next thing is the location of the web. All examples I've seen have the flexure on the bottom. I thought about that, but it made the milling more difficult. I would have to hold it so that I could come within 0.060 of the bottom. If I left the web in the middle of the piece (as shown) it would be much easier to hold while using the mill.

    However, there is a question that arises from having the flexure in the middle rather than the bottom. That is the effect of geometry - it's not on line with the feet. Well, neither is the one on the bottom, but a quick geometrical calculation showed that the parallax effect would introduce an absolute reporting error of about 0.000008 inches... Yes, that's a lot of zeros. I.e., if the meter said the probe-foot had risen 0.0001 inches it would be off by up to 0.000008 inches ... 8%. However, it's a consistant systematic error. I could live with that.

    More details to come...
    I always thought that the front piece was independent of the rear. I had no idea that the base was a continuous piece. These are used to measure surface discrepancies in the millionths in some instances and I would think any connection between the two base pieces would hinder it's accuracy by stiffness alone regardless of how thin the connecting section is. It's like a living hinge ! Why doesn't it free float??

    Other thought ..... your not sure what the material is that you used, if it's not stable and decides to move over time won't that lead to inaccuracy??

    JL.............

  3. #13

    Default

    Quote Originally Posted by JoeLee View Post
    I always thought that the front piece was independent of the rear. I had no idea that the base was a continuous piece. These are used to measure surface discrepancies in the millionths in some instances and I would think any connection between the two base pieces would hinder it's accuracy by stiffness alone regardless of how thin the connecting section is. It's like a living hinge ! Why doesn't it free float??

    Other thought ..... your not sure what the material is that you used, if it's not stable and decides to move over time won't that lead to inaccuracy??

    JL.............
    The front piece needs to be able to move up and down at the end where the indicator is. That is, it's essentially tracking just the one foot at the front. Conceptually you could have just had an indicator out front riding on the surface plate itself but the surface texture would make it impossible to read. The front section just damps out all the tiny local variations.

    I wouldn't expect material changes to affect it. The main body rides on a three point base so it's always stable. Even if you imagine a twist at the flexure, the front pad should still be able to map the broad hills or depressions of the surface plate. The only thing the repeat-o-meter couldn't ferret out would be a plate that has a slight, perfectly uniform concave or convex shape. Unless you can set the indicator on a verified flat surface, then move to the plate you're checking and find a uniform indicator change all over. Then verify that when you go back to your flat plate you once again have a zero indicator.
    .
    "People will occasionally stumble over the truth, but most of the time they will pick themselves up and carry on" : Winston Churchill

  4. #14

    Default

    Quote Originally Posted by Dan_the_Chemist View Post
    Yes. I started in woodworking, and used to demonstrate old fashioned woodworking at a living history center. I've spent hours and hours on a spring pole lathe so my Jet wood lathe is a real luxury.

    I have a lot of odd cutoffs of oak and maple that I got from the dumpster of a furniture making factory. So, it's easy to make wooden handles for things.

    You know how sometimes you think (imagine?) faster than you are reading?

    I was just finishing the "spent hours and hours on a spring pole lathe" part of your sentence - - - and thought - - - He did those on a spring pole lathe???!!!

    Then I finished the "my Jet wood lathe is a real luxury" part.

    I love working on my lathe and - as I am sure you have discovered - making "only one" pieces are much easier than making "another one to match".

    Nice job on your tool - although being a woodworker - I have no idea what you are supposed to do with the darn thing!


    I know metal work is supposed to be expensive, but let me just say this. I have more invested in my routers and related bits than I do in my LeBlond lathe and related tooling.

  5. #15
    Join Date
    Aug 2016
    Location
    Appalachian Ohio
    Posts
    869

    Default

    Quote Originally Posted by 6270 Productions View Post
    I was just finishing the "spent hours and hours on a spring pole lathe" part of your sentence - - - and thought - - - He did those on a spring pole lathe???!!!
    I think the spring pole lathe is underrated. When I was doing it 8 weekends a year for years on end I could turn one piece wooden baby rattles of the style shown (the picture is a multi part... I did it from one solid chunk of maple). It's not that I'm a particularly gifted turner... it's just by the time you've done a few dozen it gets easy.

    I have made table legs, stools, spinning wheel parts, etc., all on the spring pole lathe. It's really not that bad.

    And it really builds up the leg muscles. I was young, single, and dating for much of that time. Being in good shape was a real benefit.



  6. #16
    Join Date
    Aug 2016
    Location
    Appalachian Ohio
    Posts
    869

    Default

    Quote Originally Posted by TGTool View Post
    The front piece needs to be able to move up and down at the end where the indicator is. That is, it's essentially tracking just the one foot at the front. Conceptually you could have just had an indicator out front riding on the surface plate itself but the surface texture would make it impossible to read. The front section just damps out all the tiny local variations.

    I wouldn't expect material changes to affect it. The main body rides on a three point base so it's always stable. Even if you imagine a twist at the flexure, the front pad should still be able to map the broad hills or depressions of the surface plate. The only thing the repeat-o-meter couldn't ferret out would be a plate that has a slight, perfectly uniform concave or convex shape. Unless you can set the indicator on a verified flat surface, then move to the plate you're checking and find a uniform indicator change all over. Then verify that when you go back to your flat plate you once again have a zero indicator.
    I tried the idea of using a 246 block, attaching a magnetic indicator arm to it, and putting the indicator off to the "front" of the block. The meter needle bounced around a lot. I tried putting a button foot on the indicator and that helped a little, but it still bounced. After I built the Repeat O Meter I saw that the huge mass of the flexxed part of of the meter seems to serve several purposes.

    1) It presses the probe foot down firmly
    2) It acts as a shock absorber to isolate the meter tip from the jounces caused by local surface imperfections

    As far as the idea that a Repeat O Meter won't see a uniform concave or convex surface... that's my understanding too.

    Look at the diagram of the blue surface plate... It has a uniform curvature and the Repeat O Meter shows a small + deflection everywhere. If you had zeroed it out on the left and then pushed it to the right, it would show 0, 0, 0 all the way across.

    The Repeat O Meter shows a sign change in the deflection for each change in direction of curvature. So the sandy colored diagram has one change, and the Repeat O Meter shows +, +, 0, - The pea-green diagram has two changes in direction, and it goes +, 0, -, +



    Of course if you don't sample close enough a small feature can be missed. The wide spacing of only 4 readings on the dark blue plate would not reveal the actual shape of the feature. The red colored diagram shows that if the Repeat O Meter is moved much smaller distances it would more accurately signal that there are three changes in the direction of curvature.




    This analysis is complicated by the fact that the probe is offset from the centeroid of the three feet on the reference block. A meter with the probe in the center of three feet on an equalateral triangle would be much more easy to use analytically. As it is, Repeat O Meters tend to be used mostly as qualitative tools - to tell if:

    "The plate is good/needs work."

  7. #17
    Join Date
    Aug 2016
    Location
    Appalachian Ohio
    Posts
    869

    Default

    Adjustments

    This topic caused me a lot of bother. I wanted to be able to provide coarse and fine adjustments. My Mahr Meter is adjustable +/- .002, so I have to be able to get within that region. After some false starts, the result is shown here:



    Originally I had thought of trying to make the lower adjustment into a differential screw. Since the formula is:

    1/P(effective) = 1/P(1) - 1/P(2)

    It means the closer pitch 1 and pitch 2 are to each other the better the effective pitch. I made a small excel spread sheet and decided to try a differential screw of 32 and 28 tpi. I made one up and played with it in a small jig. It worked more or less as expected, but it wasn't very good. I realized one would have to make it very precise with NO wiggle to be able to use it to adjust to +/- .0001 The big problem is that there is NO loading on it. The Mahr meter has almost no pressure. So any tolerance in the threads shows up easily. My taps and dies were not good enough to make threads with no wiggle. I tried using the taps to make the inner threads and the lathe to single point the outer threads, but the difference between not-fitting and "whoops there is wiggle" was pretty fiddly. I have made 4 different differential screws now, and they all wiggle a bit.

    Meanwhile, I had an epiphany about the other adjustment screw. I had been thinking that a (for example) 10-32 screw pushing against the lower arm would move it 1/32'nd of an inch per turn. But then I realized that both the upper and lower arm would move and split the difference. That meant one would move up by 1/64'th and the other would move down by 1/64'th. Since 1/64'th is 0.015", that means that an adjustment of 23 degrees would move it by 0.001. I can turn a screw that finely.

    Even MORE... I realized that if the two arms were different thickness they would bend by different amounts corresponding to the different thicknesses. The thicker would bend less, the thinner would bend more. In this diagram I show it using two steel arms, one of 0.375" and the other of 0.350" thick. Thus my plan became to build the thing, and it if turned out to be too hard to adjust the arm carrying the meter to range needed, I could thin the other arm a little to make it slightly more bendy.

    It turns out that I didn't have to worry. Once I built it I found that I could easily get it to the right range with tiny little adjustments on the 10-32 bolt.




    That meant I didn't have to worry about making the bottom bolt a differential screw. I just made it a 1/4-28 and the brass is a lock nut to hold it in place. It adjusts up and down as the coarse adjust, then the 10-32 is used to get the needle on scale, and then the meter's adjustment knob is used to zero it.

  8. #18
    Join Date
    Aug 2016
    Location
    Appalachian Ohio
    Posts
    869

    Default

    Chapter the last... "STOP MOVING"

    The original Rahn Repeat O Meter has a couple of arms to prevent the front block from moving too far. That could either put strain on the flexure or bang the meter hard. Instead of two point arms I went with three point arms. Just a gut feeling, no physics.

    I didn't have any shoulder bolts of the right sort of dimension so I made some. I bought some knurled brass 10-32 nuts from the local box store and turned off the inside part to make a .245 dia shoulder. I locktited them to some 10-32 screws to create the shoulder bolts. I drilled and then reamed .250 holes in the steel keeper arm so that there is a little wiggle room, but the front is stopped after it moves a bit. By careful adjustment of the meter position it's possible to prevent the front block from raising up high enough to peg the meter. It can drop away a bit, but I don't mind that.






  9. #19

    Default

    All very nicely done, well thought out and well explained.
    .
    "People will occasionally stumble over the truth, but most of the time they will pick themselves up and carry on" : Winston Churchill

  10. #20
    Join Date
    Dec 2015
    Location
    Chilliwack, BC, Canada
    Posts
    5,648

    Default

    What's the weather like around your area when you made the meter? The rusty spots I'm seeing are a lot like finger smudges. So I'm thinking it was warm and you were sweating. And assuming you oiled the steel perhaps the salt and acids in your sweaty digits had started the damage already? Perhaps simply a good cleaning with some rubbing alky prior to a final light oiling of the meter would have prevented the issue given how the cleaned away sharpie pen area didn't rust.

    Nice meter and good explanations on using it.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •