5mm id bearings showed up today.  They do fit the binding bolts I have. I ordered some brass sheet to make the cover plates out of but it won't show up until Saturday or Sunday.  The binding bolts I bought came in a kit with 6 different lengths and non appear to be close enough in length to work.  Once the brass and some 1mm thick 5mm id washers I ordered show up I can verify.   If one is close but just a little bit too long I can grind/sand the shafts down.  The threads in them are shallow though so there isn't much extra room for the screws they came with.  I think the idea should work, although looking at how small the 5mm bearings are I kinda wish I had a 1/4" version of bearings and binding bolts to try. I think I already have some .25"id bearings. 

The 5mm bearing are 10mm OD and so went to amazon to buy a 10mm forstner bit.  They reminded me I already bought one in February this year.  Sure wish I knew where I hid it. 

 

48 minutes ago, 4DThinker said:

They reminded me I already bought one in February this year.  Sure wish I knew where I hid it. 

It'll resurface the day after the new one arrives. Might check the pencil drawer.

1 hour ago, Grandpadave52 said:

Might check the pencil drawer.

That is exactly where I found it.  A black bit hiding in a black drawer.  Of course always the last place I looked. Drilled a test hole in a scrap of cherry and given that my micrometer claimed the bearings were exactly 10mm diameter they were loose in the hole drilled by the 10mm bit.  The bit was cheap and looks it so I'll blame the bit for the imprecision. Next I'll see if I can CNC a hole just snug enough so the bearings will press in. Get it nailed in before making the holes for it in the actual legs of the cherry version. 

Edited March 22, 20242 yr by 4DThinker

Broke down and ordered some 1/4" o.d. binding screws and bearings to fit.  Scale looks more appropriate for the legs than the 5mm option. Advantage is that the bearings are rated for 28,000rmp which should be just about right for a TV tray table that sits static most of the time.    Once you decide to switch the table from horizontal to slanted you can do it as fast as you want. Worth the time spent searching as a visit to McMasterCarr.com always teaches me something new about what hardware is out there and available. 

4D

23 hours ago, 4DThinker said:

the bearings are rated for 28,000rmp

 

What would "rmp" be 4D?

2 hours ago, Cal said:

 

What would "rmp" be 4D?

A typo.  RPM was the intended. Rotations Per Minute.

12 hours ago, Cal said:

 

What would "rmp" be 4D?

Or "Repeating and Memorable Pulses. 

1/4" id bearing, .06" thick brass plate, and 5mm washers all arrived today. Had a thought about doubling up the 5mm bearings on each side to see if I could get close to the length of a binding bolt I have.  "Very" is the best unit of measurement I can come up with to describe how close.  Have to make the brass plate.  Still waiting on the 1/4"od binding bolts. 

Took a few minutes to measure, mark, and glue up the wood for the outer cherry legs. Trying a different design than the previous two for those outer legs.  I can mark where the pivot and connection to the top should be centered, but need to finalize what I'll use in those connections before making any holes in the legs. 

 

Used my CNC this morning to cut the profile shape on the top of the outer legs.   

Created the vector for one side, then mirrored it for the other. Zeroed from the same relative corner for each.  Some sanding cleanup required to eliminate the pass depth lines.  This bump is where the offset top connection hole will be that allows the design to both fold up flat and flip inside out to end up in flat or slanted positions. 

4D

Figured that while I had the CNC on I might as well do the simpler rounded ends on the legs.

I have an internal battle with myself about how many ways I could have done this.  In my pre-CNC days I'd have drawn the arc on each leg, probably used the band saw to remove the wood close to the line, then at my disk sander spun the legs in an arc to sand up to the line.  My precision might vary a little being a human with non-robotic control over my arms and stance, but the end result would have been OK with a little sanding to "fix" each one.  Or I might have drilled a 1/4" hole into one side at the center of each arc, then made a jig with a 1/4" pin sticking up to impale each end on so I could rotate the leg in a perfect arc against the disk sander or a spinning spiral router bit to send/cut the ends perfectly. Then made 1/4" plugs to fill the holes.  On the inside of each leg no one would ever notice them.  Using my CNC is a bit of a guilty pleasure.  It still takes some attention and careful setup to clamp down each piece and zero out the bit to the end, side, and top. A tap of "R" on my keyboard runs the file and leaves the end perfectly rounded a minute later.   If I had to do 100 or more of these I would refine the toolpath to do a final full-depth pass that wouldn't leave the level marks you can see.  I'm retired now.  This project is just for fun and mental exercise. I don't mind the time I'll spend sanding the ends a bit. 

4D

The last technical challenge to do with this table is to see if the bearing/binding post idea works.  Made a test sample using my CNC in a couple scraps and found the 10mm hole made by the CNC to be too big by .03" or so. Bearings drop in rather than fit snuggly.   Either the 3/16" spiral upcut bit I used was slightly oversized, or play in my CNC is stealing the precision. The sample did demonstrate to me that the binding post screw end will need to be locktited in when assembled or it may unscrew itself.      

Made another test, this time reducing the toolpath with a .003" allowance for the bearing holes.  Fit was better.  No sideways slop, although not a press-in fit. Bearing slipped in easily. Seems to be a little suction when trying to pull them out though. This might be fine enough when I add the cover plates to hold them in.  Two 4mm thick bearings in each side. When the table is assembled there shouldn't be any sideways racking force applied to the pivot connections.  Now to see if my little CNC mill can cut out brass cover plates. 

 

Larger 1/4" shaft binding bolts showed up today. I've already got the 1/4"id bearing that fit them.   Looking for a couple more scraps to make a test pivot connection.  I've got the toolpaths drawn up to cut out the cover plates.  Had to figure out a way to clamp the brass plate down on my small CNC milling machine bed. Now I just need to work up the courage to see if that baby CNC can do the job. Letting it stew in my mind to make sure I didn't forget anything about milling variables for cutting the brass.  Much like I felt the first time I cut anything on my new CNC Shark many years ago.  

Ruined one piece of brass.  Bit apparently slid up in the collet as the first holes weren't cutting through the brass.  Stopped the job, reset the Z axis and tightened up the collet.  Started the job again only to see that the X axis had shifted 1/16" or so to the right and even the new holes in the wrong place weren't cutting through.  Brass plate didn't appear to have shifted although it might have moved a tiny bit with each X/Y circle movement of the bit. 

 

Thinking now about how else to make the cover plates.  Might use my regular CNC with a 60 degree V bit to mark the center of all the holes and score the outlines.  Then use the drill press to take care of the holes before cutting the outline sides using my metal bladed bandsaw with a guide fence.  Then round the ends with a center pin fixture on the disk sander.   Bought 3 of the brass plates. So far ruined one,  Two chances left. 

Giving up on using brass.  Should have remembered from my previous attempt to CNC brass to make a push button depth stop for my drill press that brass has it out for me. 

Next strategy that woke me up today will be to see if I can file a shallow groove around a metal bolt shaft using my variable speed lathe.  Then go back to using a binding bolt to trap the pivot bolt once inserted but let it spin if it wants/needs to.

If I had access to the fab lab I retired from teaching in I would use their metal lathe and likely would have already finished this project using that idea. 

4D  

My old rat tail file (inherited from an uncle) appears to be too worn out for making a groove around a metal bolt.  Just ordered a set of 5mm diamond coated round files from Amazon.  Pondering how I might make a jig to control the file's position and depth of cut over (or under or beside) a spinning bolt shaft.  A couple ideas popped into my head just as I was writing this.  Guess I'll go play in the garage to see what I can rig/jig up.  

Almost forgot I have a radial axis on my CNC.  Drew up a toolpath to make a 2.5mm radius .0625" deep groove around and precisely positioned on the 1/4"d bolt shaft.   In theory the carbide ball end spiral bits I have would be harder than the steel bolt shaft.  Small 4% stepovers using the moulding toolpath around the shaft.  Only thing I don't like about the moulding toolpath for this is that it goes back and forth. Climb cuts in one direction and conventional cuts in the other direction.  Now I just need to get up the nerve to try cutting one.  

Answer is "Yes, you can cut a groove around a metal bolt shaft using a conventional CNC radial axis."   Binding bolt shaft added for clarity.  Should keep the pivot bolt trapped, but allow it to spin. 

Bolt will be cut off about 1/8" down from the groove. Any threads left will be filed down to the shaft so they won't interfere with the bolt sliding into a snug hole.  Now to cut 3 more. 

4D

All four.  Shafts on these were slightly imperfect, but the grooves should be perfectly centered along the axis of the bolt.  When my diamond files arrive tomorrow I'll give the grooves a little kiss with them to smooth them a bit.  

The challenge with having so many tools to use and play with is that the best tool for any job might be one of many options and of course the last one tried. 

4D

Edited March 30, 20242 yr by 4DThinker

Just an observation. With that deep and wide of a groove in the binding bolt and assuming grade hardness is 2 or less, won't the groove significantly compromise the bolt strength and potentially make it react like a shear bolt/pin. Just my opinion, but I see this as a failure point over time.

1 minute ago, Grandpadave52 said:

Just an observation. With that deep and wide of a groove in the binding bolt and assuming grade hardness is 2 or less, won't the groove significantly compromise the bolt strength and potentially make it react like a shear bolt/pin. Just my opinion, but I see this as a failure point over time.

The groove is in the pivot bolt.  1/8" from what will be the end of it. There is still 1/8" of steel in the perimeter groove, which is only there to hook around the binding bolt to keep it from working out..  The shaft through the outer leg and into the inner leg is still just a little less than 1/4" diameter.  That will never shear given how little the table is folded up or flipped inside out between flat and slanted.  This isn't a chair which might have to handle 250lbs or more.  There won't be any racking forces or pulling forces that I foresee to stress the bolts in normal use.  I have some KD hardware that reduces to 1/8" or less with a head on the end to slip into a cam.  Never seen one of those fail either. 

 

Lastly, this is just for me. If in 10 years it fails for some reason I'll learn something, replace or change something, and keep using it.  If a manufacturer wants to use the idea then I'd expect them to consult an engineer to make sure the hardness and design of the  pivot bolt will last the predicted life of the table. 

4D