Learning by doing is the only justification for making your own clamps like this. I probably have eight hours tied up in this one, and you can buy one in much more durable steel for around 20 bucks. But learning was my goal, and I did learn!
The project's genesis was a perfectly knurled thumbscrew I made for my finger plate clamp tool (link). I'll repeat that photo here:
I had noted in that blog post that it was my best ever, but what I didn't admit is that it was pure luck - my knurls were usually not so good. But there was indeed learning - I learned that my equipment was capable of making a perfect knurl. I set out to figure out how to do it consistently.
My initial trials were terrible. I was having problems getting the lathe to do anything right. I finally disassembled the compound (the top part that holds the cutters) and found Lesson No. 1: Make Sure Your Equipment Is In Good Repair. This photo is of a very important bolt, one that holds the compound secure, and that allows lots of variability in results if it's not tight.
Obviously, that bolt is stripped almost completely at the end. But it was the "almost" that got me. See the black line right at the stripped end? That's a partial remaining thread that held the nut well enough to make it look like all was OK. But it wasn't - there was enough play to let the compound rock back and forth.
Once I fixed that, I was rewarded with clean cuts on the lathe. Time to understand a concept I had heard of, but never internalized. It's Lesson 2: Know the Right Diameter to Knurl. If you look back at that perfect thumbscrew, you'll see that a knurl is a diamond pattern that repeats around the face of the rod. That circumference must be an exact multiple of the width of the knurl's diamond, or things won't come out even. It is the mismatched diamonds that made my imperfect knurls previously.
There are charts and spreadsheets to help you figure out the correct diameter, but there's an even easier option - a free phone app called "Knurling Calculator." You have to know two bits of information: the "threads per inch" value of your knurl (mine is 33) and the current diameter of your workpiece. You enter that, and the app tells you the next closest even diameter. In this example, my rod was at 0.475", and if I just removed a tiny bit to get to 0.473 (I don't even pretend to work to four places), I'd get a perfect knurl.
That is a necessary condition, but not sufficient for success. You also need Lesson 3: Knurl At Your Lathe's Slowest Speed. On my lathe, that means opening the side, loosening the belt tension at the idler, and moving the belt around:
And of course, Lesson 4: Set Up Your Knurling Wheels Correctly. I use a "squeeze" type knurler that mounts on the quick change tool post. The technique that worked for me is to lightly clamp, and briefly cycle the power on the lathe. A light knurl will print. If it looks good (and it always does, because the light pressure allows the wheels to align), then tighten the wheels as tight as you can by hand, turn on the lathe, and use the power feed to advance the wheels for the desired distance.
Once I had all that figured out, my knurls were perfect every time. I got to work on my little clamp, making two knobs that were threaded at the end, adding some threaded rod liberated from some long brass screws, and very carefully making a brass "keeper" to mesh with one of the knobs.
That keeper was tricky to make. Its task in life is to move the clamp arm open when the screw is retracted. It turns out in this first attempt that I got it a little too thin, so you can see the small spacer I had to add underneath to keep things from binding:
The clamp body was mostly straightforward milling, drilling and tapping. Brass is expensive and I'm cheap, so I squeaked both pieces from a single offcut from a larger bar:
Look to the top picture to see the result - it works great! I've already used it for real work.
After the clamp was done, I continued to experiment and tried a new way of holding the tap while threading holes on the lathe. The problem was that, when I simply chucked the tap in the tailstock, it would sometimes slip. I got the idea of using a slip-resistant adjustable Tap Socket (link) to hold the tap, and then use a hex drive socket adapter in the tailstock.
Since all the surfaces being clamped have flat surfaces, they are clamped tight and nothing can slip. Seeing an exploded view will probably help see what I mean:
I tapped the hole, and then used my new bottom tap to carry the threads all the way to the bottom of the hole. Since the threads were already half an inch deep, I could easily complete that operation by hand, using a socket wrench to turn the tap.
Phil Oles taught me about these very inexpensive brushes from Harbor Freight (link) to clean out the debris from the tapped hole.
Even with careful brushing, a Q-tip will usually get a little more:
Ready to test! I confidently threaded the knob on a brass rod and... what the heck? It's crooked!
I figured that chain of adapters must have allowed me to tap the hole off-center, so I made another test piece with the taps directly in the tailstock chuck, using both regular and bottom taps as before. I was surprised that the second test was also crooked. It was then I realized that I could actually rock the knob back and forth on the threads. It was really loose!
That's when I learned Lesson 5: Know What You're Buying. When I bought that bottom tap, I just picked one out of the catalog. It was only later that I learned that taps have different tolerances. I check the printing on the side, and mine was "H3" - one with a good bit of slop. I'll have to buy again to get an H2, which is more suitable for general use.
I set out to learn, and learn I did!
No comments:
Post a Comment