Screwing Around

I am sorting hundreds of screws and nuts. Twenty years ago, I paid my son to sort all the fasteners that ScopeRoller used, with each drawer labeled by thread, length, and drive type (socket head; Philips, standard). 

Over time, entropy happened because I was often too busy filling orders to return fasteners to the right drawer.

You may be wondering, "Isn't your time worth enough to throw them all away, and buy what you need, when you need them?" 

Right now, I live a short drive from a Tacoma Screw retail store, Ace Hardware, and True Value. I am hoping for a Milky Way sky in the new house. I will have few nearby hardware stores.  At least some of the current hodge-podge was because I could not find screws (especially thumbscrews) that were present in one of dozens of drawers, so i bought another bag retail or from Amazon. 

The sorting process is easier than I expected. The 6-32 and 8-32 screws are easily distinguished visually. Ditto for 8-32 from 10-24 and larger. In some cases, looking in a container of nuts and bolts, you can immediately see that these ten nuts are the same size. Length and finish are also useful. The aluminum 1/4"-20 hex head screws (that I bought to see if I could make ScopeRoller less marring to tripod legs) stood out from finish alone. 

I was thinking last night that a fastener equivalent to CoinStar would be useful. You drop bags of fasteners in the top and it sorts them by diameter and length, rhen attempts to gentlely turn them into threaded holes until it finds a match.  I would have paid $30 to have used such a machine to produce bags of M3x0.6, M4x1.0, 4-40, 6-32 etc. If it could sort them by other criteria (head type, fully threaded, finish), all the better.

Remote Desktop to Linux Works Very Well

XRDP is astonishingly responsive. It feels like I am sitting in front of my Linux desktop in the shop. No lagging. 

Of course Linux is running on a 2015 desktop with 32GB of RAM and an SSD. My PC running remote desktop has 128GB of RAM and the connection is at least 70Mb/sec. so it darn well better be fast.

The attraction of remote desktop is my office has a beautiful view of the marmots running across the lawn and a better chair 

Silly Question

 Let's say you want to cut a piece of aluminum angle in 1.5" segments on a chop saw with a 100 tooth blade. It is rough on blades but cuts well.) I would prefer 1.5" wide -0", +.05". Can anyone suggest a method of doing this? Yes, I should use a band saw but I do not have one right now.

Solution: dig through my scrap pile to find a 1.522" wide piece that I milled. Put it against the blade. Press a piece Delrin against and clamp it in place as a stop against the back fence. Remove 1.522" piece. Push workpiece against stop.

The first two slices after running over the belt sander were 1 517" and 1.509". Taking them down to 1.50" on the mill will be easy  

Remote Desktop to Linux Box in Shop

 Using XRDP, I am able to do all my development from my office now:

Once temperatures rise, this will be much nicer than doing this in the garage.

So I Acquired a 64-Bit PC to Run Newest Version of LinuxCNC

A gal from church was clearing out her fleet of PCs too primitive to run Windows 11 and I told her that I needed something that was 64-bit that could run Linux. It has 32GB of RAM, and several TB of both hard disks and SSD--far in excess of what LinuxCNC needs.

It had no built-in parallel port. LinuxCNC needs a real Centronics parallel port for timing reasons. I bought a PCIe parallel port made by StarTech. LinuxCNC would not control the mill! Searching for guidance gave me only a sensation that the parallel port might have the wrong address. Supergrok saved the day. It helped me find that the built-in parallel poets are at 0x3F8. This rang a bell from when I wrote a terminate-and-stay-resident PC-DOS (stop laughing) program that intercepted data going to the parallel port and converted Epson control codes to Postscript. It was wonderfully elegant in how it allowed programs with no Postscript support to use Apple LaserWriters 

Changing one of the LinuxCNC setup files to talk to the parallel printer at 0xC010 and problem is solved.

Complicating this was that F2 (enter BIOS) and F12 (boot from somewhere besides hard disk did nothing. It turned out that unless your monitor was hooked up to a DisplayPort monitor, they did nothing except assume you wanted to reinstall Windows. So I dug up an HDMI to DisplayPort cable. (I never throw away cables that might be useful in the future. The 1Mbps Ethernet cables were obsolete.)

UPDATE: After installing LinuxCNC and all the development tools that I use (emacs, ddd) and restoring all the software i have written, i decided to install Eclipse. When I worked for Corrections, I used Eclipse for Java development and I really liked it. It is certainly more pleasant and easier to  use than emacs and ddd. But once installed, i was out of disk space. Worse; attempting to remove stuff i did not need somehow permanently locked root so I needed to reinstall.

But instead of repeating a process that had disappointed me, i decided to take advantage of having way more storage than the last desktop. This has a 250GB SSD, a 1TB hard disk and a 2TB hard disk. I put root and all the CNC software on the SSD, and home on the 2TB drive. I will never run out. The SSD still is 33% free.  The home drive has about 1.9TB free.

I had SuperGrok walk me through partitioning. I used to know mostly how to do this, at least at the /etc/fstab level. It was still nice to have a friend walk me through it.

Eclipse was sufficiently difficult before I reinstalled that i am not sure that I will try to use it. All I wanted was to load my existing Makefile and sources into it. Perhaps SuperGrok can help me 

The one reason I upgraded to the 64-bit version of linuxCNC was modern browsers. This makes using SuperGrok and reading X possible. 

UPDATE 2: And I had SuperGrok set up a daily backup of my home directory to a 1 TB hard drive not currently in use.

Advantages of Having a CNC Mill

The part that holds the diagonal mirror in a Newtonian reflector is called a spider. (It has either three or four legs and a body in the middle.)

I replaced the 1960s antique Edmund Scientific spider (which had no collimation screws) a couple years ago with a very nice modern one that uses four 10-32 screws. I have never been very successful at collimation of the main mirror. When trying to diagnose the problem, i noticed that the round part through which the four 10-32 screws pass had a stripped thread.

The screws go through the round part to tip the mirror holder relative to the optical path. Obviously a stripped thread is not going to do that very well.

Fortunately, the round part with stripped screw hole was a pretty easier part to machine. The original was 1.53" diameter, .41" thick with five 10-32 tapped holes (four for adjustment; one to attach it to tge mirror holder). I had a piece of .5" thick Delrin which was easy to machine into a copy.

Once I put the new part in the spider, I discovered another nuisance: the torque to the adjustment screws was enough to repeatedly rotate the mirror holder so the laser beam would go everywhere it should except the center of the main mirror.  The only solutions would be to add a lock nut to the mirror holder where it connects to the shaft or figure out how to make the adjustment screws zero friction through the center piece. I already had it in the tube, so i adjusted for center on the mirror,:rotating the mirror holder back each time to get back to mirror center.  The final collimation step only requires turning wing nuts on the mirror cell. If I have occasion to remove this again, i may tackle one of these fixes. For now, the mirror is collimated.

Have You Ever Wanted to Work With 21st Century Material?

 I mentioned that my wife put the kibosh on carbon fiber composite machining. The dust gets everywhere and it is not the healthiest stuff to have blowing around, even if you are wearing a respirator. Before I put these pieces on eBay, I though I would all of you a chance to buy these pieces. You will be amazed at its lightness, stiffness, and tensile strength. I do not really have a price for any of these. If you can pay me for postage, that would be fine. These are all Dragonplate Economy Plate CFC:

Random scrap, mostly 1/2" and 1/4" thick. I have no idea for what these little scraps might be used, but they are amazingly stiff and strong.

2 6x6x1/8" glossy:

2x6 x 1/2 matte:

3 1/2 x 6 x1/8 glossy

6 x 2 1/4 x 1/4 glossy

6 x 2 x 1/8 glossy

I have no idea (my notes do not match the picture sequence)

1 1/2 x 2 3/4 x 1/4 matte

Today's Machining Lesson

The attempt to cut the slice went poorly because the direction of motion and velocity unscrewed the chuck from the rotary table. It was not enough to come off but enough that the tube moved away from the endmill. Solution: cut in a clockwise direction so it does not loosen.

I Was Not Successful at Using My Parting Tool

I wanted to cut a piece of aluminum tube down to 1.23" length but I could not find a feed and speed that parted it without pulling the tube out of the chuck. I thought of using the chop saw but I feared it would bend the fairly thin wall tubing while cutting it.

So I mounted it on the rotating table.

I needed to make very slow and shallow cuts because again the chuck is unlikely to hold the tube as tightly as it should. But why write a program specific to this operation? So I wrote a program called mksub that accepts zStart, zStep, and zEnd parameters as well as a string of gCode commands to execute at each step of Z. In this case, a move to degree 0, then to degree 360, and back to 0.

This is general enough that almost relatively simple gCode command can be repeated to any arbitrary depth with only a trivial change in the Makefile.

There Is No Substitute for Sharp Tools

That includes drill bits, saw blades, end mills, and today's subject, boring bars. I am boring an aluminum cylinder and cutting was going poorly. I looked carefully at the carbide tip and the problem was obvious. Fortunately, this is a rotatable carbide tip. To do that you need a teeny tiny Tor-X (T7) which is Alford slightly larger than this letter o on a 13" screen. (Do you remember 13" computer monitors?)

It bores so much better now.

Handle for Televue-85

Once I stopped fighting the inability to mill a thin piece of aluminum, it was pretty easy. I cut a .5" slice of 4x2 rectangular aluminum tube, drilled two  201" through holes 1.250" center-to-center, went to Tacoma Screw for two stainless steel 10-32 x 3/8" socket head screws and just screwed it in place using an Allen wrench. 

It is easy to pick up and hold in position with one hand while using the other hand to tighten the bolts that hold it in the Vixen saddle.

I still have enough rectangular tube to make probably 30 of these. There is very little work involved.

1. Slice a .5" piece on the chop saw using a clamped stop to get consistent slices.

2. Put it in the mill. Run a program that uses a center drill to make the through holes. (On first article, I used a center drill to make a pilot hole, then a .2010 twist drill, but I can use a center drill to make hole in one operation.

3. Sand all external surfaces and cut edges.

4. Use file to break all corners to make sure it will not cut skin.

I could do these assembly line in about 3 minutes each. The rectangular aluminum tube is a sunk cost. The only marginal costs are the 10-32 screws and my labor. If I can find buyers on CloudyNights, I might do a production run.

I was using this evening. What a change from grasping it with one hand!

Painful Learning Experience

Painful only on time. I tried to get around that problem that my mill vise could not hold the vise in an orientation for side milling the workpiece to .620" so I put so that I could face mill it to that width...

That should have been easy: slice .001" per pass at low speed.  But instead cutting on the 1/8" wall just bounced everything around enough to knock it out of the vise. So I went back to the chop saw and worked very had to make a slice. 625". I got .6".  Close enough.

As i was writing this is realized another method that I did not try but I might try tomorrow just to see if it works. Put the 4" wide piece in the vise with half of it above jaws. Side mill to width. Turn it upside down. Repeat.

If I do this again, I will use 1/4" wall.

That approach also did not work. Trying to mill such thin material just sets the material to vibrating, which simply shakes it loose from the vise and causes the vise to shake loose from the clamps to the mill table.

The 4"x 2" Aluminum Rectangular Tube Arrived

I should have considered this in advance " my mill vise is 3.8" wide at maximum opening. If i had a 4"+ long endmill I could just put the 4" side vertical and side mill it from the roughly 0.8" slice to 0.62". Endmills that long are expensive! (I should have ordered 3.5" x 2". This would have fit better to the ring anyway.)

This is not a Kurt vise or a Kurt clone so the jaws are not reversible.

Trying to use hold downs to lock the workpiece to the mill table did not work. You need something under the workpiece to sacrifice when the endmill reaches the bottom. I foolishly tried to use Delrin for this. Aluminum on Delrin is very slippery and you cannot tighten these hold downs down enough to hold them.

Supergrok suggested gluing the workpiece on something that will fit in the vise. I found a 3" piece of aluminum and Gorilla Glued the workpiece to it. Tomorrow it should be solid enough to try and mill it.

The Gorilla Glue was not strong enough. So I went back to my first approach face milling off the excess. Perhaps because the attempt at side milling evened up the surface enough, it seems to be cutting better

Handling Something Scary

 One of the reasons that I do not pull my Televue-85 out and use it as often as I would like is that it is a very expensive telescope.  To put it on the mount requires holding the tube with one hand while putting it in the dovetail while using the other hand to tighten the clamping screw.

As I( was putting it on the mount last night I realized that I have a solution staring at me. This is the 3" Televue clamshell ring:

This ring both simplifies positioning it on a mount and attaching various finder attachments. Those are 10-32 screw holes. After a long useful chat with SuperGrok attempting to find a furniture handle with the right dimensions, I realized the solution is to buy a 4" x 2" x1/8" wall rectangular aluminum tube. Cut a .625" slice, then drill through holes 1.25" apart in one side. Sand everything smooth run it over the polishing wheel, and use some stainless steel socket head screws to lock it down. The hole will fit my hand comfortably and I will be far less afraid to mount it with a really solid grip on it.

I can convert the rest of that 12" tube into 15 more handles for sale. There is almost no labor involved. Cut the slice, mill to .65" wide. Put in mill and drill two holes in one side, sand with 320 grit, then polish. I suspect that I will find vuyers.

If Godzilla Owned a Machine Shop

 I have a benchtop lathe:

Here is a big lathe:

My Fingers Are Grateful

I bought a parting tool for the lathe.  This is a very sharp thin blade that mounts on the cross-slide for cutting off pieces of the victim at precise locations. 

In the past, if i need to cut a small amount off of a tube, I had to either hold the tube closer to the blade than I preferred or if the amount was small enough, use the lathe to slowly plane off a few thousandths of an inch, again and again.

In this case, I needed to remove 1.85" from the PVC tube.  It was so elegant. 

For You Materials Science Nerds

 Why 304 stainless steel is less corrosion resistant than you may assume:

And if, like me, you have ever wondered what distinguishes 304 from 18/8 stainless steel.

What AI Is Good For

Define the problem.

For a 6" diameter 6061 aluminum disc supported at three edges with a 20 pound weight at center how thick should it be?

After showing how it calculated it. Grok said 0.04" then said triple that to avoid any problems with the various ways to interpret support positions. This is really quite a bit thinner than I expected. I had planned on 1/8" thick for CFC.

FreeCAD Tutor

Grok has been helping me figure out how to replace CFC with aluminum tube with minimal weight increase. It appears that .5" OD .035" wall is slightly stiffer than the CFC tube that I now use for the axes. This lets me reuse my existing polar and declination housings, bearings, and drive gears.

I have been attempting to learn to use FreeCAD and no tutorial seems to get me over the first step of converting a two dimensional drawing into a 3D object.  I know at least one of you has admitted FreeCAD experience. I would be quite happy to pay someone to use Remote Desktop Connection and help me past this obstacle.

UPDATE: Just realized that I do not need to replace existing housing and axes  They are already cut 

Turning Workpieces That Are Too Short for Chuck

I am about to make handles for turning the slow motion handles. I made a couple of these a while back but they were not attractive being sort of a Greek cross made from 1/2" CFC.

My plan was to cut two 1" circles with a .25" hole in the center where the D-shaft. While clearing around the mill, I found a 1" (actually. 951") diameter CFC piece that fell out when cutting a 1.25" hole. Perhaps, I could reuse this piece? 

So I mounted it in the lathe and drilled a .25" hole through the center. 

But there was some crud on one end that needed turning off.  But the jaws were not deep enough to hold a 1/2" thick workpiece with the end exposed.  So I dug through my screws and scraps and made this up.

The white acetal locks in the jaws.  A 1/4-20 screw passed through it, locked by a nut, and then the workpiece, locked in place by another nut. Now the workpiece was entirely outside the jaws where I could turn the exterior.