Strutt Epicyclic Train Clock (maybe?)

[kvom]

The first part of the machining encompasses almost the entire outline, leaving less than 2" of the top to finish on run 2.  Since the work can only overlap the front part of the bed, I need to flip and reverse the plate to finish it.  I plan to use the two upper pillar holes as locators.  I'll pre-drill them and also drill and tap holes in a backing plate, allowing me to use threaded dowel pins for mounting the stock.  The backing plate will have two sets offset by a suitable amount in Y. Y0 for the CNC program will be the center of a dial pin located with my soon-to-be-working probe.

[crueby]

The first part of the machining encompasses almost the entire outline, leaving less than 2" of the top to finish on run 2.  Since the work can only overlap the front part of the bed, I need to flip and reverse the plate to finish it.  I plan to use the two upper pillar holes as locators.  I'll pre-drill them and also drill and tap holes in a backing plate, allowing me to use threaded dowel pins for mounting the stock.  The backing plate will have two sets offset by a suitable amount in Y. Y0 for the CNC program will be the center of a dial pin located with my soon-to-be-working probe.

Sounds great!   

[kvom]

Today's adventure was cutting the two 12x12 brass plates to get 9x12 workpieces.  While cutting on the bandsaw might have worked, I decided part them on the Bridgeport using a .108" carbide endmill.  I was able to position the plates so that the parting line was over the front t-slot.  Since I have a drive on the X-axis, I didn't have to turn the cranks.  The recommended cut from G-Wizard was 5 inches per minute at .01" depth of cut, 2000 RPM.  That meant 20 passes.  The drive is variable speed and my DRO shows feedrate, so it was just a matter of raising the knee after each pass.  Time consuming, but the endmill survived.  It's 4-flute, where 3-flute would be better got brass or aluminum.

I positioned the sample acrylic plate onto the 9x12 acrylic sheet, and marked the approximate positions of the upper pillar holes.  This would be the setup for the first pass.  Then I reversed it and placed the pillar holes over the same mark to show the second pass setup.  By reusing the same hole positions, X0Y0 will remain the same between passes.

[kvom]

Change of direction.  I found a small 2-person company with a waterjet that can cut the plates for me at an affordable price.  I'm going to have them do the ring gear as well.  Going to see them on Saturday.

[crueby]

Neat!  Does the waterjet have the accuracy to do a good job on the gear teeth?  I've only had a piece cut once, a weathervane, where accuracy was not critical so I didn't look into that aspect.

[kvom]

Neat!  Does the waterjet have the accuracy to do a good job on the gear teeth?  I've only had a piece cut once, a weathervane, where accuracy was not critical so I didn't look into that aspect.

From their site: "Our Water Jet is capable of attaining accuracy down to .0005 and repeatability down to .001 inches"

[Jasonb]

I did find that you can get a slightly tapered cut, although I think some of the latest machines can tilt the jet to compensate. Finish is similar to sand blasting so will need cleaning up, again the people I use offer about 5 different grades of finish. You pay more for the fine as the machine has to run for a much longer time. I think this was #3 so middle of the range, material is 5mm and 8mm thick

[kvom]

That's good to know and look out for.  None of the edges on the plates are critical, and the pillar holes won't be seen. I'm going to reduce the size of the hole for the center arbor in the DXF, planning to ream it later.   Rough edges on gear teeth is likely not a good thing.

[kvom]

I was all set to cut out the body of the depthing tool when the CNC mill's spindle refused to start.     So I added it to the list of parts to be cut out by the water jet.

I drove 90 minutes to the shop to have the parts cut, but the waterjet machine was refusing to communicate with its controller.  So I left the brass with them, and I'll either go back up or they'll ship me the parts and the remaining stock when finished.

The couple who have the shop just opened it in December as a retirement business.  The man is a career waterjet pro who wanted to stop traveling full-time doing training and installs, while his wife does the books.  They also have a large CO2 later and a large Laguna router for other custom work.

Some info on the waterjet:  kerf is .035".  It can cut very thick material (up to 10-11").  Speed/finish can be set from 1 to 5, with 5 being the best.  He said that for thin brass like I have there wouldn't be a huge difference. 

Parts to be cut in addition to the depthing body in 3/16" sheet are the 2 main plates, the barrel covers, ratchet wheel, and the great wheel.  The ring gear is from 3/32" sheet.

[kvom]

Some work on the spring barrel.  Cut tube to length, face the edges, then turn and bore to clean the inside and outside.  Then drill and tap 10-32 for the catch screw.

[kvom]

After some tribulations, I've managed to get my probe installed on the DMC2 Mini mill.  As can be seen in the attached photo, the probe has an interior green LED that lights when power is applied.  When the probe tip becomes grounded the light turns red.  So even without using Mach3 probing the probe can be used as an edge finder.  The tip is a 2mm tungsten ball, so the edge would be 1mm outside.

The connection is a USB cable containing 3 conductors.  Two are for power that can be 5-24V DC.  The third is connected to the Mach3 card and is normally an open circuit.  When the probe touches, the circuit closes (e,g. resistance drops to 0 ohms).  Power is provided by a two wire connector from the board supplying 12V.

I'm hopeful this is the last time I need to mess with the wiring.   

I still need to install and test the Basic routines in Mach3 to enable actual probing.

Because the spindle is ER20, it's possible that the probe might vary in concentricity between separate mounts.  I'll do a test of the runout as the probe body has adjustment screws, but I expect that any difference will be less than the expected precision of the mill.