Sort of 4th Axis

Rotary control of 4th axis on a mill to save time.

Similar projects worth following
I need to drill lots of holes. Radial holes, 6 to a part, 4 parts per an order. The first run is 500 parts, 3000 holes. I prefer not to clamp and reclamp each part 6 times to get a hole every 60 degrees. I've got extra Arduinos and a NEMA34 stepper from eBay that I never figured out how to use...

There's got to be a way to make it so I hit a button and my part spins 60 degrees so I only have to clamp/unclamp once for each part to be finished.

Prototyping began with a typical Arduino Uno and once the design finalized, I chose to use the Pro Mini which saves space. Also, the Rugged Motor Driver shield has mods from previous projects (reassigning pins), so it was easiest to use the driver board in standalone mode via the green 6 way terminal block.

When supplying the Arduino from USB power, the sketch and driver behavior was different than when only running off the 5V input coming from the power supply I scavenged from a laser printer (details in build log). It irks me when this happens as my projects tend to be stand-alone and I like to power them from typical 5V supplies. So I desoldered the fuse next to the Uno's USB jack. Now the Arduino is fed from the 5V I want to give it and behaves properly when troubleshooting on serial over usb.

None of the parts in this project get hot, so I was able to box everything up nicely without cooling concerns. The Rugged shield's two driver chips eventually become warm to the touch. When trying the other driver hardware, quite a bit of heat would be generated. The stepper never got warm, so it's up to the task.

Once the electrics and control were proven, I worked on the mechanical design. There was some iteration; I needed to see that the stepper could hold the shaft firmly enough that I could use an end-mill on its aluminum arbor without the shaft moving. Only the Rugged shield was able to accomplish this.

The arbor fixed to the stepper has a hex head that wedges into the parts we will be machining. Read the mechanicals entry of the build log for details, but the parts to the right firmly clamp the part to be machined to the stepper's arbor while allowing the part to rotate with the arbor.

Note that this isn't "NC" as the stepper is not moving against the cutting tool, it is used to automate part positioning between machining operations, which will be performed manually by the operator.

We finally got in the parts and I was able to try out this tool -- it worked great! I expected the pieces to need some adjustments or repositioning. I might not have tightened the screws enough on the outer cylinder and side plate, so as they worked loose, I noticed these pieces shifting when I clamped and unclamped parts. Also contributing to the screws loosening could be if/when the drill chatters. This makes pretty strong vibrations. I ended up running the spindle at 2400 rpm and when the drilling is started without hesitation, there is no chatter. The "drill bit" is actually a Size 5 countersink.

Got through the first hundred parts in under 2 hours. Not seen in the video, the parts coming out of the tool are well over 100 degrees, so I added air cooling pointing where the countersink hits the part. After turning up the spindle to 2400 rpm, I got into a good rhythm and was cranking out the parts much quicker than in the video. I used the yellow spray can (Maxima Multi Purpose Penetrant Lubricant - MPPL) between each part, sprayed at the countersink as I started the first cut. The countersink is double ended and should get me through the remaining 400 parts.

View all 8 components

  • It totally works!

    smashedagainst06/23/2015 at 20:32 0 comments

    I received the parts to be machined and everything went better than I thought possible. I only ruined one part and that was because I pressed the button while the drill was still in the hole, preventing the stepper from turning. I was lifting the spindle, so the stepper turned only a portion of the 60 degrees. I discovered the mistake when the last hole ended up too close to the first one.

    I made a quick drawing of the tooling in the vice:

    And the firmware for the Arduino is in this Github Gist. It's really only the sample code from Rugged Circuits changed slightly for serial debugging and with values appropriate for this stepper and the 24V supply.

  • Boxing it up

    smashedagainst06/04/2015 at 19:11 0 comments

    I really like the design of the Samsung cheap laser printer we bought for front office's desk. It lasted longer than expected and when it eventually died, I took it apart to keep all the guts. The back of this PSU has a legend printed on the PCB of the connections with voltage and current. CON1 is the 115vac input. CON3 had 24V at 1.6 A and 5V at some other amps. CON3 is the small gray connector with lots of pins at the end away from the 115 input connector. CON2 is the white connector and supplies 115VAC to the high voltage power board.

    I really appreciate how this PSU came with switch and IEC socket. I cut (well, machined) slots in the lower half of the case to hold these. The IEC's socket was designed for thin sheet metal, so I had to cut back the locking tabs and bend them out some to secure the socket in the case.

    Red wire is soldered to the large pad underneath feeding 5V to the gray CON3. Green is soldered to the 24V pad. The black wires are soldered to the ground plane surrounding the DC section. This ground does not have continuity to the metal heatsinks nor either pin of the input 115V.

    I placed clear acrylic plate above the PSU to keep my wires and the various 115V lines and connections away from the PSU's internals.

    This highlights what I consider one of the bigger obstacles to small projects: connections between modules. I really like the 4 way screw terminal connection block. It has no exposed metal that can touch other bits and accommodates all wire sizes I use in projects. The output is trailer wiring, which is cable tied to the bottom of the case's battery holder. The psu uses a "trailer connector" and my project has the plug normally at the back of a car. While backwards-ish, I had only 1 "trailer" end and 2 "tow vehicle" ends, so thought this would allow the PSU to be used again.

    Plus, the tow-side connector is available at many local stores, so it's easy to buy more. Second advantage to using the trailer wiring is how durable they are made; I can step on these wires without fear of damaging them.

  • Mechanicals

    smashedagainst06/04/2015 at 07:42 0 comments

    Using the drawings from the NEMA 34's data sheet, I threaded M6 holes in the base plate to attach the motor. At the other end of the plate is a "design" that applies pressure against the end of the stepper motor's shaft. The clamp's purpose is to allow the easy removal and replacement of the part to receive machining.

    The big round thing is just a cylinder to hold the central shaft and the clamp. The central shaft has a bearing at the CNC end. This allows pressure along the shaft and also rotation about the shaft. The other end of the big round thing is a simple plate that holds the Harbor Freight toggle clamp.

    The toggle clamp acts on a M6 x 50 socket head cap screw that's threaded into the right end of the central shaft. The springs on the right end of the bolt retract the device when the toggle is opened to allow easier change of the part to be machined.

    Affixed to the stepper's shaft is an arbor I made. The items to be machined have a hex recess broached into them, so the arbor firmly holds the part to the stepper. There is a picture of me machining the arbor while it is in a hexagonal collet block. The other end of the shaft was bored to 14 mm to make a snug fit on the NEMA's shaft and a set screw clamps the arbor to a flat on the shaft.

  • Best motor driver for my 24 volts

    smashedagainst06/04/2015 at 06:49 0 comments

    SainSmart CNC controller:

    I went and bought a SainSmart CNC controller. It's all professional and serious, so I thought it would be a great start. I watched various youtube videos of the module to better understand this device and gained some confidence so tried it out. I worked through the various switch settings and arrived at what I believed to be the setting with the best holding power and I was still able to turn the stepper's shaft with my fingers. It was quite strong, but not enough.

    To make sure it wasn't the power supply's fault, I bought a lab power supply off Amazon and hooked it up. It read 24V and max current draw was just under 1 amp. With this "bench" power supply, I could still turn the shaft with my fingers.

    Rugged Motor Driver:

    Next up is the "Rugged Motor Driver" from rugged circuits. I was able to get this shield to work with the stepper motor. I was eventually able to coax the sample sketch to manipulate the stepper into doing things I want. And then I was able to gain confidence in what was happening and I was happy. Happy-ish, anyway, but the stepper would "sing" loudly. Everything seemed ok otherwise, as the stepper remained cool to the touch and the driver chips on the shield were only slightly warm to the touch.

    Big Easy Driver:

    I purchased one of these for a previous project. The small easy driver was super easy to use for the big one should be, too. What could possibly go wrong? Oh, the smoke. It fell out of the big easy driver. I hooked it up the same as the other two boards prior (and the small driver with 12V prior) and then zzzzzzzzzip pop smoke.

    OK so Rugger Motor Driver it is...

  • 12V isn't enough...

    smashedagainst05/28/2015 at 04:10 0 comments

    12V with lots of amps (4 or 5 from power supply's label) isn't enough to hold the spindle in place while machining. Going to try a 24V 1.8A supply next.

  • Electrics

    smashedagainst05/27/2015 at 04:37 0 comments

    I used the "spare" parts left over from the TPS test bench's Arduino stage as the quick start to this project. That's a proto shield with a SparkFun Easy Motor Driver soldered to it. I found an Arduino Uno and a SparkLite project case and spacer.

    I also reused the code from the TPS bench and worked up code to alter the magic numbers from the stepper code there to work with the huge NEMA34. ( hmmm, how do we share code here? )

    I did fab up some aluminum to make "the tooling" and I'll share that once it's proven working. Which next brings up the question of how do we share CAD work here? :D

View all 6 project logs

Enjoy this project?



Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates