For my mastery piece I decided to build a hand-held polar CNC machine called **The Rocket Surgeon**. You can read about it and it's expectations in Week 18.
_**[A MESSAGE FROM A FUTURE CARL]**_
####"Everything will be alright in the end. If it's not alright, then it is not yet the end."
#####- Indian Proverb
That's what I'd tell myself if I could go back in time. The journey has been crazy for me personally, with emotional and financial issues. But it has certainly been an amazing adventure: my whole life has changed, I finally went to Europe, I found a better girlfriend, I met some genuinely incredible people and I also managed to fit in some time to build cool stuff too.
So if you're reading this looking forward in time, just think about what your future self might be able to tell you when this is all over. You will not regret pushing yourself harder and harder towards the finish line!
###What went well
It was a big project, so of course there was a lot of success and failure. Things that went well: I managed to include many of the digital fabrication processes and gained a very dramatic amount of insight into various new methods and even more in-depth understanding of methods I was already familiar with.
I managed to get all of the individual units functioning in isolation before the final assembly. I had three steppers running simultaneously, the BLDC spindle control working fine, power management and heat management under control. I also managed to arrange the packaging reasonably dense (given the time I had left), and the 3D printed chassis came out fitting comfortably within the hands with an acceptable weight distribution given the stepper motors were all in the top rail.
The unit was reasonably simple to service, and the 3D printed bearing/collar worked adequately for stiffness - I was able to pick up the device by the rotating shaft which was quite satisfying.
The speeds for the stepper motors were almost perfect, so there was no pressure to implement microstepping last minute, the gearbox ratio and lead-screw ratios happened to be perfect by luck. Also the joy-stick control was surprisingly straight forward, no noise, drift or other thresholding problems. I was quite satisfied simply controlling the steppers by joystick control!
###What went wrong
We only had one 3D printer functional in that final week, so it was shared between Bára, Birita and myself on an ongoing basis. Luckily, Bergþóra was in France in that week so she had the machine all to herself! But there's no doubt that one more person trying to use that machine would have tempted Murphy's Law. We were lucky that no prints failed catastrophically, but getting things done on time was really stressful.
Due to my efficient (lean) mechanical design, the PLA parts were only just strong enough to handle operating loads. However, some instances where I had to press-fit or drill out a bolt hole to clearance, the parts twisted and I needed to re-print them. Unfortunately the PLA gearbox didn't last very long on the stepper motor shaft, and I ended up having to hot-glue weld the shaft onto the pinion gear.
I also had a few screw-ups on the electronics side: One of the headers for my stepper motor board (the main board featured in Week 12) popped off the substrate, and I had to hot glue it back in place.
Another horrible last minute thing that happened was that my power meter was wired in correctly but not labelled - the current sense leads have red and black wires coming from them, where red connects to the negative of the circuit to be measured and the black lead connects to ground.
Because of this red/black problem I ended up blowing the 328P and potentially some other components on that board. I replaced the 328P chip last minute, but the board had some other problems and it was too late in the game to go chasing the remaining bugs. I reverted to the previous board I had machined for the final presentation, where only two stepper motor drivers were available. At least I was able to show two motors in operation.
Lastly, and I only realised this as I went to assemble the main body, the spindle control pins I had exposed were using XTAL pins on the 328P which aren't accessible from within the Arduino IDE. This was a bummer (but not a deal-breaker) so I wasn't able to get the spindle to run for the demo. (Though I had tested it comprehensively during unit tests).
###Bill of materials
Updated from when originally posted in Week 18.
|Parts |Source or Self made? |Cost(AUD)|Process(es) |Additional: |Design Files: |
|--- |--- |--- |--- |--- |--- |
|Spindle |Huaqiangbei Markets |$16 |Market haggling |400W | |
|ESC |Huaqiangbei Markets |$18 |Pretending I was gonna buy more |Mystery brand, 60A running at 12V | |
|Milling bit |Carbide Depot |$3 |Finding it |1/8th end mill, though not important | |
|Stepper Motors |Jameco |$57 | |Qty:3 (4.2V 1500mA bipolar) | |
|Stepper Driver |Carl-brand |$7 |Circuit-milling |6x A4953 drivers |Output Week |
|Lead-screws |Found it |$5 | |Came with threaded slides | |
|Rail-tube |Found it |$7 |Hand-machining |Sanding them clean and chamfering ends | |
|Enclosure |Carl-brand |$40 |3D-printing |Was waiting on Nylon that never arrived so used PLA on Ultimaker 2 Extended |Project Files |
|Mounting hardware |Carl-brand |$3 |Laser-machining |Used up a bunch of scrap from the scrap box |Project Files |
|Control boards |Carl-brand |$8 |Circuit-milling |ATmega328P, 5V regulator, 1206 LEDs, Caps and Resistors as per Eagle |Output Week |
|Shaft couplers |Carl-brand |$12 |Salvaging skills | | |
|Primary head bearing|Carl-brand |$2 |3D-printing |6x skate bearings used (22x7mm, 8mm hole) |Project Files |
|Gearbox |Kevin Lee, open-source |$0 |3D-printing | |Kevin's site |
***TOTAL: $178 ***
*(Yeah, I know.. Whoops!)*
Code for manual control of the motors at the time of demonstration are available at Project Files.
You can also find the full Rocket Surgeon assembly on the Fusion360 share link.