Shapeoko 3 Dropped Connection

I received my Shapeoko 3 in June of 2015 and couldn’t wait to get it up and running in my shop. The design is solid, much stiffer to the X-Carve and Shapeoko 2. The machine gave me the ability to quickly and accurately cut wood, aluminum, brass, circuit boards, and plastics.

Starting out running simple jobs that didn’t require any tool changes I began to notice that the connection would drop upon turning off the DWP611 router that I was using as a spindle. This would result in a lost zero, and would occasionally drop the spindle a couple inches letting the cutting tool gouge the work. It made running complex jobs requiring tool changes nearly impossible.

I contacted carbide 3d and they recommended adding ferrite beads to all the cables. I ordered a USB cable with ferrite chokes build in from amazon and a bag of 7mm clip on ferrite chokes for the router and any other cables that needed them.

I installed everything, and the problems persisted. I continued by trying just about everything short of buying a big universal power supply.

  • Plugging the router into a separate circuit in my house.
  • Adding chokes to the stepper motor wires.
  • Adding multiple chokes to all wires.
  • Removing the controller board from the machine and moving it behind a piece of sheet metal as a shield.

After these solutions failed, I sent several emails to Carbide3d that all went unanswered.

Finally, I landed on a post in the Shapeoko forums where a user said he had added a shielded three conductor extension cord to his spindle. Unfortunately he didn’t provide any instructions (so hopefully this will help those who are hesitant).

THIS INVOLVES WORKING WITH 120V AC POWER. INCORRECT WIRING CAN KILL YOU! IF YOU’RE NOT SURE EXACTLY WHAT YOU’RE DOING GET SOMEONE WHO DOES TO HELP.

I dug an old PC power cord out of a cabinet and went to work. Before I made anything permanent I mocked everything up. The router uses crimp on spade connectors, and luckily I had a couple from a previous project. I did have to bend the power wire connector on the switch 90 degrees to fit it into the location of the original. Once finished I power cycled the router and…NOTHING. No connection drops!

Mocked Up and Ready to Test
Mocked Up and Ready to Test
Crimp on Connectors tying in Power, Neutral, and Ground.
Crimp on Connectors tying in Power, Neutral, and Ground.

After everything was working, I tidied up the wiring I bit. I had to slit the router casing using a Dremel to allow the ground spade connector to sandwich between the router cap and metal body. Once lined up I used a dab of hot glue on the back of the connector to hold it in place.

Slot in router cap to allow terminal to clamp up on router body.
Slot in router cap to allow terminal to clamp up on router body.

I located the switch back in its original location, routed the wires in a manner that they wouldn’t be pinched, and reinstalled the top cover. Once the cover was installed, I metered continuity between the ground prong and router body to ensure the terminal hadn’t shifted out of place.

Done, ready to re-install the top cover.
Done, ready to re-install the top cover.

I’m still using a ferrite bead on the power cable, and the USB cable with built in chokes. Since the mod (~4 months) I’ve only had one or two dropouts, and at least one was caused by static from my shop-vac hose when vacuuming up Kydex (plastic) chips during a job. Compared to the drops every-other power cycle I consider this a success.

Robotics Season is Officially Open

wabbitseason

As of Saturday 1/9 FIRST Robotics season has officially begun. Over the next six weeks the team of high school age kids I mentor will attempt to build a 120lb robot capable of traversing medieval castle defenses, throw boulders at castle ramparts, and lift itself roughly three feet off the ground. This years game is FIRST STRONGHOLD:

I’m going to try and post weekly updates as the build season progresses and my sanity diminishes.

HobbyKing Mini Composite DLG Build

Well, HobbyKing managed to suck me in with their Black Friday sale this year. I’ve been teasing myself with the possibility of a Discus Launch Glider (DLG) for the past year, but kept shelving the idea because of the initial cost (even the budget-minded Libelle will be ~$200.00 to get in the air), so when I saw the 2 channel HK Composite DLG ARF on sale for $39.00 I had to jump at the opportunity.

All Dressed Up
All Dressed Up

Parts List for this Build:
-HK Mini Composite DLG: $39.25 (on sale)
-2S 300MaH nano-tech: $5.95
-Orange R615X Receiver (Overkill, but I had one on hand): ~$10.00
-Assorted JST connectors: $1.11
-Home-Brew BEC: ~$1.00 in parts

Orange R615X, LM7805 BEC, Turnigh Nano Tech 2s300MaH Batt.
Orange R615X, LM7805 BEC, Turnigy Nano Tech 2s300MaH Batt.

The glider showed up a week later and overall the construction impressed me. Covering was tight, wooden joints appeared to have sufficient glue, and all the aerodynamic surfaces were straight and square. The tail boom appears to be unidirectional wrapped carbon, and the connections to the fuselage pod up front felt sturdy (only time will tell). The wings are one piece with a healthy dose of dihedral (you know, no ailerons and all) and had a finger peg pre-installed for a right handed pilot (they include a second carbon peg for lefties too). The vertical stab is glued into the tailboom out of the box, and the horizontal stab and wing attach with a couple of screws.

After mounting the horizontal stab and connecting the pushrod to the elevator I quickly discovered the biggest flaw with this model: lack of pushrod support. The pushrods for the rudder and elevator are thin (0.020″) wire and are only supported along the tail boom every three inches. With any resistance on the rudder or elevator the pushrods buckle resulting in mushy control throws for up elevator and left hand turns which put the pushrods in compression. This should be an easy fix in the future, however I decided to build the model as-received to see how much of an issue it would be given the low weight of the glider.

One thing HK did right was provide plenty of room up front for the 300MaH battery,  de-cased RX, and BEC. RX was attached using some double sided foam tape, and battery was held down using a strip of electrical tape for easy removal during charging.

Plenty of room up front of electronics.
Plenty of room up front of electronics.

The 6ch Rx and 3s 300MaH battery (both total overkill for this model) resulted in a slightly nose-heavy model so some paperclips were added to the end of the tail boom to bring the model into the recommended CG range. The paperclips can be easily removed down the road when I add additional weight to the tail to stiffen the pushrods, or swap out the Rx and battery for something more sane (HK Orange R410 Rx).

High Tech Balancing System
High tech tail weights

After everything was stuffed, I powered her up and set the recommended control throws in the radio. I had to set servo throws biased in one direction in order to compensate for the pushrod flex, but it all worked out to get me into a flyable range.

Maiden consisted of a couple of gentle overhand style tosses to make sure the CG was in a flyable location (you never really know what to expect with the recommended CG and control throws from HK) and then some very light discus tosses. The glider felt slightly  tail-heavy, which was remedied by pulling one of the paper clip weights from the tail boom. Properly balanced, I attempted a few (more powerful) launches and easily achieved altitudes of ~50-60 feet and glide times of 20-30 seconds.  The glider tracks straight during launch, and down elevator and left hand turns did result in the expected mushiness due to the control rods buckling under compression. The quick maiden turned into an hour out in the park experimenting with different launches until my hands went numb from the cold. The model flies great (even in a light breeze) and almost doesn’t want to land; a slight breeze blowing over the trees was enough to create some lift an extend some of my flights by more than a few seconds.

TOW of 5.6oz (158.7grams)
Flying weight of 5.6oz (158.7grams)

I can’t recommend this DLG enough as an intro into unpowered flight. I’ve had the model out a couple times this winter as weather has permitted and enjoyed each time.  In the end, the only downside I see with this model is that people who find out they do enjoy flying DLGs will quickly outgrow this tiny, 2ch, balsa-winged model (I’m already looking into full-blown glass/carbon models). Those who aren’t quickly hooked will be left with a fun model that they can take along anytime and fly anywhere.

Future Improvements to Look For:
-Smaller 3-4Ch Receiver
-Supports for pushrods to prevent buckling.