Friday, March 10, 2017

Margin of Safety Version 1

Lets fix the problem with the weapon bearing coming out. Put the bearing in an aluminum holder that maintains dimensions, and put that against the plastic to keep the same degree of shock mounting.

The bearing is held in with M4 flanged button screws. Eight of them.

It slots into the frame.
The robot uses the same dimension blades as before.
Blade down
It's flat.
Finally, this robot uses 12 mm brushless motors connected to Silver Spark 22:1 gearboxes for the drive. To control them, I made my own triple brushless control board. It's based on the STM32F3 and Allegro A4960. It was squirrelly and hard to drive, despite gyro correction in the controller.

I'm not perfect.
It then competed at Dragoncon 2015, winning first place, and a local event SLAW, also winning first place. Post event pictures follow.
Still together

Evidence of the blade pushed into the frame, but this is anticipated.

Tip is a little worn.

I fought a robot with large wheels that could crawl over this robot (though that's not hard). It was equipped with a vertical bar that hit from above.

A few knicks in the back.

Went face to face with a drum. Not bad.

The brushless motors run super hot. To solve this, I bought thick heat conductive padding for the motors. It worked great, even though the conductive padding is more rigid than the standard polyurethane rubber.

Both sides.

Brushless motors stuck on a silver spark.

Pinion appears alright?
Hub motor.


I actually go on to use this hub on quite a few other bots.

Good condition.

Stator fine.

The pieces.

Saturday, March 19, 2016

In the Margins: Save the drive at the expense of the weapon

In the Margins is a beetleweight combat robot. It is the least documented bot I have. Sorry.

Its frame is almost entirely UHMW. I am not a fan of the material, since it likes to slip out of vises and collets, but it does absorb impacts pretty well, which keeps the drive motors happy, which keeps me happy. It competed in Motorama 2015, with a record of nine wins, two losses.

Its goals were to:

  • Not destroy its own drive motors.

Before the Event


After the Event

The plastic around the weapon deformed enough to allow the large bearing to escape its screws.

Videos

Two Ply: An axial flux successish story combat robot.

Two Ply was an attempt to make a miniature axial flux motor robust enough for robot combat (in the three pound weight class. It did work, with significant heat output. It competed at Dragoncon 2014 and Combots Cup 2014. The goals for this bot were:
  • As thin as possible.
  • Use an axial flux motor.
The motor was a 15 tooth, 16 pole single sided thing with 600 Kv or thereabouts. It had heat and weapon blade hardness issues, but it did work otherwise. It record is one win, two losses.

Before the Event



The way the motor was designed, I would need flat wire to get the desired number of turns with the required current capacity. No one sells flat wire in the aspect ratio I needed, which was 2 mm x 0.25 mm, or 10:1 ish. So I curled my own copper sheet.

Added double sided sticky Kapton as insulation. This is a test coil. Looks good.

Dammit bandsaw.

Top half of rotor.

Drilling corners.


Vise flats, dat design for manufacturing.

To make these vaguely dong magnet holes.


Since the rotor is aluminum for weight, there is nothing to complete the magnetic circuit, which results in weaker magnetic fields. To get around this, I set the magnets up in a Halbach array, which creates a single sided magnetic field without the need of backing iron.

All magnets superglued in.

Vacuum chambering epoxy into the gaps.

Test fitting waterjet parts.

So thin.

Sooo thin.

Making windings.


15 windings

The tricky part is connecting the inside of one coil to the outside of the next coil.

Just flat enough.

Ready for potting.

Potted with thermally-but-not-electrically conductive epoxy. The backing iron for the coils is cold rolled steel.






During the Event (Dragoncon)



Unsure if scorching or magnet rubbing.

Holds up good.


 After the Event (Combots Cup)

AR400 is not a weapon steel.


AR400 is not a weapon steel.







The weapon rotor had a angled ball bearing for radial and axial loads in one direction and a thrust bearing for axial loads in the other direction.

Eight M4 screws are good at holding the blade in.



The blade was deflected enough to cut the motor wires.


One of the silver spark motors stopped working.


Brushes are fine


Commutator ok.



Hah cracked magnets.

A pittance of thermal compound to titanium, the best heatsinking metal known to man! (sarcasm)




AR400 is not a weapon steel.



AR400 is not a weapon steel.

Playlist of all matches.