October 22, 2018, 10:02:52 pm
Welcome, Guest. Please login or register.

Author Topic: Del Mar Bot Depot  (Read 4490 times)

daggius

  • Roboteer
  • ***
  • Posts: 102
    • View Profile
    • bot videos
Re: Del Mar Bot Depot
« Reply #15 on: July 27, 2018, 04:33:59 pm »
Greetings all.  I have been lazy with my bots lately due to getting massively addicted to mountain biking however I would like to jot down some notes so i don't keep forgetting lessons learned due to long breaks in between events.  Sorry for the crappy writing, i dont intend anyone to actually read this, and maybe ill post some juicy pics later.

Seattle Bot Battles 2018

Awesome event and Seattle in June is great, riding limebikes randomly around city is super fun, plus follow up trip to the olympic national forest was amazing!

RTR was basically unchanged since a year ago.  I had 3d printed some new arms from 910 instead of PETG and was set to swap them in last minute before the trip but it turned out due to the lowered stiffness of the material the bot was no longer able to self right!  That was fun with some late nights  reprinting and trying to reassemble everything and get it working at the last second.  I got things working before the flight and all set with a test drive at the arena.  Then while tinkering with the radio, one of my servo managed to fry itself minutes before my first match.  Murphys law! haha.  After frantically swapping out a servo (the wrong kind too having not brought adequate spares), only had arms working partially for first match.  After this was able to realize that one of the new printed arms had cracked and so needed to scramble to hack some aluminum sheet onto it as a sketchy repair, though after another match with arms not really working I began to remember what been learned and forgotten from previous year:  the front wedge is so low and nice on RTR that it drives around the arena scooping up massive amounts of debris into the internal compartments.  The debris wreaks havoc on the mechanism, jamming up the pivots and thus causing arms to break or servos to fry themselves.   Got a great tip from someone at the tournament to just use a worm gear instead of servos to deploy the arms so that it will be way more robust.  I would totally due that if I had the energy to rebuild from scratch! Luckily the arms are basically just there for self righting and taunting, in SBB you can usually get a pit in beetles without actually lifting them, so was able to go 3-2.  Could have won 4 in a row if I didnt self pit due to reckless driving.  The frame and front wedge are really solid and barely take any damage.  So  I have printed a new arm and raised up the internals pivots by .094" so the arms can have some clearance with teh bottom mounting plate when fully stowed.  Hopefully this helps with debris issues and I can also focus on getting some stronger servos in so that lifting opponents and self righting is more reliable.  Its really important to remember to aggressively clean out the bot between matches to avoid jams.  Anyway it is great fun to drive this bot, and I am still surprised that was basically a joke bot turns out to handle so well and surprisingly durable for being 3d printed.  it could definitely be champion if i just dial things in a bit more and drive better!

Flex 2.0 was unveiled, having been in the works since Flex 1 was retired after a massive beatdown from Pizza Slicer in SBB last year.  The goal was to be faster, stronger, lift opponents high enough to go over a 4" wall, other such awesome things.  Despite my attempts to make the new bot more robust and durable, due to the multitude of design changes, of course nothing really worked out right since new bot issues, and the thing came out way worse than the original.  The one solid improvement was self righting allowed by the longer lifting arm, which was very reliable.  However the lifter in general was more fragile.  Thinking I would make it very flexible to attenuate impacts and protect the servo, including flexures between the servo and the arm, as well as flexible materials (nylon).  Although my folly was still leaving one stiff link in, the aluminum servo horn, which allowed any lateral load to be transmitted directly into the servo, breaking it by the end of the tournament.  So I have updated the arm to include some aluminum, and can soften the linkage to the servo horn, to hopefully take the load into the chassis from there instead of thru servo.  Lateral load needs to go into the chassis, torsional load onto the servo needs to be absorbed by flexures or servo saver.  havent got it totally dialed in yet.  Luckily 2.0 has easy servo replaceability since it is not glued in anymore.  Learning/remembering a lot about stiffness this tournament. Another fun thing was mysteriously breaking motors.  Which happened twice despite the wheels never taking direct hits due to wheel guards/foam wheels.  I think what is happening is that because of the way I am mounting the FT motors, basically by "hose clamping" around the gearbox instead of mounting to the tapped holes on the end of the gearbox, that they are exposed to more shocks going thru the chassis due to being directly in the load path.  So that while Flex 1 was impervious to damage and would never break internals, now even glancing blows can render it immobilized, which freaking sucks.  I burned out 2 motors during the contest.   My other theory is that since I am running 4 or 5S now, the loads on the gears are higher, but this sholdnt be breaking them according to the FT tables, just causing them to fry brushes faster so wear out faster long term basically.  It was also fun to learn that going up from 3S didnt even help due to low amount of traction.  It's critically important for a 2 wheeled bot to have the wheels close to the CG, but not so close that you wheelie, something i forgot during time off.  So traction was garbage due to my new wheel placement being poor, and increasing voltage to motors didnt help, and also due to traction, could barely drive around after having people in a clamped lift position.  Furthermore, having failed to get some titanium parts together, my 3d printed 910 and PTCPE wheel guards were both too puny to withstand damage (while solid UHMW in the same thickness is basically impervious to damage) and were destroyed nearly instantly.  also new super fast servo was so twitchy that it was hard to actually use correctly, and new clamping mechanism has significantly less grip.

After going 2-2 with flex 2.0 the highlight was taking on 5 fairyweights in a hilarious 5v1 rumble and getting annihilated by team dark forces (which thankfully exposed the lateral stiffness issue of the arm).  Anyway there are various bodges available to address these shortcomings, such as adding magnets, angrily binge drinking in the garage etc, hope to come up with some crappy solutions soon!   battlebots is truly an iterative process, you must continue testing and refining till you finally get things dialed in correctly or ragequit! cheers
« Last Edit: July 27, 2018, 04:40:55 pm by daggius »

daggius

  • Roboteer
  • ***
  • Posts: 102
    • View Profile
    • bot videos
Re: Del Mar Bot Depot
« Reply #16 on: August 01, 2018, 01:07:32 pm »
Returned from another awesome event at DUTS.  Good to see everyone and some hilarious bots.  Some notes from my fav battles

https://www.youtube.com/watch?v=ucAdD1knixE

First battle against Shockwave.  I should have sent the eventual champion to the losers bracket but my servo burned out!  After taking it apart as well as the one that failed in Seattle, I realized my folly -- Savox servos don't like to run at 9V since one board was totally burned out (servo had also got really hot) and the Seattle one, while it was still somehow working till the gears broke, the board was also singed :(.  I kind of painted myself into a corner since these servos I also didn't realize are slightly smaller than a standard servo, and I 3d printed the perfect containment dimensions for it, so I should probably just address this by running a lower UBEC or reprinting a larger chassis and using a servo that does 9v and is a bunch cheaper too!  After these changes, and since I now have a nice squishy servo saver in place, hopefully the damn thing will stop breaking!

https://www.youtube.com/watch?v=vrl6F4Ctho8

Battle with Syncope.  RTR spent a lot of time spinning the tires in pushing match while riding on Syncopes wedgelets.  I should probably add a buttload of magnets since I have 0.3 lb to spare, which will also help with handling.  Riding on the wedgelets for so long that actually I found about 8 screws had backed themselves out after the match just from the vibrations of his bot! Including all 3 screws holding one of my drive motors (despite having loctite) on which was just wobbling around at the end, haha!  Actually seems like a good idea for a weapon, kind of like an electric tooth, just grab the other bot and vibrate them to pieces.  RTR's arm mech and servos seems reliable now, having no debris or servo issues, and only damage was to outside armor and wedge.  The 910 armor got badly cracked in back so I did some melting with butane torch to stick it back together which is my preferred repair method rather than bothering to print spares.  The Ti wedge got some deep cuts where it was riding on the STD

Smee and Duct Spartan

https://www.youtube.com/watch?v=lF3TcU6CWyg

https://www.youtube.com/watch?v=D_R3bSICN5Y
« Last Edit: August 01, 2018, 01:10:44 pm by daggius »