Never Do This, and Other Good Ideas Gone Bad

[114Klaatu0x72]

During Overdrive, our team decided on using a pivoting arm with one or two 80lb constant force springs, along with an underpowered gearbox, Chinese cast iron gears and "pwm cables" aka painted string, and a broken anti-back drive mechanism. It worked about once, and then the gears sheared off half their teeth, the anti-back drive pins fell out, and a cim burnt out in the manor of seconds. We switched out all of the motors with pneumatics between two competitions, and some of our problems went away. Later that year it knocked me out twice from a blows to the head, one time during a presentation to a middle school, as well as being just terrifying to be around in general. All of those problems pretty much stemmed from a few simple issues: a lack of prototyping, a love of cheap parts, and miscalculations in math. If we spent another week prototyping and a few hundred more dollars, and double-triple-well-beyond-the-point-of-redundancy checked our math, we could have done great that year. So, pretty much anytime you think something will work, despite how sure you are of it, try it and make sure it works.
Also, large amounts of force absolutely suck to work with, try to avoid designs that use them, as they tend to fail in some way after a sometimes very short while.

[Tristan Lall]

Originally Posted by 114Klaatu0x72

During Overdrive, our team decided on using a pivoting arm with one or two 80lb constant force springs, along with an underpowered gearbox, Chinese cast iron gears and "pwm cables" aka painted string, and a broken anti-back drive mechanism. It worked about once, and then the gears sheared off half their teeth, the anti-back drive pins fell out, and a cim burnt out in the manor of seconds.

I'm curious about the gear specs and the load you put on those gears; would you mind explaining further? (Besides, I also want to know the supplier of the gears, so that I can consider avoiding them!)

Originally Posted by cire

2003... Never use #25 chain for your drivetrain. It may seem like you will never hit that several hundred lb limit, but when you throw in missalignment and shock loading, well lets just say we went through a lot of chain that year.

188 broke about 13 strands of ANSI #25 in 3 events that year. Of course, we didn't exactly have the best setup—small drive sprockets on very powerful transmissions,* rigid and convoluted chain guides instead of idlers, and a bit of misalignment.

On the other hand, on a different drivetrain, ANSI #25 chain can be a perfectly reasonable choice—this year, for example, with all the high-speed driving, teams can probably expect substantially less chain load, especially if they've got room for big sprockets on their big wheels.

*Actually, as far as I know, they were the most powerful transmissions on any FIRST robot to that point; each transmission had a CIM, a Bosch drill motor and a Fisher-Price motor.

[114Klaatu0x72]

I don't have the exact gear specs, but I beleive the gears which broke were something attune to:
*80 and 16 tooth
*face width of a quarter inch
*14.5 or 20 degree presure angle
*pitch of either 16 or 20
*Made of cast iron, cheap cast iron.
There was maybe about 20 to 40 ft-lb of torque going through them, but its hard to say exactly because really nothing is left of that robot (cad, docs, notes, pictures) excpt for a stripped out frame and a single, half bald gear on our wall of shame. As for where we bought them, all I remember was that they came from some Chinese distributor and had BOSTON embosed in them. Unless you are constantly looking up cheap items and can read Mandarin, you probably won't come accross them.

[jwfoss]

Based on my years on Aces High (FRC176) and Division by Zero (FRC229) heres my list of things to avoid in season:

Never Agains:
1. Sissor Lifts
2. Tank Treads
3. Vaccuum Devices

Avoid at nearly all costs:
4. Cast Iron Worm Gears
5. Gravity Fed Hoppers
6. Gravity Dependent Lifts/Arms

And until done completely and perfected in an "offseason"
7. Crab Drive/4 Wheel Steering/Holonomic Drive

[Newo95]

In 2008 we used a pogo- stick and pneumatically actuated arms. It had 6 wheels with two omnis in the front, and chains to drive them. Unfortunately, the arm couldn't grip the ball tight enough, the chains were too heavy, and the pogo stick needed more repairs than any part on any robot since then.

[Grim Tuesday]

Never change your design with only two weeks until ship date.

[XaulZan11]

Originally Posted by Grim Tuesday

Never change your design with only two weeks until ship date.

Not sure about this one. There have been plenty of teams that changed their design late in the season or during the season and have done very well. The best example of this is probably World Champion 67 in 09. Just off the top of my head others include 469's gripper in 07, 2039 in 09, 33 in 09, 16 in 09. All of those teams either won regionals or made it to the eliminations at the championshop with their new design.

[Chris is me]

Don't change your design unless you really have to and can put in the man hours to make it work. The wrong thing well is better than the right thing badly.

Examples of "you really have to" would include the pincher rollers of 2010. Every competitive team last year at least considered that change, I guarantee it.

[Alex2614]

Originally Posted by IKE

I will let someone else cover scissor lift.

Ugh... don't even get me started on the scissor lift.

Go to a competition without replacement parts, including an extra minibot.

At SVR this year, the small gear that controlled our who arm system snapped in two after a team hit it with their claw (accident, of course!). We had no other gear to replace it with, and spent the remainder of the competition playing defense (which we weren't that bad at).

Also, our minibot got executed. During yet another collision, our minibot somehow fell off our robot, got stuck under our wheels, and got half torn apart by us. Then, once we got off, the other teams, including our alliance, ended up running it over 1 way or another (again, all accidents)!

To anyone who attended SVR this last season, if you found a neodymium magnet attached to you your robot somewhere, and you went against 256 in a match, can we have it back?

[Andrew Remmers]

So far I would have to say my worst experience was 8020 elevators... But thats me.

[pfreivald]

Originally Posted by Joyride_67_1902

So far I would have to say my worst experience was 8020 elevators... But thats me.

Last year we had excellent success with Bosch Extruded (e.g. metric 8020) -- totally rocked the Finger Lakes Regional with a highly successful endgame. (I think we failed to lift only twice, and both of those times were because of issues totally unrelated to the performance of the lift itself!)

Bosch has slide carriages that fit their 30mm extruded and work extremely well. We used more of it this year for our minibot deployment.

[JesseK]

80/20 (and maybe Bosch too) sliders really start to bind on multi-stage compound elevators. You also have to keep the 80/20 perfectly square to minimize binding, and at height there's a higher tendency for torsion. We scrapped 80/20 in favor of powder coating + teflon for slide friction that was close to what a bearing slide would have; the box aluminum also resisted torsion alot better.

[pfreivald]

Originally Posted by JesseK

80/20 (and maybe Bosch too) sliders really start to bind on multi-stage compound elevators. You also have to keep the 80/20 perfectly square to minimize binding, and at height there's a higher tendency for torsion. We scrapped 80/20 in favor of powder coating + teflon for slide friction that was close to what a bearing slide would have; the box aluminum also resisted torsion alot better.

We have not had even a single issue with Bosch slider carriages binding on multi-stage compound lifters -- and from trackballs to minibots to entire robots, we've really put them through their paces! I can't speak for 8020.

[Kyler386]

I don't know if the majority of my team would agree with me on this, but here's what I think:

2006: No more turret designs. According to members & alumni, 06 was an all around horrible year, robot and all.

We broke this rule in my first year. (2009) Needless to say, our robot scored once that whole year.

2010: Do not try to stop a pnumatic cylinder halfway. We had a "Bump Mode" that allowed our robot to go over the bumps. To do this, we had to activate our pnumatic kicker, and then immediately bring it back about halfway so that the kicker would clear the surface of the bumps without the kicker breaking the plane of the bumpers. Sounded good in theory, (which, according to a forum regular, is a nice place) but it did not work in the real world.

And the new one for Logomotion: NO MORE BUMPER SKIRTS!!! Those things are horrid.