Never Do This, and Other Good Ideas Gone Bad

[hallk]

Epoxy works great when you use it on it's intended materials that have been properly prepared. It doesn't work work quickly at all and expecting it to hold together for the next match in 20 minutes will never work. It is also a disaster when someone loses the caps to the epoxy and then just puts it into a plastic bag at the bottom of the tool bag.

Trying to speed hole extruded Aluminum will just result in bad things.

Based off this year's experience with mini bots, don't use Tetrix motors and expect them not to fry.

Make sure to pack your controls.

Not robot related but..don't let someone handle the bookkeeping if the have no accounting knowledge. Make sure someone goes over said books before the night that the sponsorship application is due.

[gren737]

Don't use rigid couplings or pin shafts. Love joy 3 piece couples are great as are offset couplings.

The benefit is 2 fold, it gives you some wiggle room in case things aren't perfectly lined up (like that would ever happen on a FRC robot) and it also makes changing parts much quicker in the event of a failure, you can build spares right up to the coupling.

my $.02

On my teams I will not allow scissor lifts, mecanums, multi stage elevator lifts, winches, large pneumatics or belt type conveyors (poly cord ftw!)

[pfreivald]

Every year we seem to have to relearn the same lesson: Don't engineer things too closely.

If something works but *barely* works at home (meaning that it might be fantastic, but the tolerances are very tiny), the Unmutable Laws of The Universe dictate that it will NOT work at competition.

In other words: All values are nominal values, even if they aren't nominal values.

[RayTurner1126]

team 1126 will most likely never use a cable winch again, after failing several times, including in 2010 when i personally worked on our double-barreled (kicker and elevations) monstrosity of a winch, i understand our failures. we found that cable just frays and gets tangled much too easily

this year we used a chain winch to power our elevator and it worked GREAT, sure, there is extra weight involved, but it's worth it to not have to deal with that cable! i don't think we'll ever go back to cable!

[pfreivald]

Oh, right:

NO CABLE WINCHES.

Also, AVOID QUALIFYING MATCHES WITH 1126. It doesn't seem to matter how well either of our teams are doing in any given year, if we have a qualifying match where we are on the same alliance, it is always a disaster for both teams. This has been true at every FLR since 2005. No legitimate reason -- we always get on just fine in the pits, and love each other so... But when it come time to play a QM, disaster. Every time.

[Kyoshirin]

1) Build the most important part last:

Did this on an offseason robot last fall. It had a bucket that lifted via parallel arms. By the time the frame and bucket were done, we had no room for the parallel arms unless we used flat. The bucket never fell back inside the frame when we let it down. We almost completely rebuilt our robot after this.

2) Use flat for parallel lifting arms:

It makes things go up and down fine. It also has the unfortunate side effect of things waving around to the sides.

3) Not read all of the manual:

We used both a DeWalt and a Globe on our robot this year. Luckily I was glancing through the manual one night (procrastinating homework) and noticed. We spent the next two weeks trying to mount the window motors.

4) Have someone build a part for a design they hate:

It either will be done wrong, or just not done at all.

5) Non-Polycord belts:

We have found that they don't work for us, from slipping to coming off the roller to being so ineffective it's useless.

6) More than one gate latch on an arm:

The amount of sway and overall inefficiency as well as a horrible lack of control prevent this from being a remotely good idea.

7) Human player fed ONLY

We did this for an offseason competition. We probably would've won if we didn't have to cross the field to reload, score, reload, score, etc.

8) Build something without a definitive strategy, or a good idea of how to make it work:

Here is a list of robots we've (recently) had to completely rebuild after the first competition fro this reason;
FIRST 09 (from Archimedian screw to polycord conveyor)
FIRST 10 (kicker and hanger both didn't work)
Offseason 10 (See #'s 1 & 2)

[Nick Lawrence]

Our 2010 robot drivetrain.

Just don't.

7) Human player fed ONLY

We did this for an offseason competition. We probably would've won if we didn't have to cross the field to reload, score, reload, score, etc.

I might suggest you reevaluate this one. Have a look at this robot and this robot. Both robots human loaded. Both robots played on Einstein their respective years. You can get away with human loading if you do it right.

-Nick

[StevenB]

Never use KOP servos as a means of grabbing the game piece, no matter how light the game piece may seem. "We need something really light to power the tube gripper out on the end our long multi-jointed arm... I know, servos!" Our 2007 robot was hastily dismembered after the season, and the veterans try to shut it out of their minds.

Really, it seemed like a good idea at the time.

[the man]

I must say our old team, 2645, used set screws all the time with no problems, we used mechanum wheels, a scissor lift, and a cable winch. All on one robot with no problems except a broken logitech controller. One of those examples of engineering gone right.

[AdamHeard]

Originally Posted by the man

I must say our old team, 2645, used set screws all the time with no problems, we used mechanum wheels, a scissor lift, and a cable winch. All on one robot with no problems except a broken logitech controller. One of those examples of engineering gone right.

A lot of things mentioned in this thread aren't mechanically wrong, they're just not the best way to do things (for various reasons). A robot with mecanum wheels and a scissor lift has no reason it couldn't be mechanically reliable all season, it's just not likely to keep up with a more optimal design.

[Billfred]

Something we learned the hard way at Bayou 2010: Don't put off numbers if they have to go on the bumpers...if you're having to use paint outside, the result will look like crap.

JVN's silver Sharpie tip was a revelation this year, though that still took us a couple of hours in our shop with a small group of people tracing and going back over the numbers a few times. (I think we put at least three or four coats on our robot's bumpers, which we did with a rather thick stroke--we bought six markers and had usable ones afterward.)

I wouldn't want to do anything resembling that level of effort in the pits ever again.

[Tristan Lall]

Originally Posted by Billfred

JVN's silver Sharpie tip was a revelation this year, though that still took us a couple of hours in our shop with a small group of people tracing and going back over the numbers a few times. (I think we put at least three or four coats on our robot's bumpers, which we did with a rather thick stroke--we bought six markers and had usable ones afterward.)

To make the numbering process faster, couldn't you cut the markers open, extract the plastic-wrapped felt ink reservoir, and use it like a big, wet crayon? (Who cares that you can't write with the markers anymore? And if at a competition, there will be plenty of other teams with inadequate bumpers who could benefit from the leftover ink.)

[Billfred]

Originally Posted by Tristan Lall

To make the numbering process faster, couldn't you cut the markers open, extract the plastic-wrapped felt ink reservoir, and use it like a big, wet crayon? (Who cares that you can't write with the markers anymore? And if at a competition, there will be plenty of other teams with inadequate bumpers who could benefit from the leftover ink.)

I haven't experimented with this--in our situation this year, we had some time, plenty of cheap labor students, and me leading the process in clothes for work that didn't need to be silver if I could help it.

Unless we were in a panic, I'd probably keep with how we did them this year (better photo: http://twitpic.com/40kj4i)--the only improvement would be a true stencil. (We traced old iron-on transfers that we used unsuccessfully in 2010--but we didn't have a 5, so we improvised it using other numbers we did have.)

Of course, if someone can show how it's done (and without creating a total mess), I'm all ears.

[Molten]

This is more of a bad idea gone bad but it certainly applies to the side topic of team numbers on bumpers. In 2007, 1766 forgot to have our numbers on our bumpers and had to improvise at the competition. Going pit to pit we found a team that had some paint and got permission to borrow it. We cut a stencil out of paper and cardboard and went outside to spray it on. In the end I held the bumper with one hand while spraying with the other. My arm had orange on it for the rest of the competition. I would give credit to the team that gave us the paint but I can't remember the number. It was a tiger themed team at the buckeye regional in 2007 if anyone else recalls the number.

My point to all this is that there is always a worse way of doing things. I never once thought of a sharpie and immediately went looking for paint.

[pfreivald]

Bumper skirt.... Bad idea. They almost got us DQ'ed at FLR.

Next year we'll be making two sets of bumpers!