"The Little Things" - Helpful hints for all

[rsisk]

Mr. Lim, you rock! Diggin the PDB on spacers. That's going on the next robot

[Mr. Lim]

I'm just the messenger for a lot of these. Mr. Stehlik on 610 deserves the credit for coming up with most, if not all of these tips and tricks. The students worked out the details to do the implementations. We don't build particularly flashy robots on 610, but a lot of these little unseen details are some of the things I'm most proud of our students for doing. It's not glamorous work (who really wants to crimp standard length pigtails on every sensor and motor? Or machine bumper brackets?) But I think they all make a big difference, and the students who are willing to roll up their sleeves and take on these "little big things" should be celebrated.

[BBray_T1296]

Originally Posted by Mr. Lim

Surgical tubing to connect encoders and pots to shafts:

Not sure how I feel about this one.

I would think the twisting slack the surgical tubing has would be a bad thing?

[Mr. Lim]

Originally Posted by BBray_T1296

Not sure how I feel about this one.

I would think the twisting slack the surgical tubing has would be a bad thing?

There is a wood dowel inserted in the surgical tubing the entire length except for the ends that slip over the shafts.

We normally mount the encoder right next to the shaft, so the surgical tube length is very small.

This year we couldn't do this because our CIMs were mounted close together, and we couldn't fit the encoder between them. We borrowed the wood dowel workaround from 188, who had done this a few years prior. I was skeptical at first, but when the guys at 188 said it ended up working great, we went ahead with it too. In our implementation, the encoders are held in place by a floating encoder mount. It uses the cylindrical shape of the CIMs, the fact that the encoders are to big to fit between the CIMs, and the slight tension from the surgical tubing to hold the encoders in place.

[themccannman]

I'll second that, as long as you have the surgical tubing properly attached to the encoder and the shaft it works great. We always ziptie the surgical tubing on both ends.

[Thad House]

The things 610 does to make the little things easier are the reason they are World Champions. Also it took me a second to determine why the PDB was not in a case. Its a beta board. But lifting the PDB is a really good idea, and I like it alot.

[PayneTrain]

I'd seen varirations of everything, but the spacers on the PDB and the redundant solenoids and motor controllers are tiny little things that really blew my mind. Those are both literally very tiny things that can make a world of difference. I'm keeping those for later.

[yash101]

Our Electrical team did a great job with their setup. However, since we had a very compact and low cog robot that can climb to the third rung and that can shoot frisbees to the high goal, our electrical board was scrunched inside the worst possible places! However, this was just a design flaw. Also, VELCRO! VELCRO is a friend!. We velcro'd many things, inclugind the bridge and the drivers. Mostly every electrical equipment except the crio and the ds. The velcro allowed us the move the electrical part to access another part. We could simply move the bridge to get full access to the shooter board. Also, our robot was quite modular. We could completely remove the shooter by unfastening a couple of bolts!

[Al Skierkiewicz]

OK,
I have been asked to weigh in on R65 so here goes. Yes, gluing is in violation of this rule (tampering includes ... gluing) but I think it is more important to look at the practice and it's repercussions. Hot glue is 'HOT' by nature and the devices you are trying to glue are not. Often the glue does not adhere to the parts in the way you want so you have a false sense of security thinking the parts are secure when they are not. They just have a big blob of useless glue attached to them. Gluing connectors into a speed controller can render the $100+ part useless when someone rips the #24 or #26 wires out of the PWM connector. If the person who is applying the glue gets a little over indulgent, the glue gets down inside the device either damaging parts as it is applied or by holding in heat that otherwise would be dissipated in the atmosphere. Rapid heating and cooling can tear parts from the circuit board. The result is the same, early failure of the internal component. We have allowed teams to perform their own repairs on electrical parts but gluing parts to them makes repairs almost impossible. As I say to people all the time, "Think for a minute and I bet you will come up with a better idea." We mount everything on perf stock. It is easy to work and all of our parts simply tie wrap to the stock. For PWM cables, we take the wire down to the deck and tie it in place. The tension holds the connector in place and a pair of wire cutters will remove both the connector and controller in the event of a failure. Gluing also makes it almost impossible to reuse the parts next year.
My Murphy's Law corollary is "Something will always go wrong at the worst possible time, on Einstein." You want to win and that means being able to change failed devices quickly. As posted earlier, we use quick connect tabs on our controllers. These are available in boxes of 100 from Digikey. So no other tools but a wire cutter is needed to change out a controller and you will have no dropped screws to look for. If everything is labeled (using colored tape or with real labels) then there is no thought needed to swap out a suspect part be it electrical mechanical or pneumatic.

[pfreivald]

Originally Posted by BurkeHalderman

Use two different colors of pneumatic tubing.

An excellent idea!

Originally Posted by BrendanB

Don't use bolts to fasten electronics that may need to be replaced on a dime (motor controllers and sidecars).

Indeed: use industrial-strength velcro, and a little extra wire. Being able to pull a speed controller (or CRio, or sidecar, or solenoid, or any other electronic component) off of the robot without disconnecting anything, where it can be jimmied/fiddled-with/tightened/etc, is a godsend for fast electronics repair! And it holds up great even under competition impacts--if anything, it helps with shock absorption.

Mr. Lim's ideas are all excellent, too. We do the additional speed controllers and solenoids thing, but never thought of mounting the PDB on spacers!

Some of 1551's standard practices that I haven't seen (or glossed over):

--keep a detailed inventory of what you have, in what quantity, and where it's stored.

--clean up every night, even if you're just coming back in the morning.

--do off-season projects that require design, and have potential applicability to FRC (manipulators, drive trains, etc). Include but don't limit to improvements on the previous year's robot.

--if you can't build two robots, at least build two drivetrains, and give one to your drivers as fast as possible.

--have tryouts. Some people think they're much better drivers than they are, and vice-versa.

--when a student says they don't handle pressure well, don't put them on the drive team.

--solder, don't crimp. (Or solder AND crimp!) (Yes, I know this contradicts other advice in this thread... As two people, get three different answers! You should also have quick crimp connectors available for really fast, omg-it's-timeout-and-we're-in-the-finals repairs.) On that note:

--invest in a high-quality crimping tool.

--consider tensioned steel cable for upper mechanical structure. It's very light, and somewhat annoying to work with, but can provide amazing strength and stability for almost no weight. But if you use it, use swedges with a good swedge crimper, and turnbuckles to account for expansion.

--invest in a mini-mill and mini-lathe for competition. Mount them on a cart, use that cart for (labeled) storage of all your bits and bobbins. I can't tell you how much use these devices have seen at FLR and Buckeye, much of it for other teams--and it keeps our pit organized as well!

--build homework/dinner time into each build night, and have a quiet room for students to work until they're done with their schoolwork; but be vigilant about off-goofing and shenaniganery!

--invite your sponsors to everything you do. Everything.

GREAT THREAD! Thanks for starting it!

[Nate Laverdure]

I think most of us can agree that hot gluing your PWMs is not an elegant solution. We can put this aside and focus on whether it's really a practice that can cause equipment damage.

Originally Posted by Al Skierkiewicz

Hot glue is 'HOT' by nature and the devices you are trying to glue are not. [snip] Rapid heating and cooling can tear parts from the circuit board.

This particular concern sounds silly. Your standard 99.3Sn-0.7Cu lead-free solder melts at ~440 F (PDF link). Eutectic 63Sn-37Pb solder melts at 361 F. Low-temperature "hot melt" glue guns range from 170 - 250 F.

[Al Skierkiewicz]

Nate, it isn't desoldering I am concerned about. It is the thermal shock and the resulting expansion/contraction that may force components off the board or produce cracks in the circuit traces.

[pfreivald]

Originally Posted by Al Skierkiewicz

Nate, it isn't desoldering I am concerned about. It is the thermal shock and the resulting expansion/contraction that may force components off the board or produce cracks in the circuit traces.

So, I'll admit to 1551 being hot gluers... (We didn't realize it was a violation, had always passed inspection even at CMP, and will not be doing it in the future now that we realize it's illegal.)

In our experience, the pwm connections tend over time to jiggle free, even with strain relief. Once we started hot gluing, we never had another problem.

Given that this is a helpful hints thread, what do teams do to secure pwm connections on the sidecar and on speed controllers without using hot glue?

[Alan Anderson]

Originally Posted by Al Skierkiewicz

Nate, it isn't desoldering I am concerned about. It is the thermal shock and the resulting expansion/contraction that may force components off the board or produce cracks in the circuit traces.

We're talking about the PWM connector on a Victor here, right? The top of its shroud is nowhere near any circuit boards or components, and the typical practice has been to use less than a pea-sized bit of glue to keep the plug from vibrating. Removing it is trivial if you have a thumbnail. I do not believe there is any danger at all of thermal shock to the board. The only thing that will get remotely warm is the surface of the plastic. It's not like anyone is suggesting using the hot glue to make a watertight connection around the entire connector.

That said, it's still glue, and gluing a speed controller has been explicitly forbidden by the rules.

[FrankJ]

Don't disconnect plugs by pulling on the wires. Be it battery connectors or PWM cables.

Keep metal swarf away from electrical components.

I am going to leave off with the hot glue thing with "it is against the rules so don't do it." Makes the advantages/disadvantages of doing so moot.