Re: 95 Hard at work
 

James,
Where can you get those current sensors and how exactly do you employ them?

Re: 95 Hard at work
 

Current sensor: http://www.digikey.com/product-detai...7-1-ND/1680542

Break-out-board: https://www.sparkfun.com/products/494

The current sensing chip is just a hall-effect sensor, so you only need to attach it to the wire that its measuring current. Ideally split apart for other wires as much as reasonably possible. The snag is that they must be individually calibrated since their sensitivity is dependent upon location (mainly distance from the wire).

Re: 95 Hard at work
 

Attending the Week 0 event in Merrimack NH, hosted by #166, The Chop Shop. A well-run event, met a lot of great people, drove our robot a bunch, and then we broke it!

We got a basic autonomous program working (drive forward, score in low goal). When the robot was mechanically sound and setup properly it worked nearly every time.

Had a few issues with our intake rollers, one jumped a chain once and has since been fixed. Another one sucked in a spring and completely jammed up. We're going to replace the springs with rubber bands in the hopes that rubber bands will shred instead of getting tangled up! It made us really appreciate having two intakes/scoring mechanisms, we were always able to use the other if one of them went off-line.



By the end of the match in the following video one of the WCP colson hub wheel bearings had shattered and the other was on its way out. The drive train chain had jumped and jammed inside the robot. With a spare bearing from 1058, PVC Pirates, we managed to get out to the field with a marginally functional drive train to lose in a brutal eliminations match.

I have no idea why the field people/human players kept tossing balls onto the field... at a few points there were 2 or 3 red balls on the field at the same time.

https://www.youtube.com/watch?v=0vojAM_yHwI

Overall a very good learning experience, got a lot of little things sorted out on the robot and got a lot of good drive practice in too. We're looking forward to sorting out the exploded wheel bearing issue soon and fine-tuning a few other things on the robot.

Re: 95 Hard at work
 

Attending my first NE event... yeah brutal matches. Bot looked great out there James, can't wait to see it again soon.

Re: 95 Hard at work
 

Thanks Andrew! I was really impressed with your team's robot, 125 will continue to be a seriously competitive team.

Re: 95 Hard at work
 

So... last night we did an autopsy on our drivetrain.

One wheel's bearings had failed completely at week 0. Its other bearing was close to failure, and two other wheel assemblies' bearings were also close to failure. :eek:

We dug up our spares (6inx1.5in Colson wheels with WCP dead-axle hubs and bearings), which have never been used, and found that the bearings in one of them were already 'cogging' or 'clicking' or 'crunching,' however you want to describe that behavior directly associated with bearings that are about to fail. Very disappointing. One of the dead-axle hubs wasn't even tapped!

After popping some of the bearings out we found that the lips on some of the WCP hubs were peened in asymmetrically, presumably to retain the bearings. We had a few extra bearings so I tried pressing them into the hubs after trying to clean them up a bit. In one hub the bearing actually cracked as I pressed it in!

Whatever was going on, WCP had some serious quality control issues with their dead-axle colson wheel hubs and bearings. I've never been disappointed with any OTS FRC hardware until now... and very frustrating to have it happen right at the end of build.

New bearings are on the way from McMaster and AndyMark, hopefully we can get something together before midnight tonight, even if we have to re-bore the colson wheel hubs.

Re: 95 Hard at work
 

Sorry to hear James. Hope your new hardware works better. Good luck.

Re: 95 Hard at work
 

Thanks.

Well... nothing is ever easy! The replacement bearings from McMaster had a lip on them that we had to take off with a surface grinder. On the plus side surface grinders are fairly amusing to use...



After taking apart the remainder of the WCP hubs we found evidence that some of the bearings had been pressed in at an angle, likely contributing to the premature failures we had. We reamed out all of the hubs to 0.875in to give the replacement bearings a light press fit/slip fit, which worked really well.

On that front: WCP has gotten in touch with me about the failures we saw and seem genuinely concerned about what's happened and intend to figure out why and make it right. It's nice to have see a company go out of their way to stand behind their products. I would, and probably will, buy WCP products in the future despite this frustrating experience given how good their customer service has been thus far.

Back on track... with fresh bearings we reassembled the drive train. Runs like absolute butter now, and the robot seems to be faster and to track straighter. We can't even hear the chain anymore. We also replace the coil springs that keep some string under control with rubber bands. The rubber bands get shredded if they get sucked into the carwash, instead of tangling up the carwash completely like the coil springs did (see earlier post).



We spent a bit of time tweaking and rebuilding the carwash mechanisms. Most notably we replace the mini-CIMs with RS775-18 motors and CIM-U-LATOR gearboxes. This drops about 1lb of mass from each assembly which helps deploy/retract times, makes the robot less tippy, and gets us a more powerful motor with an output shaft that can deal with side-loading well.



We also added some ultrasonic sensors to the deck that allow us to see an incoming ball and automatically catch it. We may need some additional tweaks to get it running smoothly, but have cleanly demonstrated that it works.

At 11:11 we bagged the robot. It was fully-functional, had all of its mechanisms in place, had been driven in ~10 matches and lots of drive practice, had a working autonomous move, and was 10lbs under weight. Man did that feel good after the roller-coaster that was our week-0 experience, and a LOT better than last year.


I hope everyone liked reading our 'build thread.' If enough of you did then I think we'll do another build thread next year. I know its not as tantalizing as a bunch of teaser pics/videos, or as dramatic as some of the well-done reveal videos, I just hope it was a little helpful and a little inspirational to see a team's whole manufacturing process. Maybe next year we'll include the prototyping and design processes.

What say you, CD?

Re: 95 Hard at work
 

Looks complicated.

Re: 95 Hard at work
 

Complexity is a relative thing. In many ways this is the simplest robot our team has made in many years. It certainly doesn't posses any particularly intricate mechanisms (beyond the purchased shifting gearboxes); just simple chains, sprockets and pneumatics.

It terms of design, fabrication, maintenance and use it's pretty straightforward. Everything is just mirrored left/right and front/back, and the most complex mechanisms were iterations of things we've done in previous seasons or during the fall. It might look more complicated than it really is because we have two of everything.

Re: 95 Hard at work
 

What I meant was that it looks complicated. I have found that making something that you understand [i]seem[i] more complicated, then you seem smarter. It works almost every time. It's awesome.





Good luck everyone!

Re: 95 Hard at work
 

No trickery here. Are you implying that we're trying to look smart by making our robot appear more complicated than it is?

Re: 95 Hard at work
 

I love how well labeled your equipment is. We will definitely be doing that!

Re: 95 Hard at work
 

No, I'm saying that you ARE smart. The robot looks complicated to me because I didn't make the bot. To me, my robot looks simple. to other people, my robot may seem overwhelmingly complex. No offense intended in my earlier posts.

Re: 95 Hard at work
 

Thanks! At first it seemed excessive... and then when we went to debug things and assign all of the controllers in code and it all paid off. We will be just as thorough next year.