Shown below is what we are taking to the upcoming DC Knights competition. Feel free to ask questions or make comments!
Note: Wiring still needed to be cleaned up at the time the photo was taken.
Wait why do you have 2 scoring mechanisms??? 
The front claw is a de-scoring mechanism primarily. Our game strategy does not revolve around scoring entire stacks of tubes.
:)I understand this, but why cant your claw just score all of the tubes, and then take of 2 or 1 and score it on an other goal post? Or just make a chain collector that can de-score somehow? like 2921? find a more simple way than having 2 mechanisms on 1 robot. You and your team have to make the choice to choose one of those mechanisms. 
It is that team’s choice to have more than one mechanism. If they want to take it off, they will. It may even be beneficial to have more than one scoring mechanism. Personally, when I look at the mechanism on the back, I see a hang. Please tell me it can hang 
The team decided it was more advantageous to use both a claw and a ‘snorkel’. We played with the idea of a de-scoring mechanism folding out of our arm, but had trouble getting the geometry correct. The length of our arm and snorkel prevent us from folding it down far enough to use a ‘fork’ style de-scoring mechanism. Plus, one of the students had already created the claw, so it was just a matter of fitting it into the robot’s starting box.
The other nice thing about using a claw is that you can relocate entire stacks of tubes to less convenient locations for the opposing alliance.
Another photo showing an alternate angle - pardon our mess.
Comments/descriptions:
Aggressive 3x speed up gearing with Dual motors each side on 4" hi-traction tires in rear, HS chained to 4"omnis in front;
Sturdy square frame with 1x5x1 dual steel C channel rails.
No ultrasound/line follower/polycarbonate/hanging/tube nest visible.
Frame may provide nesting to goals.
One 2AH NiMH battery.
Looks like needle pickup can work in front or back.
Probable motor allocations, as far as I can tell.
2L + 2R drive, with L,R encoders for feedback
1 claw clamp w curved wings, but not 4bar, no feedback?
1 clamp lifter on 4 bar arm, with spring assist
2 for arm shoulder joint, 1:9 torque ratio?, w pot feedback
1 for arm elbow joint, 1:5 torque ratio?, w pot feedback
1 for needle roller pickup
---- 10 total motors
Questions/suggestions:
- Don’t your drive motors overheat with this gearing?
- Or do you have advanced drive motor programming to help avoid that?
- Pic or Cortex CPU?
- If you clean up the outside wheel rails, by moving the encoder axle bearing to the inside wall of the wheel rail, you’ll be less likely to catch on things like walls, ladder, your alliance partners robot, that you are driving near.
- More feedback sensors on your descoring system might be helpful.
Good luck at Competition.
Aggressive 3x speed up gearing with Dual motors each side on 4" hi-traction tires in rear, HS chained to 4"omnis in front;
Sturdy square frame with 1x5x1 dual steel C channel rails.
No ultrasound/line follower/polycarbonate/hanging/tube nest visible.
Frame may provide nesting to goals.
One 2AH NiMH battery.
Looks like needle pickup can work in front or back.
Probable motor allocations, as far as I can tell.
2L + 2R drive, with L,R encoders for feedback
1 claw clamp w curved wings, but not 4bar, no feedback?
1 clamp lifter on 4 bar arm, with spring assist
2 for arm shoulder joint, 1:9 torque ratio?, w pot feedback
1 for arm elbow joint, 1:5 torque ratio?, w pot feedback
1 for needle roller pickup
---- 10 total motors
Questions/suggestions:
- Don’t your drive motors overheat with this gearing?
- Or do you have advanced drive motor programming to help avoid that?
- Pic or Cortex CPU?
- If you clean up the outside wheel rails, by moving the encoder axle bearing to the inside wall of the wheel rail, you’ll be less likely to catch on things like walls, ladder, your alliance partners robot, that you are driving near.
- More feedback sensors on your descoring system might be helpful.
Good luck at Competition.
Great observations, you have a knack for dissecting a robot quickly.
We do have some polycarbonate on the back of the robot to help protect our cortex and all of the wires plugging into it. This will also serve as the area that holds our batteries. The polycarbonate shield detaches / attaches quickly with zip ties.
A potentiometer was added to the de-scoring mechanism’s 4-bar joint after this photo was taken. We may also add a limit switch or another pot to the claw if necessary.
While we get plenty of height to try for a hang, we are out of motors and don’t see it to be as valuable as scoring & de-scoring tubes.
We haven’t over heated the HS drive motors yet and plan on using the power expander to devote an entire battery to the drive. The robot is right at about 12 pounds. We can lose a little weight in our de-scoring mechanism and arm if necessary, but so far it looks like our weight is ok for our drive’s gearing.
We plan to clean up the outside rails, one of those last minute details.
You didn’t mention Cortex or PIC.
Re power expander: Its not so much adding another battery, as adding another thermal breaker to share current sourcing. PIC is 4A limit, Cortex 2x4A, a power expander adds another 4A. Once the thermal fuses in the supplier, or in the motor, start to open, they open more easily everytime.
Re hanging: Its part of the game, may give other teams a reason to pick you, and is interesting to design for. In the “learning mode post season”, you might consider an add-on module with 2 motors that would implement a hang, as an alternative to the two motors on your descore mech.
We are using Cortex for the first time actually, we like it quite a bit.
We will definitely keep an eye on the robot this weekend during competition; I don’t predict we will trip breakers, but if we do then we will have to either modify our drive’s gear ratio or lighten the robot. We have our drive motors in ports 1 & 2 and 9 & 10 to hopefully avoid tripping them. We will most likely plug all of the motors related to the arm and snorkel into the power expander and have the descore motors on 5 & 6 of the cortex. Thanks for the information!
Regarding hanging, it is absolutely part of the game, just how being able to prevent hanging is part of the game. Not many teams in our area are focusing on it, though I also realize that many teams, including international teams, are doing it quite a bit more. In my opinion, if a team is taking 15 - 20 seconds to high hang, we can probably negate that entirely by descoring two goals and scoring our own tubes on them.
I like the idea of having an option to hang though, so maybe the team can develop that between now and the championship.
I want to start by saying I am NOT actually a motor-design engineer, and that this was my idea – not something from the engineers at VEX. I suggest to teams that they plug their high-load motors into ports 2-9 on the Cortex, using the Model 29 motor controllers. This way, if a controller fails due to continued abuse, it is cheap and easy to fix. If the built-in controllers on ports 1 or 10 fail, it’s a much bigger problem. I suggest that ports 1 and 10 be reserved for mechanisms that are unlikely to have high loads.
We have a dozen Cortexes at the school where I help out and none of them have had bad ports, but I figure it is better to be a little paranoid.
For some reason, now that VEX has given students bigger motors, they are pushing them even harder than before. You know what would happen if VEX provided Ultra-Super-Strong motors, don’t you?
I like it a lot. You’re able to carry a large number of tubes while being able to descore and dump as many stacks as you please. Great job.
That makes two of us on the motor-design engineer comment. We will definitely keep this in mind though, makes sense to me at least. We’ll need to pick up a couple more motor controllers to try it though and as of now the only 2-wire motors we have are the 393’s. I wouldn’t mind putting a 393 on our arm though… wouldn’t hurt. But then we’re outside our HS motor limit… basically we need to order some 269’s.
Thanks Chris. The students really wanted to be able to selectively score from the start and that strategy has been fairly successful. We didn’t have a descore mechanism at our last event though and that proved to be our downfall. Now if I can just get the students to register on Vex Forum…
I think you guys won us over on the claw debate. The thing is beast-mode.
Pictured below is more or less what we will be taking to worlds:
The claw has gone through several design iterations, this being the final one. Major props to the KTOR guys for giving us some inspiration on this final design (haven’t seen your guys’ robot in a few weeks so I’m guessing it has changed quite a bit)
We also finally ran into the problem of tripping our breakers for our 1:3 geared drive train. After troubleshooting the issue quite a bit, we finally just decided to change our gearing to a 5:3 and still use the torque setting on the 393 motors. No more breaker tripping problems since we did that.
Currently we can hold 9 tubes at once, giving us a nice ability to de-score and immediately score, or just throw a ton of tubes under the ladder.
Looking forward to competing at the championship event!
I do feel obligated to point out that a direct drive gearing with 393s in speed mode should be identical to a 5:3 gearing in torque mode. Probably more efficient, too.
You are correct, I should have mentioned the process we went through while attempting to get around the breaker tripping issue. We changed to a 5:3 gear ratio and then changed the 393’s to their speed setting to see if the reduced overall rpm output would help. It didn’t.
So by this point we were running out of meetings, it just made sense to put them back into the torque setting and get more driver practice than spend another meeting re-gearing the drive.
I’ll also say that we looked over the drive quite a bit to see if there was anything causing friction or binding and we couldn’t find anything. We went through a whole competition with the 1:3 ratio and didn’t trip a breaker once. The next competition we tripped them constantly, it was very strange.
did you add any extra “weight” on your robot?
also did one competition have more “dusty” tiles? 
cool looking bot!
one of the rare “needle + claw” designs i have seen!
will definitely want to check you guys out at worlds!
c ya there
No weight was added, we’ve remained in about the 12 pound range. Not sure on the tiles, wasn’t looking at them too much.
Thanks for the compliments, definitely drop by our pit we should have some buttons to hand out.
Here are some ideas why a less efficient external gear ratio might be more robust.
Hypothesis1: External gears have larger gear slack. At startup, the motors can start moving to take up the gear slack, so they are not starting up from stall condition, which is easier on the motors.
H2: Motor bearing has more friction than axle bearing?
A test for these hypotheses might be to compare 1:1 external gearing with motor geared for speed, to the existing 5:3(speed) external gearing with motor geared for torque. This would equalize the external gears and bearings.
well you also have to consider that the internal gear is 160% (1.6:1)
while the external gearing is 5:3 (1.6666666666666666666:1)
theres that extra .0666 difference…
i dont know how much difference that is in real life though
maybe not even noticeable…