31-10-2014 16:18

Ty Tremblay

Re: pic: pre season chassis prototype

This looks very similar to what the kit chassis from AndyMark is rumored to look like. Using it for your frame and adding your improvements would save you a lot of time.

31-10-2014 17:02

75vs1885

Re: pic: pre season chassis prototype

plot twist: he is andy mark!!!!

31-10-2014 18:00

Deke

Re: pic: pre season chassis prototype

It looks good!

Suggestion: Place your encoders on a drive chain idler sprocket or gearbox output for positive engagement. The idler wheels may slip on the ground and provide inconsistent or false information to the encoders. It could save you some money and weight as well.

31-10-2014 18:01

AdamHeard

Re: pic: pre season chassis prototype

If you're looking for precision, the idler wheels actually yield much better results (especially in turns).

If you're not looking for huge precision (most teams aren't) or aren't trying to keep position during turns, doing them on driven wheels is fine (and what most teams do).

31-10-2014 18:04

Deke

Re: pic: pre season chassis prototype

Interesting, I haven't heard of that before, good to know!

Does the placement need to be close to the center or dropped wheel then for good contact with the ground? I would think the rocking of the tank drive could impact the traction of the idler.

31-10-2014 18:06

AdamHeard

Re: pic: pre season chassis prototype

We actively powered them down with a 5/16" bore pneumatic cylinder.

We had them installed far left/right extreme for precision in turns, and roughly centered between the front/middle wheels on our heavier side.

It allowed us to do 6 or so linear movements and drive about a combined 40' (and probably 720* of turns) and still measure within 2" (maybe better) our final position.

31-10-2014 18:10

Deke

Re: pic: pre season chassis prototype

Gotcha, is the wheel scrub from the skidding of the tank drive the cause of the reduced accuracy of drive encoders?

31-10-2014 18:19

AdamHeard

Re: pic: pre season chassis prototype

Yes.

So for linear movements only, especially if the control keeps you from apply full voltage at standstill (and skidding out briefly), you get plenty of accuracy w/ encoders in the driveline.

For keeping position during turns, which most robots don't need, they will lose position.

31-10-2014 22:59

cad321

Re: pic: pre season chassis prototype

Yes the idler wheels are for increased precision in autonomous and to ensure they remain on the ground while rocking I plan to have a spring attached to the idler pivot bracket to ensure it is always touching the floor.

01-11-2014 00:41

AustinSchuh

Re: pic: pre season chassis prototype

Grabcad is pretty cool. I only run Windows in January and February, but it let me see what you are doing and give some feedback.

You really need some flanges on both of the bottom edges of your drive rail. The bottom edges are quite weak right now. While you are at it, tie those into your belly pan. They will bow under load.

The flanges on the top of your frame rail are way too short. They look like 1" flanges or shorter. We have bent frame rails that had 1.5" long flanges with another flange off the back side of that flange for even more strength.

Also, any reason why you are using angle iron in the corners rather than adding a flange to the side rails?

Nice work making the two side rails the same part. That will help with machining. Most people miss that one.

Consider adding access holes in the outside of your frame for bolts holding your transmissions on. Being able to get a wrench in there easily is very useful.