06-02-2013 22:50

rcmolloy

Re: pic: 3992 Drive Train

Out of curiosity, what is the thickness on the side plates. My guess is that it looks like 1/8.

06-02-2013 22:50

ksafin

Re: pic: 3992 Drive Train

Correct.

06-02-2013 22:51

sdcantrell56

Re: pic: 3992 Drive Train

Looks very pretty. Seems like you chose to compromise a tremendous amount of strength for appearances though. I hope you are planning on using the bumpers as a structural member

06-02-2013 22:51

dodar

Re: pic: 3992 Drive Train

Just out of curiosity, how come you guys went with belt drive mecanum rather than the normal direct drive?

06-02-2013 22:56

ksafin

Re: pic: 3992 Drive Train

From some collision tests and general observation, we don't see too much of a lack of strength. It's more than strong enough all-around - the only primarily weak point is at the "X" cut out on the long side-plates, but even that is fairly strong and will be stabilized further with standoffs yet to come.



When you say direct drive, do you mean direct attachment to a gearbox?
There were a few logistical reasons we didn't do that, but overall because we had the resources to make a belt drive right away.

06-02-2013 22:58

dodar

Re: pic: 3992 Drive Train

Ya thats what I meant. So it wasnt for like CG or spacing for other things? It was just because you could?

06-02-2013 23:00

ksafin

Re: pic: 3992 Drive Train

CG was one of the points that came up, actually (I forgot about it as I replied to your initial inquiry).

For our climbing mechanism we need as much weight center as possible, so gearboxes at both ends were a barrier.

But otherwise it was less of "we could" than "we could do it given what we have now, or wait to order more parts to perform the same function" so we went with what we had.

06-02-2013 23:05

Gregor

Re: pic: 3992 Drive Train

I'm going to quote my post in the other thread you posted with your practice robot.

Have you actually driven this yet? It is far too fast.

06-02-2013 23:06

rcmolloy

Re: pic: 3992 Drive Train

Good to hear this statement. I could see definitely see some buckling of the material inwards with a solid hit from another robot at the right angle. You guys can keep the aesthetic appearance of the drive super nice with pocketing that benefits not only the appearance but functionality as well.

06-02-2013 23:12

inkspell4

How are you planning on mounting your bumpers?

06-02-2013 23:14

ksafin

Re: pic: 3992 Drive Train

Using special shaped C-Channelish brackets that are ~ 4.8" high and about ~0.75" in. IE, it basically wraps around the side and front plates. Then the top and bottom of the C-Channel have a hole that mounts to the plate itself, and the long part to the bumper plywood.

06-02-2013 23:24

Brandon_L

Re: pic: 3992 Drive Train

I second this, are you planning on having a 30fps drive?

06-02-2013 23:29

ksafin

Re: pic: 3992 Drive Train

We had the previous drivetrain (the 80/20 one) driven, and it seemed to be fine.

I'm not particularly sure what the software team did, its possible that they set a cap on the motor output speed (such as setting 75% as a max for the motor controllers).

06-02-2013 23:31

ksafin

Re: pic: 3992 Drive Train

But I can say for sure that it was nowhere near 30 fps.

06-02-2013 23:32

Gregor

Re: pic: 3992 Drive Train

If so, you are limiting the torque you are getting. This *shouldn't* be bad for mecanums, as I assume you won't be in pushing matches, but it really hurts your acceleration.

06-02-2013 23:34

akoscielski3

Re: pic: 3992 Drive Train

The speed you have on your speed controller's will never help you with torque. You WILL NOT have enough torque to turn your wheels, and decreasing the current won't help that. You need to change your gearing. Have you tested your drive train? If not I would HIGHLY suggest it.

My team had direct driven Macanum's (6in from CIMple boxes) in 2011 and we could only drive forward/backwards and turn, we couldn't strafe because we didn't have enough torque to turn the wheel opposite to eachother. PLEASE CHECK YOUR DRIVE NOW!!! Before it's too late and you're a sitting potato on the field.

06-02-2013 23:44

ksafin

Re: pic: 3992 Drive Train

What do you characterize as not being able to strafe?

Did you not have enough torque to strafe period?

When we tested this on the previous drivetrain, it succesfully strafed, albeit it a slight angle rather than directly perpendicular.

06-02-2013 23:46

akoscielski3

Re: pic: 3992 Drive Train

it would attempt to strafe, and would kinda move, but not really.
There was not enough reduction, usually people would use toughboxes on macanum's we were dumb and didn't
did you have 150lbs on that robot? that's when you need to the torque.

06-02-2013 23:55

ksafin

Re: pic: 3992 Drive Train

What about putting a Mini-CIM in each gearbox for additional torque? Or RS-775's on CIM-U-Lators?

07-02-2013 00:01

Gregor

Re: pic: 3992 Drive Train

Thats not a bad idea, and probably the fastest switch, but switching to Toughbox Mini's would give you the best results.

07-02-2013 00:10

ksafin

Re: pic: 3992 Drive Train

We'll see what we can do.

We'll test with this and a heavy load on the chassis and see if the results are poor enough to warrant the switch.

If anything, I think we'll switch to toughbox Nanos. The mini has a front plate bigger than our actual plate so it'd be an odd fit.

07-02-2013 00:12

Gregor

Re: pic: 3992 Drive Train

TB Nanos are also a good choice, I like them.

07-02-2013 09:37

Lil' Lavery

Re: pic: 3992 Drive Train

The reason it's nowhere near 30fps is because you're pushing your motors way closer to stall than an average drivetrain. They aren't going to come anywhere near the free speed, yet they'll draw more current than a properly geared drivetrain and will accelerate much slower.

07-02-2013 13:19

Bstep

Re: pic: 3992 Drive Train

I doubt that they will realistically achieve this kind of speed with the losses in the gearbox, lack of friction with the carpet, and other losses. If they do in fact reach speeds comparable to 30 fps then more power to them. This years game has no mid-field obstacle and score will be determined heavily by how fast disks can be ferried from the feeder station to shooting position. A fast drive base could be a game changer.

07-02-2013 21:30

Gregor

Re: pic: 3992 Drive Train

This is true, all the more reason to use a greater reduction. Lets say they never surpass 18 ft/s when geared for 30 ft/s (random guess). If they gear for 18 ft/s, you get the same top speed with greater acceleration and torque.