pic: 254 Drivetrain [Archive]

Elgin Clock

16-02-2004, 01:12

Question: What has 6 wheels and goes really fast??

Answer:What is another regional winner - and maybe more????????

Nice looking bot this year guys. I was finally able to watch the Science Channel Documentary on FIRST featuring your team and I was impressed with your team's showing in CA!!

Maybe this year is your year???

LOL, but that's what we all want though, right?? To win the Championships??

See you down in Atlanta hopefully!!

BTW, what's your weight in this picture??? It seems like it's awful close to 130lbs... lol
I see lots o' goodies!!

phrozen solyd

16-02-2004, 02:10

Whew, ok, lots of questions to answer:

Jessica - We've been using the waffle tread material since 2002, and it's worked out great. The stuff's pretty easy to find (it's actually conveyor belt material you can order from McMaster), and it provides good traction on almost any surface.

pras - There are only 2 double solenoids. The other 6 are singles.

Elgin - Haha, thanks for the props. We weighed the complete robot last night, and it was exactly 129 pounds...it doesn't leave too much weight for powdercoating, which we think will add about 2 pounds. Looks like we might have to resort to desperate measures like cutting off a couple cm off each bolt or cheese-holing everything :D

pras870

16-02-2004, 02:13

If you're using 8 solenoids, I'm guessing you're gonna be going for multi-positioning with the pneumatics this year?

activemx

16-02-2004, 05:36

yea lets hear a little more about the beafy pnumatics this year you guys are using along with 60? Also i am curious about your drive train!

phrozen solyd

16-02-2004, 19:43

Our drive train is powered by 2 Atwoods and 2 drill motors, with 2 speeds, pneumatically shifted. The gearbox is a modification of 60's design from 2001. The 6-wheeled base is modeled after Team 25's awesomely fast and agile machine from 2003, with the center wheel on each side lower than the others by 1/8 inch, so that the robot rocks back and forth. This shrinks our wheelbase to about 13", allowing us to turn quickly.

The numerous solenoids on our bot are actually used for many different functions. One is for the cylinders that shift our gears, two are for our ball corralling "flaps", one extends our bar-grabbing hook, one actuates our gripping hook for latching onto the mobile goals, one extend the "whips" that knock off the 10-pt balls, one double solenoid actuates the claw for manipulating the large balls, and the other double solenoid engages and disengages the winch for pulling ourselves onto the bar.

sanddrag

16-02-2004, 19:45

Our drive train is powered by 2 Atwoods and 2 drill motors, with 2 speeds, pneumatically shifted. The gearbox is a modification of 60's design from 2001. It must have been a fairly major modification as the drills were different then and we didn't even have Chias.

NateBot16

16-02-2004, 19:57

Looks great. Everything looks super organized, even down to the different color pneumatic tubes. Good luck this year.

eugenebrooks

07-04-2004, 23:51

sanddrag

08-04-2004, 00:23

activemx

08-04-2004, 00:57

I got to see 60's robot REALLY up close (like I stuck my head in it) at Phoenix and it is amazing how everything is packed in there. They even have this little leadscrew mechanism on their PTO winch that winds the cable evenly so it doesn't wind on top of itself. Anyway, on 60's it appeared that the 4 corner wheel's are toe'd in a bit. I was wondering what is the effect of/ reasoning for this and how does it affect the chain/sprocket alignment? Thanks.
Yea thats why they have the 6 wheel caster driver just like 25 last year. They have an offset for the center wheels a little lower then the front and back wheels. so the front and back are pretty much like a powered caster. It tips over on to either front or back generally depending on the direction its driving in. This allows them to have a great turning radius.

sanddrag

08-04-2004, 01:05

Yea thats why they have the 6 wheel caster driver just like 25 last year. They have an offset for the center wheels a little lower then the front and back wheels. so the front and back are pretty much like a powered caster. It tips over on to either front or back generally depending on the direction its driving in. This allows them to have a great turning radius.Yes, I know all about the 6 wheel drive and the center one being a little lower. It is just like a Home Depot shopping cart. What I was asking though is if the 4 outer wheels have toe in. On Kingman's bot, it did appear to have about 2 degrees of toe in per wheel (or "toe out" depending on your perspective of front and rear of the vehicle)

phrozen solyd

08-04-2004, 01:49

How do you fasten the tread material to the aluminum so it doen't peel off?

The treads are held into the wheels with screws.

Travis Covington

09-04-2004, 02:51

Yes, I know all about the 6 wheel drive and the center one being a little lower. It is just like a Home Depot shopping cart. What I was asking though is if the 4 outer wheels have toe in. On Kingman's bot, it did appear to have about 2 degrees of toe in per wheel (or "toe out" depending on your perspective of front and rear of the vehicle)

Just a guess, but I am willing to bet it wasnt intended to be like that (if it truely is more than just an optical illusion). My guess is that their tensioning devices have a tendancy to pull the sprocket side of the axle closer to the gearbox (in essence making it look like the wheel is pointing inward) I would guess it is partially due to the way they canteliver the wheels out, but also the way the chain is wrapped on the inside sprocket and how the wheel is moved on slotted mounts to tension it)

When I took a "close" look at 254's robot, I did not notice it at all. I would be willing to bet it was not intentional if it was anything more than an optical illusion...

254/60 care to enlighten us?

phrozen solyd

09-04-2004, 02:58

Just a guess, but I am willing to bet it wasnt intended to be like that (if it truely is more than just an optical illusion). My guess is that their tensioning devices have a tendancy to pull the sprocket side of the axle closer to the gearbox (in essence making it look like the wheel is pointing inward) I would guess it is partially due to the way they canteliver the wheels out, but also the way the chain is wrapped on the inside sprocket and how the wheel is moved on slotted mounts to tension it)

When I took a "close" look at 254's robot, I did not notice it at all. I would be willing to bet it was not intentional if it was anything more than an optical illusion...

254/60 care to enlighten us?

I can't speak for Kingman's bot, but on ours, the wheels don't normally toe out like that. Unless the chains are overtensioned - we noticed this on Saturday at SVR, and fixed it soon after.

Natchez

09-04-2004, 04:37

Posted by phrozen solyd
The treads are held into the wheels with screws.

How do you cap the screws? Do you countersink them? We tried all kinds of things last year but we could never convince ourselves that we weren't violating the "Traction devices may not have surface features such as metal" rule. Plastic fasteners were not as good as glue. Of course, that's not what the four students that went around with their fingers stuck together for two weeks thought.

Thanks for the help,
Lucien

Dennis Jenks

09-04-2004, 09:43

How do you cap the screws? Do you countersink them? We tried all kinds of things last year but we could never convince ourselves that we weren't violating the "Traction devices may not have surface features such as metal" rule. Plastic fasteners were not as good as glue. Of course, that's not what the four students that went around with their fingers stuck together for two weeks thought.

Thanks for the help,
Lucien
If you look at the wheels you can see that they are of two piece construction. Each side has a small lip on it so that when they are sandwiched together they grip the tread and keep it in place. The wheels are tapped and we use a pair of small pan head screws, not to hold the tread down but simply to keep the tread pulled tight. These are positioned so that they sit in the waffle well below the tread. If you are using screws to physically hold the tread down then you would need to use a larger head which may not sit deep enough to pass inspection.

Glenn

09-04-2004, 10:46

There is no toe in unless something got bent, maybe I had better check! But seriously no toe in or out is not required to make the robot maneuverable.

The six wheel drive is the best system we have found so far it is simple, light, maintenance free, very little loss of horse power due to mechanical friction, and you will get the same traction as any other system.

Many people think you get better traction with a track system, not true. The only two things that effect traction are weight and the coefficient of friction of the treads. Surface area is not part of the formula. Ask your engineer.

Cory

09-04-2004, 17:15

ChrisH

09-04-2004, 18:10

Correct me if I'm wrong, but doesnt surface area matter in instances when the tread material deforms a surface such as carpet, and "interlocks" (for lack of a better word) with the individual fibers?

Cory

Glen is right as far as he goes. But sometimes the co-efficient of friction varies with either the load or the load divided by the area. This usually happens in systems where something is deforming enough to be noticable. Generally these interactions are also non-linear and need to be quantified emperically. At least that is easiest for our type of systems.

There were a couple of very good white papers posted by IBApril180 about this a while back that address this question. Here are the links.

http://www.chiefdelphi.com/forums/papers.php?s=&action=single&paperid=120
http://www.chiefdelphi.com/forums/papers.php?s=&action=single&paperid=119

sanddrag

09-04-2004, 18:36

There is no toe in unless something got bent, maybe I had better check! There definitely was some toe at Phoenix. I'm willing to bet it is what Travis said above.

Collin Fultz

10-04-2004, 00:06

would it be possible for 254 or 60 to post a white paper about their drivetrain? of course...i guess I can always see it in NEWTON! or just e-mail me information

collinfultz@att.net

SteveK254

10-04-2004, 01:01

would it be possible for 254 or 60 to post a white paper about their drivetrain? of course...i guess I can always see it in NEWTON! or just e-mail me information

collinfultz@att.net

Hi Collin,

We don't plan to generate a white paper on the transmission, due to focusing on our collaboration white paper. However you are welcome to any of our drawings as well as an invitation to unlimited questions. Please stop by our pits (60/254) at the Nationals and we would be happy to answer any of your questions. You can also see the working model. :p

Stephen Kowski

18-04-2004, 23:20

I'm wondering how 60/254 packed their solenoids together sooo tight.....is there room for the hose/fittings....can anyone elaborate...?

Travis Covington

18-04-2004, 23:29

I'm wondering how 60/254 packed their solenoids together sooo tight.....is there room for the hose/fittings....can anyone elaborate...?

It is a SMC solenoid manifold. It is available through any SMC distributor (Motion industries etc). It has a common pressure input, and has ports for the solenoids on top and quick disconnect fittngs on the side, 2 for each solenoid. Block off plates are available if you decide to remove a solenoid or cannot purchase a small enough manifold. (in our case we needed ports for 7 solenoids, and thus had to purchase an 8 port manifold, we blocked off the 8th port) I think they come in manifolds of 2 and larger (maybe 4, I can't remember) and I recall them being in multiples of 2.

Hope that helps!

Collin Fultz

20-04-2004, 00:32

Hi Collin,

We don't plan to generate a white paper on the transmission, due to focusing on our collaboration white paper. However you are welcome to any of our drawings as well as an invitation to unlimited questions. Please stop by our pits (60/254) at the Nationals and we would be happy to answer any of your questions. You can also see the working model. :p

i'm terribly sorry that I didn't get to see the model...i was either talking to judges, trying to shoot balls, or ogiling over the HOF displays...plus i didn't see your post until today. if there is any other way to get the information, digital pictures, gearing ratios, or anything else that would be great...thank you!

magnetnerd

04-05-2004, 12:28

How did you make your tires, or did you order them? Specifically, the wheels and not the waffle treds.

Cory

04-05-2004, 12:50

They made them. I believe it was explained elsewhere, but they are 3 separate pieces. The hub, and then the two side plates that sandwich the hub. By design, the two pieces trap the tread in the wheel well, so that screws aren't needed to fasten it.

soezgg

04-05-2004, 16:52

bad $@#$@#$@# wheels.

Solace

04-05-2004, 17:23

just a question, but did you use open or closed loop for steering? because it would seem that at the speed the robot usually goes, it's turning speed is much too fast for it to be controlled by open loop.

JessR

04-05-2004, 18:25

:confused: But I think if the chain is open loop, you can only drive a few feet befor it falls off? The closed loop works much better for that.

Cory

04-05-2004, 18:33

I think he's talking about programming...

Aignam

04-05-2004, 18:53

Gee, Kingman had something really similar to this... ;)

Greg Ross

04-05-2004, 19:18

:confused: But I think if the chain is open loop, you can only drive a few feet befor it falls off? The closed loop works much better for that.
Wiseacre engineer. :)

Solace

04-05-2004, 21:29

i mean in programming. I know that 254 and 60 both did the steering wheel/joystick combo, but how exactly did they do the controls in programming?

Greg Ross

04-05-2004, 22:31

i mean in programming. I know that 254 and 60 both did the steering wheel/joystick combo, but how exactly did they do the controls in programming?
Yes. I'm still waiting for them to answer too. :) I think I know the answer (which is that it's purely open) but I'll wait for the proverbial horse to speak. ;)

Travis Covington

04-05-2004, 22:34

Yes. I'm still waiting for them to answer too. :) I think I know the answer (which is that it's purely open) but I'll wait for the proverbial horse to speak. ;)

Jay Lundy talked about this in one of the other threads.. Well kinda...

http://www.chiefdelphi.com/forums/showpost.php?p=248213&postcount=14

Greg Ross

04-05-2004, 23:05

Jay Lundy talked about this in one of the other threads.. Well kinda...

http://www.chiefdelphi.com/forums/showpost.php?p=248213&postcount=14
I guess I was wrong. Alex Baxter's post (10intheCrunch) (http://www.chiefdelphi.com/forums/showpost.php?p=248211&postcount=13) in the same thread comes close to saying that they (254) used closed loop control.

Cory

04-05-2004, 23:07

Greg- as I understand it, they used closed loop control for their arm.

Jay Lundy

05-05-2004, 09:36

The drive train loop is open when you are turning (see the other posts linked to for more information about that), but the straight drive loop is closed. The striaght drive code is run whenever the steering wheel is at the center position +- a certain deadzone. It uses the proximity sensors to get a speed error (left count - right count). For the robot to go straight, the speed error should be 0, but it usually isn't, so the speed error is multiplied by a constant and subtracted from whatever side is going faster.