29-10-2008 15:10

Zflash

Re: pic: FRC488 - 80/20 Drive Train

Looks solid. Do you have an estimate of what it wieghs in it's current design state? Also the question I really have is are you using two servos to shift the dewalts, if so have you done any testing with it?

29-10-2008 15:13

Madison

Re: pic: FRC488 - 80/20 Drive Train

It's 22.25 lbs. as shown. It's missing chains, obviously, and some brackets and fasteners required for mounting the gearboxes. It'll probably hit 28-30 lbs. with those things and with an ABS base for control system components.

We had some trouble shifting the DeWalt gearboxes with a single servo last season, so we've been thinking about adding the second to each. We haven't tried it yet.

29-10-2008 15:18

sdcantrell56

Re: pic: FRC488 - 80/20 Drive Train

How are the bearing blocks held in place on the 80-20? Other then that it looks like s very good simple drivetrain. I guess you didn't see a need for the second set of cims last season?

29-10-2008 15:20

Madison

Re: pic: FRC488 - 80/20 Drive Train

The bearing blocks are 80/20 pivot pieces. We're going to open the pivot hole from 1/4" to 1/2". They tie in to the 80/20 from the top using their existing mounting holes. They're part no. 4181.

We had some trouble shifting the Dewalts and made no use of 3rd gear last year, so we were a bit slower than I'd hoped. As always, the game may change what we put into the drive, but if we can avoid pushing matches again, we'll probably stick with 1 CIM per side. Similarly, it's shown with four traction wheels, but we may again use just two and put omniwheels back onto the corners.

We expect to have our hands full with the new cRio system, so anywhere else that we can remove complexity, we'll do it.

29-10-2008 16:27

Lowfategg

Re: pic: FRC488 - 80/20 Drive Train

I would never mount bearing blocks to 80/20 using the T-slots. They slip way to easy under the strain that is being placed on them. Drill through the 80/20 and bolt them right into it.

Thats just my opinion.

29-10-2008 16:30

Madison

Re: pic: FRC488 - 80/20 Drive Train

Well, the simplicity of this design is that 80/20's slot system substitutes for chain tensioning mechanisms. If we were to eliminate that advantage, we'd change the design entirely.

We haven't used 80/20 much, but have had success with it staying where we put it as long as the bolts are properly tightened.

29-10-2008 16:36

Lowfategg

Re: pic: FRC488 - 80/20 Drive Train

Well I guess then the T-slots are nice to have for that, but make sure to keep up with those screws. My experiences with 80/20 last year were not that great, but some people like it.

29-10-2008 16:38

AndyB

Re: pic: FRC488 - 80/20 Drive Train

Madison, this looks like a very simple, easy to put together design.

The only weaknesses are, in my eyes, angular clearance for ramps.

Assuming the wheels are centered at 3/4" above the 8020, and positioned about 3.5" back from the inside of the front rail, that only gives you about 17 degrees of angular clearance. But for a flat-field application, this looks like a very nice drive that you could easily modify at any point.

29-10-2008 17:00

Andy Baker

Re: pic: FRC488 - 80/20 Drive Train

You are right, as the t-slots do allow for the t-slot nuts to slip. However, another thing to do besides drilling a hole is to just add another t-slot nut to back up the initial nut. If there are 2 in a row, and both are tightened well (and checked routinely), then the slippage does not happen.

Andy

29-10-2008 17:02

qwertyuiop[]\

Re: pic: FRC488 - 80/20 Drive Train

Why do you need a six wheel robot if you are using omni's? also watch out, 8020 bots get heavy, fast. on our 2006 robot we essentially built our entire robot out of 8020. There is alot of un necessary weight in 8020. Do you really need all for slots in the metal? it is very strong and is very fast, so if your going for a fast build this is what you want. Also, what size 8020 are you using, the inch or half inch?

29-10-2008 17:05

Cory

Re: pic: FRC488 - 80/20 Drive Train

8020 itself is not actually particularly heavy. A very long time ago I did the math to see for myself and the 1010 series is almost exactly the same weight as 1x1x0.125" Al tube. The weight difference comes from all the connectors, t-nuts, and fasteners.

29-10-2008 17:24

Madison

Re: pic: FRC488 - 80/20 Drive Train

Yeah, this can clear a 17* angle, but nothing greater. If there's anything steeper, we'll figure out how to deal with it as a special circumstance -- or we'll turn the chassis upside down.

6WD does a bit better at getting over obstacles -- whether those be field elements or game pieces. Again, the simplicity here is that we can swap in any sort of wheels we'd like easily -- and we can add or remove "drop" as necessary.

The profile shown is 1010 -- 1" square.

We have a good history of building reliable drive trains that are also very light. We've been working lately on minimizing manufacturing time, so we're conceding some of the features that make things light in favor of those that make them quick to build.

That'll be Fred's job. Fred is the advocate for 80/20 on our team and even then we use it sparingly. If it breaks, it's his fault.

29-10-2008 19:12

Andy L

Re: pic: FRC488 - 80/20 Drive Train

Our base was almost exactly like this last year and we had a few problems because the middle wheel was more dragging than helping

We removed it for Cal Games this year and it made a noticeable change when driving, but other changes we made to the robot made it hard to actually compare.

29-10-2008 19:32

s_forbes

Re: pic: FRC488 - 80/20 Drive Train

Great design, you could build that whole thing in a day! Well, except for the transmissions... I like how you completely eliminated the need for any tensioning device with the use of 80/20, very clever. The only "problem" I can see with it is the lack of any diagonal bracing, but I suspect the electronics board or the top half of the robot will account for that. I doubt that you'd let anything like that slip by.

29-10-2008 19:57

=Martin=Taylor=

Re: pic: FRC488 - 80/20 Drive Train

I almost slapped an "FRC100" tag on it... But I'll resist...

Kinda reminds me of some of our drive ideas.

Our team doesn't actually ues 80/20 connectors, we use bolts with filled off heads, or toilet bolts. This spares us the aggravation of dealing with sliding, twisting, falling-out 80/20 biscuits.

That's very clever.

But how will you reache the bolt if the axle is in it? And how will you tension your chains if the wheel isn't in there?

We usually buy a kind of extrusion that 80/20 sells that already has the flat groove, and then we put the holes in it ourselves.

29-10-2008 20:00

Gdeaver

Re: pic: FRC488 - 80/20 Drive Train

Instead of the pivot blocks, you might look at the 10 s 2497 parts. They may handle the shock loads better and the 2 " and 2 T nuts grip better. The raw stock can also be purchased if you want to cut your own. Watch the 8020 garage on e-bay. Any reason for the 1/2" shaft? For a dead axle a 3/8" grade 8 bolt is enough and lighter.

29-10-2008 20:21

David Brinza

Re: pic: FRC488 - 80/20 Drive Train

Team 980 has used the 1010 size 80/20 extrusion in our robot chassis for years now. Our six-wheel drive with Andy Mark SuperShifters goes together very quickly and getting proper tension on the chains is easy. We use simply-machined phenolic bar stock as pillow blocks for the axles with fastener through-holes spaced at 2". Instead of the more expensive drop-in T-nuts, we use economy triple T-nuts. We haven't had problems with slippage of the fasteners, we put a bit of Lock-tite on the threads and torque them down.

I don't see us changing this approach for sometime to come.

Oh yeah, you can save some $$ if you buy from 8020 Inc Garage Sale on E-bay.

29-10-2008 20:33

Rich Kressly

Re: pic: FRC488 - 80/20 Drive Train

M-
We use the t-slots until we have chain cut and drill through and bolt once everything is in a final position. The bearing blocks we have are slotted and allow for minor adjustments if necessary. Our t-slots always wound up sliding in the 80:20 and wound up being more maintenance that it was worth without bolting through. After 4 years of this, my advice would be to go ahead with the t-slots and if they never slip, great - but leave yourself space to drill and bolt through just in case. It's a quick enough move with a hand drill and some 1/4-20 bolts.

29-10-2008 23:29

Rich Kressly

Re: pic: FRC488 - 80/20 Drive Train

Yep, no doubt about that one. We spend a little extra to save some fastener weight by using drop ins and the hidden corner connectors, but overall we choose the convenience that 80/20 brings and deal with the "extra" weight.

29-10-2008 23:46

dtengineering

Re: pic: FRC488 - 80/20 Drive Train

I'm not entirely convinced that this is either stronger or lighter than basing a drivetrain on the KOP frame from IFI, but I do appreciate its simplicity.

Mind you, I'm not counting on having that in the kit this year, and have come to expect something innovative each year from the Xbots in the way of an off-season drive system prototype/model. I also base my KoP drivetrain comments on our experiences with it, and our team has in-house access to mills and a TIG welder that allows us to do a fair bit of "hot-rodding" to it.

I do like how this simplifies the chain/belt tensioning. I would suggest adding a second block, inboard of the two outboard axle mounts that has a bolt hole drilled and tapped parallel to the frame rail and is mounted to the frame rail in a similar way to the axle mounts. This way a bolt run through the tapped hole could be used to press on the axle mount and tension the chain very carefully, then the axle mounts could be tightened. Once the appropriate location for the second block was determined, then it could be welded in place, and the tensioning bolt could be secured with a lock nut. This way even if the axle mounts' bolts were to come a little bit loose the chain would not come loose.

See you in Portland and (yes!) Seattle!

Jason

30-10-2008 08:15

Gdeaver

Re: pic: FRC488 - 80/20 Drive Train

Our team has used the KOP frame and 1010 extrusions in the past. The Kop frame and a 1010 frame are heavier than needed when the bumpers are added in. The KOP frame is time consuming to assemble. Pop rivets help. The 1010 frame most likely will not be ridged. It will be much heavier than needed. Last year our team used the 8020 quick frame material with the press in connectors. We used the flanged stock were possible . the the flange recesses provide a mounting for clear poly panels. The way we attach the bumpers prevents the tube from separating from the connectors. I was worried about the plastic connectors but, they held up fine. To mount the axles we use 2 2.5" pieces of 8611 panel mount extrusion to make clamp on mounts for 3/8" grade 8 bolt dead axles. With the 1/16" tubing allot of weight was saved and it went together very fast. Our team intends to use quick frame again unless some thing better is in the kit. Nothing like whacking a frame together.

30-10-2008 18:31

techtiger1

Re: pic: FRC488 - 80/20 Drive Train

Madison as always great cad work. I like the simplicity of the design and wish my team would try Dewalt transmissions but anyway. I do have one question I noticed you always use omni wheels up front why ? I don't have an issue with it as it seems to work fine for you but I'm just wondering. Looks great can't wait to see it in real life if 488 and 1251 both make it back to Championship in 09. Good luck and nice slick designing.

30-10-2008 18:35

AdamHeard

Re: pic: FRC488 - 80/20 Drive Train

looking good, just a few questions.

Is the switch to 6" wheels so they can be entirely COTS?

Have you considered running 2 CIMs per side? If so, would you just mirror one of those CIM mounts to the other side?

Where do you foresee attaching mechanisms/superstructure?

31-10-2008 07:16

qnetjoe

Re: pic: FRC488 - 80/20 Drive Train

Great CAD work it looks really similar to our gen 8 reference CAD design that we release every year. Take a look:

http://uncert.mines.edu/research/csr/FIRST/Gen8.htm

If I may, I would look at some stamped pillow blocks like the AMI Bearings BPP1-8 rather than try to modify some of the 8020 parts for your bearing blocks. Here is a link to there catalog page 15

http://amibearings.com/pdfs/mounted_normal.pdf

31-10-2008 14:17

Madison

Re: pic: FRC488 - 80/20 Drive Train

I think that, when we start to get to the point where we're adding band-aids -- little blocks here and brackets there -- it's best to start again and reexamine the requirements of the drive.

This design's efficacy, for our needs, is wholly contingent upon using the 80/20 slots for tensioning. If it turns out that it's not effective in that capacity, we'll go back to other designs that, while requiring a bit more time and effort, have worked for us in the past.

I wonder if sometimes, people become so attached to an idea that they'll go to the end's of the Earth to avoid acknowledging that it's not the best idea for their circumstances. I think that happens to me sometimes and I'm trying to get better at recognizing that and knowing when to start over.

31-10-2008 14:39

Madison

Re: pic: FRC488 - 80/20 Drive Train

The switch to 6" wheels came from a desire to use dead axles. It's a bit more challenging to make dead axles work with 4" diameter wheels -- as there's less area in which to bolt sprockets to the wheels and the also because the chain, potentially, runs very low to the ground. It depends upon sprocket choices, of course.

If we chose to add a second CIM motor to the drive, I'd have to reevaluate the use of the Dewalt transmissions. I haven't yet compared the resources needed for two Dewalt transmissions per side against those required for, say, a single AM setup.

Mechanisms will also bolt to the 80/20. For the past two years, we've built structures that are pinned in place rather than bolted, so that we can remove things quickly.

We've used omniwheels in all four corners for the last several years. It makes the machine very easy to maneuver, but we sacrifice some ability to stay put. It's worked well with our strategy so far, but we've always retained some ability to swap in traction wheels. I avoid the typical "drop" because that rocking motion can translate to some significant movement at the end of appendages and it makes controlling them that much harder.

31-10-2008 16:55

M. Mellott

Re: pic: FRC488 - 80/20 Drive Train

Elegant in its simplicity--very nice.

Is this to be used as a test bed for trying new ideas, or do you think we could see Team Xbot using this in a future competition robot? I only ask this because of the additional weight of an 80/20 frame and fasteners, along with other issues with the potential of loosening fasteners.

01-11-2008 07:29

Alan Ing

Re: pic: FRC488 - 80/20 Drive Train

Hi Madison,

It was great meeting with your team at the Microsoft complex last year. I must say other than the competition itself, it was the highlight of our Oregon/Seattle trip.

I am sure you know what you are doing when it comes to your frame designs as we have seen your work and this is really directed at teams in general.

Anyway, our team has been using the Bosch Rexroth extruded profile system (similar to 8020) for 7 years and except for the fact that the profiles are a bit heavier than the equivalent size 1/8" wall square or rectangular tube, we have nothing but good things to say about it and have learned over the years how to exploit its features making some pretty spiffy lifts and robot frames.

One thing we really like about the Bosch system, is that the rotating T-nuts drop in from the side and have ribs that bite into the profiles making them less prone to slip which is a problem that was brought up in this thread. Slipping T-nuts have been a rare problem for us and our machines have taken a lot of hard hits over the years. Also you can premount the T-nuts and screws to the component itself, and drop it in along the extrusion at any point.

8020 has a much larger selection of accessories than the Bosch 1" system and in all fairness we have never tried them. 8020 also has a drop in nut but they are different and if I am not mistaken, they have to be rolled in unattached to the component you are trying to mount and cost something like $2.69 each compared to the $0.60 Bosch unit.

Not trying to knock 8020, it is a great product. Many teams use it with great success. If you are worried about the T-nuts slipping, you might want to give the Bosch units a once over. Hit them hard enough and they will move too, but I believe the ribbing really helps keep them in place.

Alan

03-11-2008 22:59

Ben Piecuch

Re: pic: FRC488 - 80/20 Drive Train

I'm posting to this a bit late, but we've used a similar 8020 setup to this for many years, so I thought I'd add some notes.

You may find that there's a lot of corner to corner twist in the frame alone. When you add any upper structure, it greatly stiffens it up. But, the 1010 extrusion, and especially connectors, leave a bit of stiffness behind compared to a welded 1" x 1" tube frame. (This is a similar problem with the kitbot chassis)

The bearing blocks, while convenient to use as tensioners, are always a nightmare to keep aligned. When you need to switch wheels, you'll need to un-tension the chain, remove the shaft, and reassemble after. It's not so simple. In our experience, even when you think you have an aligned wheel and axle, it's still not perfect. We've thrown more 35 pitch chains with a similar setup than I care to admit.

I've always found approach angles to be difficult (re: shallow) when you have a minimum of 1" of frame in front of the wheels. There aren't too many ways around this issue when using a frame like this.

Carefully plan out where you'll be adding extra mounting points, and pre-assemble some T-nuts into the rail. The drop-in's are a pain, especially when you need to line them up with the mating component. 8020 makes some T-nuts with a set screw, to hold them in position. I highly recommend those on any vertically oriented channel.

One last note, when cutting 8020, always square off the edges. Chop saws work well, as do milling machines. A band saw, however, does not.

Finally, for all those weight conscious folks... The 8020 brand, 1010 extrusion is within 2% of the weight of a 1"x1"x.125" aluminum tube. The bending stiffness is also withing 2%. Torsionally it's not quite as good, but a frame doesn't see the twist like an arm/appendage would. The connections are the key, and using the double-anchors on all corners are your best options (and lightest weight...) While the connectors do add weight, you're talking around 1-2lbs when done correctly. Using the proper material for your upper structure is the key to weight savings when using 8020.

Simplicity in design and manufacturing is always good, especially if you can make it as reliable as a highly machined and labor intensive chassis. I highly recommend building a practice bot and (literally) shaking out all the issues. Oh, since we only build one 8020 frame with the new bumper requirements... Figure out a way to mount the bumpers prior to week 6. Best of luck!

Bengineer

07-11-2008 12:59

Madison

Re: pic: FRC488 - 80/20 Drive Train

I posted this in another thread, but if up to 100 people are interested in viewing the Solidworks files, you can download them in an archive from here: https://www.yousendit.com/download/Y...WlR0d0ZFQlE9PQ

They're made in Solidworks 2008.