Scuse my stupidity, but whats a translational spinner? I could interpret that as either something like Y-Pout, or something like Tip-Top, but Ive never heard the term before.
--
Fluppet
It is similar to Tip-Top in that it uses the power, and rotation of the disc to create a linear (forwar/backward/side to side) movement of the robot. It is quite a complex idea involving many parts...
Nope, its simlar to Y-Pout and why-not. The PDFs of the real robots articles can be downloaded from their website, and so quite a good job of explaining the principle. If you have Pro Desktop, I can email you a CAD file demonstrating how the mechanism works (in that you can move all the parts around and watch the wheels turn in and out).
Stupid me (*slaps himself*)...
A Y-Pout style spinner would have the problem in that you cant just decide you want to go forward, and go forward. You wouldnt be able to move in any forcefulway (trying to push someone) from a standing start as, if the motors are high rpm, thus less torque (Ed put awy the lem
Then again they may have more advantages than disadvatages. But seeing as I may be getting some sponsorship from a Bristol-Based CNC company... Id like to go the walking route to a MAJOR featherweight spinner.
Or you could try making a simple robot that works?
all wrong, it is not a tip-top and not a y-pout
its something, different
still can be described as an transitional spinner, but not like either of those 2.
See the 2004 FWS, thatll be a simple bot...
:-) If everyone was assuming a different thing, I now feel a little less stupid!
Eddy: thanks for the link (once Id worked out that their meant teamwhyachi.com, not realrobots.com) - better explanation than last time I looked at it. No luck with the CAD Im afraid, but I get how they work (although before I read the RR articles Id assumed the cam was spun by a motor running in anti-sync with the frame rather than being on a central mini-bot - *weird* way to do it). Since Ive never seen them run, I look forward to your implementation - from a safe distance.
Ewan: as with all noncircular horizontal spinners, the stuff them in the corner of the arena and keep them pinned there approach is pretty effective, so the motion is a problem, yes - especially when you rely on spinning to locomote. Although less so if youre running with 3 LEMs or 3 magmotors, as the Team Whyachi robots do (they could probably bounce themselves out).
Tip-Top is a very simple idea, in contrast - two lifters at the rear wheels and the weapon motor. Very pleased to see it at the Rumble last year, having remembered it from an early Robot Wars. Interesting engineering design. Never going to have much traction, of course, but they have my approval for doing something different.
2004 FWS?
Mike: You can, in theory, make a much mechanically simpler version of Y-{pout|not} just by varying the power to the motors mid spin - that is, dont change the pitch of the wheels, change their speed. I believe people have tried it for Stinger-like devices (allowing the robot to spin while moving). Im sure its a faff electronically, but Im surprised nobodys taken it further - it cant be *that* hard to do. (Do gyros work when you spin them fast, or would it need to be kept still-ish?) Er, said with my Ive not yet built a robot hat on, of course. Stopping the batteries from dying on impact is another matter, too - someone needs to market impact-proof batteries (internal bracings?), and theyll corner the market.
It also strikes me that the mechanics needed to fit all the motors (including the drive) to the shell of something like the Typhoon family, thus increasing the weight of the spinner component, arent all that complex. Might be heavy enough that you may as well leave the motors in the middle, though.
James: Ill look forward to seeing it when youve created it! (Im keeping my designs to myself until theyre built too - I dont mind someone copying my robot, I just dont want them to do it until Ive built it!)
--
Fluppet
Andrew-
Appologies for being unclear about the link)
I am in complete agreement that its a *wierd* way to do it, and am actively looking at other options. However, electronics is really very very difficult to impliment. Yes, the electronics are fast enough, but the motors simply cannot increase and decrease their revs 30 times a second- that kind of speed is approaching a usable PWM frequency for larger motors, so clearly the overall affect will just be that the motors see the average voltage given to them.
Batteries are a problem, but in our tests, wrapping the batts in low density foam, and placing that package into a higher density foam (like the stuff smaller flight cases use for cameras/Txs/telescope parts etc). However, we have not run our rig above 500rpm, as its only MDF and angle ali put together in a bit of a rush. Also I feared for my life when running it at that speed.
Another interesting factor to note is tyres- use tyres towards the springier end of the spectrum, because theyll give some suspension, which makes the whole thing spin more smoothly- when we tried with harder wheels, and hit a lump in the floor, the thing started dancing around due to the gyroscopic forces, and any hope of steering it went out the window- the only way to recover was to spin down then spin back up agin.
The most importent part of the design is undoubtedly getting the cam mechanism as smooth as possible. We orgionally had a piece of ali mating onto the cam. Whilst it was a nice, smooth sliding fit, and whilst translation was acceptable, putting a bearing in to slide over the cam made a world of difference- it massivle reduced the counter torque being transmitted to the NavBot (due to much lower friction).
As mentioned, Im toying with ideas for doing it differently, as I dont really want to copy whats already been done- afterall, theres really no fun in that. Making a translating spinner the way Wyachi have is really nothing more than a money spending exercise, which is dull and boring and not what the sport should be about imho.#
All the best,
Eddy
No problem re. the link - at least I got to see what RR had covered recently. Looks like I missed some interesting articles (I read a few that come in to newsagents, but my fiancee would kill me if I subscribed.)
Guess youre right about the 30Hz problem - originally discussed wrt Stinger-esque robots, which only rotate at 2-3 times/sec; the problem is exacerbated for feathers, of course. Still, methinks there ought to be a way. Ill have to ponder on it.
If youre sticking with the cam and are using a navbot, I should clarify what I was saying about motor driving the cam and why I thought the navbot was weird:
Id suggest the cam - and anything else you want stationary - could be connected to the outer shell so that it can be rotated by a low power motor. That is, hold the shell still, spin the motor, and the cam twirls round. Then put a gyro on the same stationary bit as the cam, and make it feed the small motor. There might be some issues in coping with the amount of gain (although the relative voltage change is small), but the principle is there - spin the shell one way, spin the cam the other at the same speed. Receiver drives this small motor, obviously.
Nothing needs to be in contact with the ground, and since you can attach the spinning motor to the outer frame (although not mounting it centrally might be fiddly) you get more weight spinning. Strikes me as easier than a completely separate sub-robot, which is why I thought it was weird. Maybe theres a good reason for not doing it this way.
But maybe you didnt mean a navbot in this sense anyway. (Apparently this might be a good thread for not assuming we all know what we mean. :-)
--
Fluppet
Posting Permissions
Bookmarks