PDA

View Full Version : My First Machine For Robot Combat



Vsions
09-07-2008, 08:08 PM
Hi every one, I am new to robotics and Im very excited to get involved in this hobby which I have always wanted to do. I started out with one version and then redisigned it once I purchased some 5990's. Those servos are great and incredibly powerful. I decided to create my own grippers using lynxmotion parts because I did not like anything currently being offered. They will also be powered by a 5990. I wanted a very aggresive look and I wanted the oversized claws for melee combat. I mounted a temporary 6mm airsoft cannon but will be replacing it with a good one for actual combat. This is not for mech wars by the way. I plan on doing my own smaller competition with him with very different rules and scoring system. I would like to get involved in mechwars with my next machine which is already being designed with that competition in mind. Hope you guys like it. Im on pause right now till funding is approved by the local government body (my wife) for the final phase of this machine. But all I need right now are the components to build the second arm and a sabertooth motor controller. So without further delay let me introduce you all to my first machine, I think I am going to call it Sampson 209:

This is my first version
http://i331.photobucket.com/albums/l456/Vsions/DSC00375.jpg



Now this is the redesigned version...
http://i331.photobucket.com/albums/l456/Vsions/DSC00400.jpg

http://i331.photobucket.com/albums/l456/Vsions/DSC00404.jpg

http://i331.photobucket.com/albums/l456/Vsions/DSC00403.jpg

Hope you guys like it..

Dennis

asanist_13
09-08-2008, 12:34 AM
Whoa!!!
This could be the first grappler bot in TRC!!!!!!
Nice work!!!

DresnerRobotics
09-08-2008, 10:00 AM
I dig it.

My only suggestion is fortifying where your torso tubes connects to that hex block on the rotating base. A single L bracket is the weak point for the entire chassis.

Why not simply rotate the 4 sided block you have at the bottom of the two torso tubes 90 degrees so that you have an SES mount facing the side of the hex block and mount it directly to it? Would be much more stable removing that L bracket.

Vsions
09-08-2008, 10:44 AM
I dig it.

My only suggestion is fortifying where your torso tubes connects to that hex block on the rotating base. A single L bracket is the weak point for the entire chassis.

Why not simply rotate the 4 sided block you have at the bottom of the two torso tubes 90 degrees so that you have an SES mount facing the side of the hex block and mount it directly to it? Would be much more stable removing that L bracket.



Yeah that was a concern of mine as well, tried your ideea initially, but I cant connet them together since I cant get a screw driver in there to connect them from the inside, any ideas?

DresnerRobotics
09-08-2008, 11:28 AM
A thin and long enough screwdriver to enter in the SES mount holes from the opposite side of the hex block?

Or just drop screws in from the square block side and use your fingers to hold nuts in place inside the hex block.

Smaller screwdrivers are a must when working with SES.

sam
09-08-2008, 06:48 PM
Very nice robot.

I have to say that I agree with Tyb, looks like a weak point. But other than that itès a wonderfull robot.

I love the hand!

Vsions
09-11-2008, 09:53 AM
I had a question for anyone that may know. I was interested in using a linear actuator for opening and closing the claw. I know firgelli has some but they are out of stock for the ones with the rc controller. I then thought about setting up a 5990 which I have figured out how to make it work, but I was wondering if I could use a push pull solenoid. I dont know exactly what solenoids are, I know some are valves. I am not looking for any precise movement, I just want to be able to open and close the claw with as much speed and powers as possible and keep it light. Ideally I would love for this claw to at least be able to crush an empty soda can. Any ideas from the big dogs in here?

Adrenalynn
09-11-2008, 10:49 AM
Electric Solonoids (electromagnetic solonoids) can be very very fast. But at the end of the day, they're electromagnets - TANSTAAFL (There Aint No Such Things As A Free Lunch) - and you just can't put the power into them to make them THAT strong. Are there any hobby-grade ones out there that will crush a can? Unknown. Maybe at dead-center of the drive shaft, but if you have no mechanical advantage and are putting your force a few inches out away from dead-center, I'd guess "no". Absolutely not if you want to crush an undeformed can end-to-end. Without disrupting the cylinder first, that can take a couple thousand pounds. ;) [Favorite "science of structural engineering" demonstration...]

All that said, my answer might be woof instead of woof, so you be the judge. :)

Vsions
09-11-2008, 11:00 AM
Electric Solonoids (electromagnetic solonoids) can be very very fast. But at the end of the day, they're electromagnets - TANSTAAFL (There Aint No Such Things As A Free Lunch) - and you just can't put the power into them to make them THAT strong. Are there any hobby-grade ones out there that will crush a can? Unknown. Maybe at dead-center of the drive shaft, but if you have no mechanical advantage and are putting your force a few inches out away from dead-center, I'd guess "no". Absolutely not if you want to crush an undeformed can end-to-end. Without disrupting the cylinder first, that can take a couple thousand pounds. ;) [Favorite "science of structural engineering" demonstration...]

All that said, my answer might be woof instead of woof, so you be the judge. :)


No actually, I should have been more clear, I would like it to be able to crush a can through the cylinder not end to end.

Adrenalynn
09-11-2008, 11:17 AM
I figured that was the case, I was more looking to be personally precise. :)

A quick search shows some that will do 10lbs @ 12vDC for about 1". A soda can is about 2.5", so 1" from each side should reasonably crush it.

Adrenalynn
09-11-2008, 11:25 AM
11.2 pounds of force applied through 3/4 inch of surface area will do this:

655

pretty quickly. I used 3/4" square stock on both sides of the can, can spanning it, and then set two batteries simultaneously on the square stock. Bonus points for decreasing the surface area even more...

Another thing I didn't consider in my test is that you could theoretically have 10lbs pushing from BOTH directions - 20lbs total force for short periods. As long as the surface area is kept small and you get as close to the center of the can as possible, you could be able to at least approach your design goal. I haven't looked at 24-36v linear solenoid push-pull actuators yet, only 12vDC...

Vsions
09-11-2008, 11:26 AM
is this still controllable with the sc-32 or atom pro, Im sorry if these are simple questions but while I feel relatively confident with my mechanical abilities, I am totally unknowledgeable about the electronics and programing and am still learning the basics about what exactly these components can do

Adrenalynn
09-11-2008, 11:33 AM
You would need a relay circuit that can be driven (switched) by the SSC-32 or Atom. They're high-current devices like a large motor.

In fact, looking at a few spec sheets, you're going to need about 15A each to generate 11lbs.

Vsions
09-11-2008, 11:40 AM
You would need a relay circuit that can be driven (switched) by the SSC-32 or Atom. They're high-current devices like a large motor.

In fact, looking at a few spec sheets, you're going to need about 15A each to generate 11lbs.

Thanks for your advice I really appreciate it alot, Do you think this is feasible within this platform size and weight constraints? I also want to definitely keep this thing totally wireless would this still be possible with this setup?

Adrenalynn
09-11-2008, 11:52 AM
Well, with the caveat that you'll need lots of battery @ 12v or more. Also, the solenoids that generate that kind of power are really big and REALLY heavy, so I doubt your little arm would even dream of lifting them.

Wireless and portable is no problem, but I suspect given the design constraints of arm lift capacity, you'd probably be better off with a geared servo or geared motor and keeping your contact patch as small as possible. Given infinite time and infinite gear reduction, a servo-driven pincher could cut a car in half. :)

Vsions
09-11-2008, 12:09 PM
wow, never thought about that but you are right. Im gonna think about sticking with servos. I just wanted to educate myself a little about solenoids and I appreciate the lesson. Actually I was really impressed with the strength of the 5990's right out of the box, those things caught my finger during testing and it was not pleasant.

darkback2
09-11-2008, 12:22 PM
I know this is probably overkill...but have you considered some sort of pneumatic or hydraulic ram? probably bring up a wieght issue though...It would be cool to use a CO2 cartridge and valve set up to have the claws slam shut...

Adrenalynn
09-11-2008, 12:26 PM
It's a good idea - the weight can be moved off to the base primarily. Time to look at Sam's air muscles? :) He's getting like 10x the force needed here.

Vsions
09-11-2008, 12:36 PM
interesting idea, I originaly kinda overlooked it because of weight concerns, but a co2 cartridge is doable. The only thing is I still would want it to be able to hold something not just a quick grab and release

Matt
09-11-2008, 01:30 PM
The VEX stuff should do the trick.
http://www.vexrobotics.com/vex-robotics-pneumatic-parts.shtml

FryGuy
09-11-2008, 01:48 PM
Some people use a threaded bolt with a motor on a nut to do beetleweight clamper robots. That said, I really hope this thing doesn't get used for any sort of battlebot-type combat, as it'd likely get destroyed, and it looks too nice for that.

Vsions
09-11-2008, 01:55 PM
Some people use a threaded bolt with a motor on a nut to do beetleweight clamper robots. That said, I really hope this thing doesn't get used for any sort of battlebot-type combat, as it'd likely get destroyed, and it looks too nice for that.


Thanks, dont worry I am definitely not doing a battle bots thing at all. I actually dont really like that stuff because in my eyes they are really just advanced rc cars. I am going to do something more like mech wars, but in a different way. I definitely dont want to destroy any robot.

Vsions
09-11-2008, 01:57 PM
The VEX stuff should do the trick.
http://www.vexrobotics.com/vex-robotics-pneumatic-parts.shtml


Wow you got me totally open on that, I am going to see if I can implement that system, nice find..:happy:

Adrenalynn
09-11-2008, 01:57 PM
If you're going to do a screw, you want to do a more advanced screw like a tech screw, acme screw, or ball screw. Bonus points for multi-start.

The lifter on our SciOly 'bot last year would lift 30lbs for 8" in 2.75 seconds, powered by a little 9.6v Banebots motor.

I was going to say "I know all about screws" - but I just couldn't think of how to write that in a way that wouldn't make me blush, so I decided not to.

[edit] DOH!

Vsions
09-11-2008, 02:04 PM
If you're going to do a screw, you want to do a more advanced screw like a tech screw, acme screw, or ball screw. Bonus points for multi-start.

The lifter on our SciOly 'bot last year would lift 30lbs for 8" in 2.75 seconds, powered by a little 9.6v Banebots motor.

I was going to say "I know all about screws" - but I just couldn't think of how to write that in a way that wouldn't make me blush, so I decided not to.

[edit] DOH!


Haha nice.... could you show me an example of how that works, I dont think I have ever seen that stuff. does it move a cylinder in or out by turning the screw over a threaded cylinder or something?

Thnks

Adrenalynn
09-11-2008, 03:05 PM
It's a threaded nut on a screw - very low TPI (Turns/Inch). In our case, I use a chain drive (#25 roller chain) with #25 chain cogs - all except the screw purchased here. If you end-up needing a source on the screw, I can really recommend the folks over at precision machining - I can get you their contact info.

I don't know if I have any pics on this machine...

Ok - here was some early shots:

sthmck
09-11-2008, 04:17 PM
If you go with an air powered pincher you should look into air muscles. You could use vex air system that matt linked minus the cylinders. That will take a lot of weight of the arm (air muscles weigh next to nothing), and still be very powerful. The only thing you may have to do is use two muscles. You could however get around this by using just one muscle and having a spring return it to its original position. Talk to Sam he should be able to help you out with getting a few air muscles.

sam
09-11-2008, 07:25 PM
Well, I just finished making an air muscle 13 cm long from end to end (where you connect it to mechanical parts) when streched. I don't know if I could go lower. It weight only 40 grams about. I could put a picture, it's actualy pretty funny to looks at.

FryGuy
09-12-2008, 01:26 AM
I found a link to the clamper robot I was thinking about, unfortunately there are no videos. It claimed to have 300 lbs of force on the tip:

http://www.fingertechrobotics.com/team/robots/hellrazer.htm

I can't really tell what it used, but I think it was a banebot motor (the 256:1?) with some sort of threaded screw. A big problem was getting it mounted such that it wouldn't destroy the frame when it clamped down.

I don't think pneumatic will work that well, given the amount of psi that can safely be put in a small container (100-130 psi)

Adrenalynn
09-12-2008, 01:32 AM
I could have easily gotten mine to 300lbs just by changing the cogs.

Paintball and BB-gun cylinders are a LOT more than 130PSI in the tank - are you thinking of the regulator's output? a 16g canister stores like 850PSI?

metaform3d
09-12-2008, 01:45 AM
What I found, employing a scratch-built screw for linear actuator, was that it was really easy for the thing to tear itself apart. Make sure you have limit switches that are wired in at a low enough level. Either that or design it cleverly so that if it runs unchecked that the screw unthreads itself and doesn't cause any damage.

Adrenalynn
09-12-2008, 02:27 AM
In my design above, it would pop the top or bottom off the frame at about 45lbs.

FryGuy
09-13-2008, 01:58 AM
I could have easily gotten mine to 300lbs just by changing the cogs.

Paintball and BB-gun cylinders are a LOT more than 130PSI in the tank - are you thinking of the regulator's output? a 16g canister stores like 850PSI?

I was talking about compressed air, not co2. Mainly, in reference to the vex system (which has stated it's 100psi limit on the container). I've seen the combat robot guys take them to higher psi, and the effects of rupture at the higher pressures as well.

Sorry if I'm biased towards the "battlebot" side of things, as that's what I'm most familiar with. I know that's not what's being discussed, but I thought I'd throw my two cents in.

Adrenalynn
09-13-2008, 02:49 AM
That's cool - I hadn't looked at the Vex system. I suspect it's more the regulator than the tank, honestly.

Air bottles can certainly be had up in the 3000+ PSI range.

Vsions
09-14-2008, 03:28 PM
So I have been trying to figure out what Im going to do to to close the claws, and I am thinking about the current suggestions pros and cons:

Servos-
Pretty simple to set up, decent strength/speed. Not as fast as the air muscle or as strong and it adds a good amount of weight to the arm.

Air muscle- Fast, Strong and significantly lightens the arm. Unfortunately this might be more difficult to set up and would require me to set up an air supply system on top of the air supply system being used for the airsoft cannons. I would also have a limmited amoung of times I could close the claw due to the limited amount of air available.

Screw System- Very Very strong and capable of really holding the target in place until its released. This system might also be difficult to set up and will likely be slower than the other two. would also probably be the heaviest.

Right now Im leaning towards the airmuscles, but I looked at my design and would have to reconfigure the claw slightly to fit everything properly. Sam as soon as I have somethiing configured properly fo the air muscles I will be posting some pics to see what you think about outfitting it with the air muscles. That is what I am in the process of doing. I also should be aquiring the funds soon to finish this model and begin the programming aspect and do a demo for you guys.

Later
D

sthmck
09-14-2008, 04:42 PM
Are you going to be using co2 for you air soft gun? I am asking because you may be able to combine your two air systems and eliminate some of the complexity. I am not sure how well air muscles work using co2, but I imagine it wouldn't be to much different. Other wise you could use compressed air. I have seen some pretty small air tanks used for paintball that should give you enough air for both systems.

Vsions
09-14-2008, 06:08 PM
Are you going to be using co2 for you air soft gun? I am asking because you may be able to combine your two air systems and eliminate some of the complexity. I am not sure how well air muscles work using co2, but I imagine it wouldn't be to much different. Other wise you could use compressed air. I have seen some pretty small air tanks used for paintball that should give you enough air for both systems.


Yes I will be using a CO2 system, but I wasnt sure if the c02 system would cause the muscles to freeze and become brittle.

sthmck
09-14-2008, 06:33 PM
you would probably just have to incorporate an expansion chamber. Its worth trying out though. Are you going to be using those really small CO2 cartridges?

Vsions
09-14-2008, 08:53 PM
you would probably just have to incorporate an expansion chamber. Its worth trying out though. Are you going to be using those really small CO2 cartridges?


I am hoping to be able to use the cartridges that are used on crossman bbguns.

sthmck
09-14-2008, 09:15 PM
ok I know what you are talking about. If I get some time I will work on setting up a CO2 powered air muscle system. I am curious to see how many actuations you could get from one of those cartridges.

Adrenalynn
09-14-2008, 10:45 PM
It'd have to be a very small muscle for even one expansion. Those little cartridges don't have much volume...

sam
09-15-2008, 03:12 PM
I just want to try out how many cycles one can get with 100 PSI with 9.15 cubic inches. (or 150 ml) (the vex thing)


The quatity of air :

150ml X 100 PSI /50 PSI = 300 ml with 50 PSI pressure (this is not totaly accurate, but will do here).

Estimate of the muscle : 9 cm long - 4 for the metal pieces = 5 cm.

Volume :

5X .95 (radius) * .95 * PI = 14.17 ml * 2 (one for each gripper) = 28.32 ml.

300/28.32 = 10.6 uses with each hand before it's empty. :o

Bigger tank, higher pressure? :tongue:

gdubb2
09-15-2008, 03:30 PM
Don't forget to consider the rate of actuation into things. As you draw off the Co2 which is stored as liquid, it will cool the tank it's in. Which lowers the pressure available. My combat bot had a Co2 powered flipping arm. I have actually operated it fast enough to turn the 20 oz. Co2 tank into a block of dry ice with virtually 0 pressure. When I should have got 37 shots per tank, I got maybe 10 or so.

Co2 is very predictable, but does have some shortcomings.

Good luck
Gary

sam
09-15-2008, 04:31 PM
Oh yeah. I didn't really work with C02 with the air muscles, but I know it freezes paintball guns. :genmad:

Vsions
09-17-2008, 09:23 AM
I just want to try out how many cycles one can get with 100 PSI with 9.15 cubic inches. (or 150 ml) (the vex thing)


The quatity of air :

150ml X 100 PSI /50 PSI = 300 ml with 50 PSI pressure (this is not totaly accurate, but will do here).

Estimate of the muscle : 9 cm long - 4 for the metal pieces = 5 cm.

Volume :

5X .95 (radius) * .95 * PI = 14.17 ml * 2 (one for each gripper) = 28.32 ml.

300/28.32 = 10.6 uses with each hand before it's empty. :o

Bigger tank, higher pressure? :tongue:


Thanks for those results. I looked on the vex site and they were not offering any bigger tanks. Not to mention I cant have huge tanks mounted on this thing. I really was leaning towards the airmuscles idea but 10 shots for each hand is to low. Is that all off one tank? would adding two tanks give me double the use?

Adrenalynn Ive also looked into using some sort of screw, but have not found anything small enough, I have at most 4 inches to play with, have you seen anything that fits that scale?

Adrenalynn
09-17-2008, 12:23 PM
eMail or call Wanda at PrecisionTechMachining. [email protected] or (321) 751-4212 She's awesomely helpful. Feel free to mention that I talked to her about robotics applications when I bought my lead-screws.

I ended-up cutting mine down (blasphemy because they're so beautifully machined) a few inches because I needed a different design than I originally ordered for.

They're crazy inexpensive compared to everyone else. A 6" carbon steel 1/2" diameter 10TPI screw was like $15, beautifully hand-turned with super good precision. My 12" was around $20. When I cut it apart with the dremel (hard steel, do it outside and away from flammables - wear proper gear!), I felt like I was violating a religious idol.

I'm not associated with these guys in any way other than as a really satisfied customer. I was trying to do some insanely complex theoretical calculations (duh) and asked for very technical information about the screw prior to ordering. She got me some basic info, and then took it upon herself to scramble around interrogating busy machinists to glean the remainder of what I was looking for, and had all the information I needed back to me the next day. Esoteric stuff like the angle of the threads, the depth of the cut, the coefficient of friction on the nut, ... All for a measly $35 eBay order.

I wouldn't buy a lead screw from anyone else, even though they don't make some of the more technical designs.

metaform3d
09-17-2008, 02:57 PM
I just used a threaded rod I got at the hardware store, and a little block of aluminum that I tapped. Connect it directly to the motor shaft and you're good to go.

sam
09-17-2008, 03:29 PM
Thanks for those results. I looked on the vex site and they were not offering any bigger tanks. Not to mention I cant have huge tanks mounted on this thing. I really was leaning towards the airmuscles idea but 10 shots for each hand is to low. Is that all off one tank? would adding two tanks give me double the use?

Adrenalynn Ive also looked into using some sort of screw, but have not found anything small enough, I have at most 4 inches to play with, have you seen anything that fits that scale?

No problem. Yes, adding another tank would give you twice the shots. The low pressure is really a problem. you could have a bit bigger tank and have 3000 PSI and have hundreds of shots. But then again you would need a special pump to get to 3000 PSI. But you don't have enough place. Are you thinking of Co2 thing for your gun PLUs compressed air?

Adrenalynn
09-17-2008, 04:11 PM
That's one way to do it, Meta. But a multistart hitech screw can make 3 or 4"/turn and generate hundreds if not thousands of times the force your screw will. Even a simple acme cut leadscrew will generate many times the force and won't bind the way a threadrod will.

Vsions
09-17-2008, 04:58 PM
That's one way to do it, Meta. But a multistart hitech screw can make 3 or 4"/turn and generate hundreds if not thousands of times the force your screw will. Even a simple acme cut leadscrew will generate many times the force and won't bind the way a threadrod will.


I agree, first thing I thought in regards to using a regular screw was that it would likely bind alot. Im under a huge dealine at my studio but soon as I have some time I will call that company you mentioned.

Hey Sam this is what I would like to use for the cannons, they are the co2 canisters traditionally used in hand bb guns.

http://home.hiwaay.net/~ispellan/Co2.html

metaform3d
09-18-2008, 01:11 AM
Sure, a custom screw is best but an off-the-shelf rod can do the job. Just trying to provide alternatives.

I've never had a problem with binding. Perhaps because my rider is aluminum it wobbles a little but is resilient to shear loads.

Vsions
09-18-2008, 10:49 AM
Sure, a custom screw is best but an off-the-shelf rod can do the job. Just trying to provide alternatives.

I've never had a problem with binding. Perhaps because my rider is aluminum it wobbles a little but is resilient to shear loads.


Thank you I definitely do appreciate your input. I can honestly say that I assumed it would bind without even trying it. I will definitely explore that option as well. Unfortunately I am currently under a tight deadline but next week end I should be able to get on it. Thanks Meta..