Gripper Arm
AIman
Posts: 531
Has anyone built a decent gripper arm?
I want a gripper with some strength with the ablity to move back and forth regularly. Servos don't really cut it for strength which means steppers or solenoids, probably solenoids.
Any comments or thoughts?
I want a gripper with some strength with the ablity to move back and forth regularly. Servos don't really cut it for strength which means steppers or solenoids, probably solenoids.
Any comments or thoughts?
Comments
Just one comment. You are speaking about a MAJOR outlay in time, money and electro-mechanical engineering. If you have all that at your disposal - "Press on McDuff"!
Regards,
Bruce Bates
Post Edited (Bruce Bates) : 12/12/2006 7:30:10 PM GMT
A relative works at a steel mill and a couple others are engeeniers, so getting any form of metal is bulk cost for me and the gears can be had through a few hobby stores in the area fairly cheaply.
My estimated cost for a 3 jointed Gripper running·4 solenoids, 1 dc motor, rubber for the grip and aluminum is about 100 bucks.
Unless you have some novel, new way of doing this, I suspect the encoders, to check actual arm/wrist/gripper position, may easily cost upwards of $100.00. Additionally, you will need a rather beefy power supply to operate (say) three solenoids, which is only two degrees of freedom (DOF). Depending on how you view it, a human arm has about 5 DOF. Two at the wrist, one at the elbow and two at the shoulder.
Personally, I'd love to see you overcome some of the costs that are usually involved. I once designed a small gripper with two DOF at the rotatable wrist and used two steppers and a gearing arrangement. It even had grip strength feedback to provide a degree of "feel".
Everything looked great on paper, until I got an estimate on the precision machining involved, which was over $500.00. That was the end of that project!
Regards,
Bruce Bates
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
<!--StartFragment -->
Not sure how many DOF's they had...but they weren't all that expensive!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·
Steve
"Inside each and every one of us is our one, true authentic swing. Something we was born with. Something that's ours and ours alone. Something that can't be learned... something that's got to be remembered."
Having owned a number of Radio Shack Armatrons, I can honestly say that's the neatest toy that Radio Shack ever brought to market. Now I'll tell you something that will absolutely blow your mind. The Armatron essentially had three joints plus a rotatable gripper. Each was independently movable. Yet, with all that flexibility, there was only ONE MOTOR inside! It was all done with a large cone of concentric gears, and what a MESS of them there were inside!
I made the mistake of taking my last Armatron apart when a couple of the the gears locked up, after I'd had it for 20 years or so. Biggest mistake I ever made, because I then had no way to re-time the gears to put it back together again <sigh>.
Regards,
Bruce Bates
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
<!--StartFragment -->
A BOE can run 4 servos and by using jumpers and resisitors can run some motors.
All you need to make a gripper - 3 jointed - is a servo for the 'shoulder' joint at the base, a servo for the second 'elbow' joint and one for the opening and closing of the gripper. One motor rotates the wrist of the gripper and one rotates the base of the shoulder.
If you want more power you can upgrade from servos to stepper motors and for even more power upgrade to solenoids.
I have the stuff at home to make a servo gripper or the motor run gripper without a problem, those parts are easy to come by and are scrape that was salvaged from previous products.
The goal isn't to just build a gripper, the goal is to build something decent.
Post Edited By Moderator (Aristides Alvarez (Parallax)) : 12/18/2006 4:52:52 PM GMT
I have been looking at the SMC products for a gripper on a wheeled bot.
http://www.smcetech.com/CC_host/pages/custom/templates/smc_v2/prodtree_product_2.cfm?cc_nvl=((searchPart,mhc2-16s),(CC,SMC,ACT,3007))
you would need this, a storage tank, air line, and a regulator. The air line would be 4mm OD line that could easily be hidden in the structure
Like Bruce, I applaud your efforts. However, also like Bruce, I think you're underestimating the task. You mention that it's easy to build a gripper out of servos. Well, of course it is. A servo has built in electronics and a position sensor so that all you need to do is provide a pulse of the right duration and it automatically seeks and holds that position. In addition, a servo comes in a nice rectangular package with 4 mounting holes, plus a splined shaft with matching servo arms and that already have holes in them. Furthermore, they have a built in gear train and bearings on the output shaft. You provide 5 or 6 volts DC and a pulse duration, and it does the rest.
You'll need to use a NEMA 23 style stepper motor at minimum. Anything smaller won't provide even as much power as a servo, so there's really no point in using anything smaller. Those little steppers in inkjet printers, etc? Trivial amount of torque compared to a servo. So, let's look at a stepper motor. It's... well, it's a motor. Nothing more. It has a round body with mounting holes. Okay, fairly on par with the servo. After that, it's all downhill from there. There are no built in electronics. Zero. It has 4 phases which need to be energized in the right sequence to get it to rotate. The voltage on those phases needs to be either a high(ish) voltage that is brought down with dropping resistors or PWM voltage to provide the correct current. Of course, with a 200 step motor, a full phase driver will move the motor 1.8 degrees on each step, or .9 degrees if you use a half phase driver. A micro-step driver can increase the resolution into thousands of steps per revolution, but with an increase in complexity.
In order to get the stepper to run at anything close to its maximum speed, you need to ramp the frequency of the phase energizing so the motor accelerates and decelerates without stalling. If it stalls, there is no way to know where it is because there is no position sensing device built into it. For that you'll need external devices ( such as encoders) along with corresponding electronics. There isn't even a way to set a known position on startup with a stepper. At the very least you'll need some sort of home sensor so your system knows where it is when it wakes up.
The shaft isn't splined, it's smooth which means the only thing that keeps things connected to it are set screws. This means anything attached to the shaft directly is made from metal. There are no control arms of any sort. In order to connect it to the load it is driving, you'll need to build all the connecting devices. Any stepper that has any strength is far heavier than a servo, so forget about connecting them directly to each other. The weight of strong steppers in measure in pounds, not ounces, so you;ll need metal frames to hold everything together. They are usually driven through some sort of gear reduction to increase the positioning resolution and torque. You'll also need a few amps of power minimum to drive each motor. If you're a machinist, then you're set. If not, then you're paying someone to do a lot of work for you.
You might be able to cobble something together using aluminum strips bought at a hardware store. The precision would be greatly compromised, as would the strength. You say you want something "decent", so I'm assuming you want something than can at least lift a few pounds worth of weight. All of this will require a lot of power, a lot of machining, a lot of time, and most likely a lot of money. When you switch from RC style servos to steppers, you throw away a ton of engineering and electronics that are built into each on of those little packages, and have to re-create all of it yourself.
You mention that getting metal isn't expensive. You're right. It's doing something to it that is expensive. You also mentioned building a gripper arm using solenoids. Solenoids are two position devices. They are either on or off, open or closed. For the actual gripping, this could be fine, but for positioning on the arm and joints it's a non-starter unless you want an arm that has only two positions. Because of this, you'll need steppers for the joints.
Of course, your definition of "decent" and mine may be different. To see my definition, go to my website and look at some of the stuff I have built for my Halloween display each year:
www.SoCalHalloween.com
I'm not saying it can't be done, nor that you shouldn't try. I am saying that you are underestimating the task at hand, and that you are going to discover none of it is as easy as you think.
Thanks, PeterM
To build an average gripper - if you have the parts - only takes a few hours. To build a gripper from scratch can take weeks simply because of finding parts and then making them fit. If you have some scrap sheet metal and start making very short small grippers you will learn a lot more then if you are trying to do something like a small crane.
For me to build the gripper I have in mind, aside from the programing, would take a weekend. Now to be fair, I have a gift with building things so for most it would take a week maybe a week and a half.
The key thing is to think out the picture in your head before you start, then sketch it out on paper or in Paintbrush. BE SURE to go a single item at a time and then add a single item to that. In Paintbrush that means draw out your base, then for the next drawing use the base and add a bearing. Save drawing two under a different name and then add the first screw to hold the bearing in place. Save drawing three and then add to that the second screw used and so on. By doing things part by part you can see an image on the computer and if anything is missing when you are building you can look at your plans and know right away. You don't need to mess with techy stuff, just get down where the screws go, wheres the wires go, where the gears go - stuff like that. Works wonders.
The base for my gripper is on a flat bearing with a gear mounted on it (looks like a tea plate from the top down). The shoulder is mounted on top of this gear. The up and down motion for the shoulder is run by a solenoid with 20 lbs of holding strength. The stepper attached to the gear allows for rotation.
The elbow is a similar structure with the shoulder ending as a flat gear to which the elbow is attached. The elbow bends like the shoulder via solenoids of 20 lb strength.
The elbow ends at the wrist were things start to get a little sticky. The wrist is attached exactly like the elbow and sholder (a gear and solenoid) BUT it also holds a stepper and solenoid for the gripper.
The arm reach is about 6 feet in length but holds up to a MAX of 5 pounds. Aluminum tubing is being used for strength on the arm - three to a section which allows for better strenght - and the motors are rather torqy motors I picked up at a surplus store.
In short, the steppers are similar to the Parallax unipolar stepper motor (Stock#: 27964) except that they are geared to put out a lot more power. The solenoids I found are from Midwest Motion and a different set from Jameco. They range in power from 25 lbs holding force to 10lbs.
Pnumatics and Hydraluics were explored but ruled out because of the need to haul compressors or pumps. Solenoids and steppers were selected because of the commonality of the parts. Servos of strength around 100 lbs were considered but ruled out because of being cost prohibative, they are very nice if you can afford them...
The biggest problem right now is getting enough pins out of a Stamp. Going off the Parallax drawing for a reference people can see - http://www.parallax.com/dl/docs/prod/motors/Stepper_Motor_27964.pdf - there are 4 wires going from the motor to the stamp with each wire requiring a pin.
Since I want to use 4 motors, 5 solenoids (1 for each joint and 2 for the gripper) and a few sensors I am quickly at 30 ish pins (16 for the motors, 10 for the solenoids and 6 for the 3 sensors).
Obviously the 2P40 is quickly becoming a serious consideration.
Post Edited (AIman) : 12/18/2006 5:26:47 PM GMT
What ever happen to this project? is it done?
I am still working in my's
·http://forums.parallax.com/showthread.php?p=641010
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Fernando Gomez
revinc.us
gomez-rivera.com
Never compare yourself with anyone else, there will always be someone bigger·or·smaller·than you.
Post Edited (willy1067) : 4/19/2007 7:10:23 AM GMT