PDA

View Full Version : air muscles



sthmck
11-13-2007, 09:45 PM
Just wanted to start a thread on air muscles. I was wondering if anyone here has any experience with them. Right now I am messing around, just to get the feel for how they work. I am also designing some basic systems that use air muscles for actuation. I am just curious to see what everyone thinks about this type of actuator.
thanks

Alex
11-13-2007, 10:05 PM
I've always been intrigued by these. Do you have or could you record some video of these in action?

sthmck
11-14-2007, 09:15 AM
Hey I dont have any videos right now, but I will try to post some asap. I am getting ready for finals so I will be busy for the next few weeks. As soon as I get some free time I will put something up. Most of my ideas are still just concepts. I am trying to figure out a simple solution to control air muscles. Something that can be easily implemented, so that air muscle can actually become a useful option for hobbyist.

Alex
11-14-2007, 10:36 AM
I'm a huge fan of these air muscles. They're really cool and have a ton of potential, but I just haven't seen an elegant solution to control them. Shadow Robot Company came up with a pretty cool controller for their air muscles, but it looks like you're only controlling eight of them for $1,826:

http://www.shadowrobot.com/SPCU/

It does have some pretty sweet features though:



8 high-quality Isonic valves, giving high flow rates necessary for big movements.
8 sensor inputs, allowing you to control the muscles based on pressure, position, or other data, as well as providing information to the host computer.
Integrated drive electronics - no additional hardware is required to drive the valves.
Direct control of valves, allowing soft movements to be sequenced from your software directly.
Integrated control loops, so once you've configured your hardware you only need to tell it where to go to!
Non-volatile setting storage, so configuration is a one-time deal - just power it up and run!
Serial port connection to the PC - just hook it up and get going - no need to mess with add-in cards.Whenever you can get up some video, we'd love to see them in action:D Also, feel free to post any of your pictures in the new TR Community Gallery we just installed. Just click on UserCP at the top and select "User Control Panel", then click on "Edit Gallery" and upload your pics!

Good luck on your finals!

MYKL
02-11-2008, 05:15 PM
I too have experimented with this type of actuator for some time now. Have you thought about using a non-compresable fluid in your design? I've used water and the 'muscles' strength is increased exponentially. I'm working on a servo driven version of it at the moment. I'll post pictures as soon as I button up something to the point that you can recognise the componants.

Try it with syringes first to test it. You will be pleasently suprised by the response of the 'muscles' to the non-compresable fluid.

sthmck
05-12-2008, 11:17 PM
I too have experimented with this type of actuator for some time now. Have you thought about using a non-compresable fluid in your design? I've used water and the 'muscles' strength is increased exponentially. I'm working on a servo driven version of it at the moment. I'll post pictures as soon as I button up something to the point that you can recognise the componants.

Try it with syringes first to test it. You will be pleasently suprised by the response of the 'muscles' to the non-compresable fluid.

I was actually thinking about this approach some. I think the reason that these actuators tend to use air instead of a non-compressible fluid has to do with air being cleaner. I have never seen anyone else use a fluid before. Have you made anymore progress with this approach?

While most developers tend to focus on valves as the most complex part of the system, I feels as though the pressure supply is the most important part in developing a system for mobile robots. What I mean by this is that to provide a constant supply of air for a robot to run at full capacity for any reasonable amount of time would require a strong compressor. This would mean more weight. And then you get into the whole more power more weight more power more weight thing again. The way I see it right now an air powered robot requires an air supply, valves, actuators, and some type of feedback. The air supply could be a compressors, therefore a battery would be need to power it (making the system much heavier), or a HPA system like the ones used in paintball (you can get light carbon fiber tanks, but the air supply would need to be refilled). Valves really aren't as big of a deal as people make them seem. Most electronic paintball guns have a small solenoid that controls an air valve. (I have a paintball gun sitting on my work bench right now with one in it) Six of these valves could control a three dof leg in a hexapod, and they wouldn't take up to much room. As far as feed back goes I would guess a joint position sensor in the form of a pot or an encoder or something would be the only thing needed. I guess you could throw in a cheap digital air pressure gauge. You could get one of those from a tire pressure gauge from autozone or some place like that.

I am interested in how much more power a air muscle creates when using water to actuate it. If it is a reasonable higher amount then it would be much more advantageous to use a non-compressible fluid for actuation.

MYKL I am very interested in your progress with using water to actuate the muscle. Did you make your own muscle or did you buy one?

metaform3d
05-13-2008, 02:13 AM
Seems to me (without having any practical experience, mind you) that pneumatics and hydraulics would have some very different design constraints. You couldn't just pump liquid into a pneumatic system and expect it to work. At the very least a pneumatic system could be designed to leak, since some escape of pressure at the end effector would not badly hurt the operation of the system, and in some cases could help it by preventing local overpressure. A hydraulic system on the other hand has to be sealed, and all the liquid that's pumped into it has to return. In fact, when I think about it, that's the major difference. In a pneumatic system the pressure difference that generates work is the difference between the pressurized vessel and the outside atmosphere. In a hydraulic system it's the difference between the output side of the pump and the input side of the same pump.

I was at the Computer History Museum in Mtn View last weekend, and they had a display of some early robots. One was a set of disks set along a kind of spine with three air bladders at each juncture. You could easily imagine adjusting the pressure in each bladder and making the whole snake-like arm bend. The little museum card said it was abandoned because the motions were not repeatable.

sthmck
05-16-2008, 05:25 PM
Seems to me (without having any practical experience, mind you) that pneumatics and hydraulics would have some very different design constraints. You couldn't just pump liquid into a pneumatic system and expect it to work. At the very least a pneumatic system could be designed to leak, since some escape of pressure at the end effector would not badly hurt the operation of the system, and in some cases could help it by preventing local overpressure. A hydraulic system on the other hand has to be sealed, and all the liquid that's pumped into it has to return. In fact, when I think about it, that's the major difference. In a pneumatic system the pressure difference that generates work is the difference between the pressurized vessel and the outside atmosphere. In a hydraulic system it's the difference between the output side of the pump and the input side of the same pump.


You are right, there are differences in the types of systems. With airmuscles though the differences are not the actuatot but the supporting system. If an air muscle is made properly both a liquid and a gass could be used. Where they differ the most is like you said, the return lines and the pressure supply. while standard actuators for both gas and liquid can be single or double acting an air muscle is only single acting. meaning it can only contract. To be able to move a joint in to directions you need opposing actuators. To take care of the return line in a liquid system all you would need to do is make a valve that would switch between flow in and flow out. Also I think that you could have a much smaller pump for a hydraulic system then for a pneumatic one.

MYKL
06-12-2008, 02:09 PM
I make my own Hydraulic muscles. The beauty of this simple design is that they are completely sealed. I plan on incorporating them in antagonistic pairs just like the real deal is designed.

I've found some wonderful miniature sized PD pumps that would move the fluid from one muscle to another but these mechanical marvels are a bit too steep for my budget at the moment... The first Hydraulic muscle I made was able to lift a five gallon air tank. Easilly. It was full of air sure but it probably weighs 15-20 pounds. My wife can't keep one of my actuators from contracting holding one end in each hand. The muscles she is using to try to spread her arms apart aren't her strongest but the actuator has a diameter of half an inch. I have some sleeving and tubing that has a diameter of 1". Hmm...

My favorite though are the little muscle that have a filled diameter of about three eighths of an inch and a stretched diameter of less than an eighth. Tiny little muscles for tiny little creatures.

I wish I had more time...

Sorry I've been:

1. Unemployed. (No internet access)
this post is evidence that I've remedied this problem.
2. Taking Mandarin Chinese at night school.
Took finals last monday. Now I'll have my evening a bit more free.

I have resurrected my Supernova project and have developed HiRo (hip rotation and tilt along with waist tilt and turn). Once the SN is complete cybernetically speaking I'll design its hands to scale and post some shots of the 3D model. The little Hydraulic muscles would rock in this application. I have a plan for servo-driven metering of the fluid. We'll see.

^_^

Matt
06-12-2008, 03:00 PM
I make my own Hydraulic muscles. The beauty of this simple design is that they are completely sealed. I plan on incorporating them in antagonistic pairs just like the real deal is designed.

I've found some wonderful miniature sized PD pumps that would move the fluid from one muscle to another but these mechanical marvels are a bit too steep for my budget at the moment...

Can you post some pics of this stuff? Myself and others would be interested in seeing your designs and some of the hardware you are looking at.

metaform3d
06-12-2008, 03:33 PM
A tutorial would be ever better than pictures...

MYKL
06-12-2008, 04:12 PM
Can you post some pics of this stuff? Myself and others would be interested in seeing your designs and some of the hardware you are looking at.

Yeah, I'm suprised you guys still even talk to me at all. I've been all talk and little show. I'll see if I can't scrape some evidence of my dementia together so that you can see that I do truely belong here.


A tutorial would be ever better than pictures...

A pictorial tutorial will be my memorial...

^_^

Sienna
06-12-2008, 07:19 PM
One other resource for miniture hydralics: Find forums for 1/16th scale RC truckers. A lot of them are getting into 1/16th scale construction equipment, and almost all of it is hydralic driven. There are a few companies selling the equipment itself, but it seems a lot of people make their own, and I have seen several plans / sketches for things in the past.

Alex
06-12-2008, 11:32 PM
Yeah, I'm suprised you guys still even talk to me at all. I've been all talk and little show. I'll see if I can't scrape some evidence of my dementia together so that you can see that I do truely belong here.

meh, we're all mostly professionals here in some area or another. I can't speak for everyone of course, but I'm sure most of us can totally relate to life getting in the way of robotics (wait... did I just say that? Sorry Matt, please don't fire me:D)

LinuxGuy
06-12-2008, 11:37 PM
meh, we're all mostly professionals here in some area or another. I can't speak for everyone of course, but I'm sure most of us can totally relate to life getting in the way of robotics (wait... did I just say that? Sorry Matt, please don't fire me:D)
I must be odd person out then, because I'm not a professional at anything. I am just learning as I go and need to learn new stuff.

8-Dale

Alex
06-12-2008, 11:57 PM
ha! You're actually one of the ones I had in mind when I said that RobotGuy. You have a TON of experience in robotics (I'm a bit jealous to tell you the truth:D) and really helpful to everyone here!

MYKL
06-13-2008, 08:54 AM
One other resource for miniture hydralics: Find forums for 1/16th scale RC truckers. A lot of them are getting into 1/16th scale construction equipment, and almost all of it is hydralic driven. There are a few companies selling the equipment itself, but it seems a lot of people make their own, and I have seen several plans / sketches for things in the past.

That kind of modeling is awesome and the painstaking attention to detail is inspiring. I'd like to see robotics spruced up a bit by artests like that. The problem with realizing tiny hydraulic cylendars is the surface area of the piston. The the pistons diameter regulates the force said cylendar is able to produce. Where with the hydraulic muscles the force is distributed over the entire inner surface of the inner seal. (thats what I've led myself to believe anyhow... ^_^


meh, we're all mostly professionals here in some area or another. I can't speak for everyone of course, but I'm sure most of us can totally relate to life getting in the way of robotics (wait... did I just say that? Sorry Matt, please don't fire me:D)

I'm hoping that someday one of my 'messy hobbies' will make my life easier... I'd like to be an independantly wealthy professional.


I must be odd person out then, because I'm not a professional at anything. I am just learning as I go and need to learn new stuff.

8-Dale


ha! You're actually one of the ones I had in mind when I said that RobotGuy. You have a TON of experience in robotics (I'm a bit jealous to tell you the truth:D) and really helpful to everyone here!

Anyone that wishes to learn will humble himself to the point that he/she will be able to draw the vital energy and knowledge required from any acquaintance that seems worthy of being assimilated.

Here is a video of one of my actuators cut and paste it into an address bar:

s225.photobucket.com/albums/dd108/DrageSjel/?action=view&current=06-12-08_1729.flv

Keep in mind that this was whipped up for your enjoyment only and had I spent some quality time (we've aready established the rarity of said beast) on it you would would be watching an opposed pair of muscles on a joint of some sort moveing a load. You can (barely) see how the thing contracts anyhow. If it was under a load it would stretch another half an inch or so when the pressure is reversed.

^_^

Aries
06-16-2008, 02:44 PM
Sorry to come to this discussion a bit late but i have been so amazed by this community, having just joined, that I only now have found this thread. Around a month ago I had to experiment quite a bit with air muscles for a humanoid robotics project i was doing for my school. I do not have any pics of that part of the project so i hope you will take my word for it, i ended up being to busy to visually document it. However i believe i can address some of the concerns brought up in this thread from my own experiences.

In terms of pneumatically powered air muscles the problem of consistent air flow to power draw to bulk of equipment can be solved. When i hit that road block i discovered that two to three way solenoid valves controlled via something as simple as a basic stamp can do the trick, by using several solenoid valves spread out over the base of one's rig it can easily distribute the bulk and the flow. In terms of power one experiment i did ended up being promising for that, i never was able to finish it completely however as graduation preparations got in the way, but i hope this can help anyway. when i began to deal with the pump scenario i first looked up tiny pumps, etc. but very quickly found that i either had to spend way more money than i had or order ten thousand plastic ones from china(no thanks) so i ended up having to figure out my own. Now an air pump can be easily made with something as simple as an electric solenoid valve and a sealed tube, but to deal with power properly i spread the pump components over a larger surface area also limiting my bulk and began putting it together to work in stages. Now this does slow down the start of the system, but it saves on power quite well. If anyone has any thoughts on that part i would love some as i am still toying with that detail. Now i know this wasnt addressed but if anyone is trying to work out a place to store tanks for compressed air i found placing them inside the limb structure of the rig saved space, and made things a bit more efficient, in my case i had tanks acting as the bones of arms.

Now, in terms of hydrologically powering them i wasn't able to get to that stage while i was building them other then setting up a small rig to test. The only concerns i found with what i had made was the sealing of fluid but that isnt too hard even in the current method of constructing an air muscle, which if anyone is looking for a tutorial on, there are quite a few good ones on the web to help get you started. Thanks for the info and the video on the hydraulic rig you guys set up, it helped clear up some of my own concerns.

sthmck
06-17-2008, 09:40 AM
I've found some wonderful miniature sized PD pumps that would move the fluid from one muscle to another but these mechanical marvels are a bit too steep for my budget at the moment...

What specs do these pumps have, and where did you find them.

I was thinking that you could make a pretty big robot using hydraulic muscles. If you made big ones. Like say 1 inch in diameter, then if you hade maybe two for each side of a joint you could probably support a decent amount of weight. You should look into making a tutorial about your muscles, unless of course you are planning on marketing them. What I was think about was using a very small 12 volt hydraulic pump and an gas powerd engine to generate electricity. Im not sure if there is a pump like that, but what I had in mind was basically a hydraulic pump off of a snow plow. I know these are pretty heavy so they are probably not a very good choice.
If you were able to make a hexapod that uses hydraulic muscles instead of regular cylinders, it would cut down on the weight a lot. It would be something like big dog, but probably a little lighter.

sam
07-08-2008, 09:37 AM
Hi,

There is a tutorial on instructables for building air muscles, they could probably be modified for use with hydrolics ( http://www.instructables.com/id/How-to-make-air-muscles!/ ) In this tutorial the bigger muscles (3/8) can lift almost 50 pounds and the smaller (1/4) can lift 22 pounds about. So you could maybe get close to 100 pounds for a 1/2 in. ID rubber tube. If you place it at 3-4 inch from the center of rotation you could get almost 5000 oz.in. and 10 000 If you have 2. I think that how you calculate it right?

The problem with building 1 inch diameter air muscles is the price jump. For a 3/8 ID rubber tube on amazon is about 33$ for 10 feet. For a 1/2 ID rubber tube for 10 feet is 110$. And I cant seem to find any bigger. But bigger bladder (rubber tube) create much more power. Doubling the diameter will augment force of about 4-5 times.

Electricity
07-09-2008, 11:20 AM
Hey very cool sam. When you said air muscles I assumed you meant pneumatic rams, but those seem like they would work even better!
Might I suggest http://www.clippard.com/ for pumps. I'm still trying to figure out of the air flow for the e-valves are adjustable (i.e. put x psi into the ram, making it move only n% of its total travel) or if their just kind of on or off. Which could still be worked with ,you'd just have to turn them on for a few milliseconds.

ScuD
07-09-2008, 03:41 PM
That's one of the things i don't get.
Looking at it from a pure physics kind of way, i'd say with compressed air you can't put a pneumatic ram at position X by feeding it Y psi, since its position will depend upon the force acting on the ram externally, thereby compressing the gas inside the piston, giving you Z psi.

My mind keeps telling me the only way to do this is taking a sensor which tells the ram at what position it is, and then shutting down the air flow.

With hydraulics though, you could pump a volume of X to a ram, and knowing the bore of the ram you can calculate the position it would be in, or vice versa, within the percentage of compressibility of the fluid.

So, what am i missing here? :confused:

Electricity
07-09-2008, 03:53 PM
You're right on that regard, and its one point that I'm unsure of. one thing you COULD do is put position sensors of some kind, rotation sensor perhaps? (http://www.trossenrobotics.com/store/c/2676-Rotation.aspx) with big ol' brass gears for sweet ass looks on the joints themselves?

ScuD
07-09-2008, 04:04 PM
"Build it and they will come"... I know I will... you'd be my new hero if you could put something like that together!

sthmck
07-09-2008, 04:05 PM
Hey very cool sam. When you said air muscles I assumed you meant pneumatic rams, but those seem like they would work even better!
Might I suggest http://www.clippard.com/ for pumps. I'm still trying to figure out of the air flow for the e-valves are adjustable (i.e. put x psi into the ram, making it move only n% of its total travel) or if their just kind of on or off. Which could still be worked with ,you'd just have to turn them on for a few milliseconds.

They have pumps here? All I could find were valves.

Electricity
07-09-2008, 04:14 PM
They have pumps here? All I could find were valves.
The hard copy catalog I was looking through the other day had pumps I though? Maybe not.
The valves would be all you'd need for pneumatics. but you would need a pump for hydraulics.

sam
07-09-2008, 07:06 PM
That's one of the things i don't get.
Looking at it from a pure physics kind of way, i'd say with compressed air you can't put a pneumatic ram at position X by feeding it Y psi, since its position will depend upon the force acting on the ram externally, thereby compressing the gas inside the piston, giving you Z psi.

My mind keeps telling me the only way to do this is taking a sensor which tells the ram at what position it is, and then shutting down the air flow.

With hydraulics though, you could pump a volume of X to a ram, and knowing the bore of the ram you can calculate the position it would be in, or vice versa, within the percentage of compressibility of the fluid.

So, what am i missing here? :confused:

You are 100% correct for the air compression, you can't calculate the position just with the quantity and compression of the air. I was thinking of including a rotation sensor like Elec said to my mech for the wars, keeping stability without would be near impossible. (or very long and with a lot of pain included :rolleyes:)

As for the hydraulics, Can't a liquid under enough stress compress too? This could put you off in the calculation by a certain %.

Maybe I can't go to the 2009 Mech Wars but I will make it for sure the year after. I think building a robot with air muscles will be hard enough. I don't know why I find having something that pulls easier to control than something that pushes (like rams).

Adrenalynn
07-09-2008, 07:23 PM
Can't a liquid under enough stress compress too?

As memory serves, water compresses about 46ppm/atm.

Electricity
07-09-2008, 08:42 PM
As memory serves, water compresses about 46ppm/atm.
I guess its one of those things.. For our purposes, it really really doesn't matter.

Adrenalynn
07-10-2008, 02:19 AM
It might depending upon the fluid choice. Some are rather very compressible.

ScuD
07-10-2008, 02:23 AM
Be that as it may, it will still be a lot easier to use in an open-loop system than a gas.

Thing is, in robotics closed-loop will always have the preference for controlling limbs etc.

sam
07-10-2008, 07:26 AM
Yeah, That's for sure. There's someone at the begining of the thread that said that he had managed to control his air muscles with water instead and that a lot mor power came from it. Bt in this case you would actualy have to use pumps and all that, so your robot is still working on electricity and won't be very strong due to all the pumps and water tank.

ScuD
07-10-2008, 08:17 AM
Well i think you could do with only one pump, given a lot of valves that is.
But those valves still need actuation, so you'd still need some sort of motor/servo/solenoid.

I'm curious to the efficiency of hydraulic systems though, are we talking 50%, 70%, 90% ?

Maybe some day i'll try building a hybrid, with a small gas-engine to drive either a hydraulic pump or an alternator. Or both.

sam
07-10-2008, 08:28 AM
Yeah, I was thinking using servos for the actuation. One servo for 2 actuaors that are on opposite sides of the leg.

I can probably get my hands on some ram and try to figure out the efficienty. Could I try using a small tube the same diameter of the inside of the ram and exert 50 pounds of pressure in the tube. Then I would see the force that the ram pushes and then put that number on 50 to get a %?

I would guess that it would be over 80% of efficienty.

ScuD
07-10-2008, 08:42 AM
That would indeed be an indication of the efficiency of a ram.
I was thinking more about the efficiency of the pumps though, taking the amount of power you put into the motor and seeing what kind of power you get from it.

Not sure how you calculate pressure / flow to power though?

Adrenalynn
07-10-2008, 11:15 AM
Not sure how you calculate pressure / flow to power though?

I had to pull out my handy little pocket reference. The one I *cough* carry around in the bottom of my purse *cough*

For Hydraulic and Centrifugal pumps
HP = (CFM * PSF) / (33000 * EFF)
or HP = (GPM * PSI) / (1714 * EFF)
or HP = (GPM * head-feet) / ( 3960 * EFF)

EFF = efficiency [assume 78% if unknown]
GPM = gallons/min
CFM = cu.ft/min
PSF = pressure Lbs/sq.ft
PSI = pressure Lbs/sq.in

Once we have horsepower like this, I assume any other conversions are pretty easy, right?

sam
07-10-2008, 11:56 AM
Yeah! Thanks a lot :p

The parts to build air muscles should arrive next week so I'll keep you guys updated :wink:

I guess that the power difference if you use water is due to using a more powerfull motor to pump the water or fluid.

Electricity
07-10-2008, 11:59 AM
Yeah, I was thinking using servos for the actuation. One servo for 2 actuaors that are on opposite sides of the leg.

I can probably get my hands on some ram and try to figure out the efficienty. Could I try using a small tube the same diameter of the inside of the ram and exert 50 pounds of pressure in the tube. Then I would see the force that the ram pushes and then put that number on 50 to get a %?

I would guess that it would be over 80% of efficienty.
Thats not a bad idea with the 1 servo per 2 actuators actually. Although I think a positive preasure system with e-valves is still going to be the smallest.
If you'd like a ram to mess with, pm me your address and I'll mail you one, I've got a few sitting around.

Adrenalynn
07-10-2008, 12:19 PM
Yeah! Thanks a lot :p

The parts to build air muscles should arrive next week so I'll keep you guys updated :wink:

I guess that the power difference if you use water is due to using a more powerfull motor to pump the water or fluid.

Well, the calculations are for fluid systems above. I didn't look-up pneumatic. I could see efficiency differences with a less compressible fluid than a more compressed fluid, but I suspect those efficiencies may go out the window with the challenge of pumping a more viscous fluid.

MYKL
07-17-2008, 08:53 AM
Yeah, That's for sure. There's someone at the begining of the thread that said that he had managed to control his air muscles with water instead and that a lot mor power came from it. Bt in this case you would actualy have to use pumps and all that, so your robot is still working on electricity and won't be very strong due to all the pumps and water tank.

Sam,

My idea involves useing the fluid stored in opposing muscles rather than carrying around a water tank. A Positive displacement pump is conveiniantly inserted into the fluid line connecting said actuators to one another. When you run the pump in one direction (obviously the pump in question must be reversable) the fluid is removed from one muscle allowing elongation/relaxation and forced into its opposed twin causing contraction there. When the pump is reversed the muscles actuate in the other direction. I'm using a Glycol solution (Pet safe anti-freeze) for my fluid.

My first prototype included a couple of syringe-type pistons attached to a servo to move the fluid but while it looked kinda steam-punkishly cool it is not as effecient as using the PD pump. I'm thinking about building a hex with this setup though with brass tubing and wood overlayed servos so that it looks straight out of a steam-punk novel. Crab-fu inspired of course.

Another benefit to these actuators is thier built in shock absorbtion and compliant nature. They act as natural suspension. Like your own muscles do.

^_^

scorch
07-17-2008, 11:50 AM
Wow, this sounds really interesting.

I would love to see pictures and movies!

MYKL
07-17-2008, 12:35 PM
Here is a really crappy video of one of my Hydraulic Muscles:

http://s225.photobucket.com/albums/dd108/DrageSjel/?action=view&current=06-12-08_1729.flv (http://s225.photobucket.com/albums/dd108/DrageSjel/?action=view&current=06-12-08_1729.flv)

I am going to post a servo-driven Hydraulic muscle actuated arm assembly soon...

^_^

4mem8
07-17-2008, 01:15 PM
Nice, certainly has it's uses in robotics.

Matt
07-17-2008, 01:24 PM
Here is a really crappy video of one of my Hydraulic Muscles:

http://s225.photobucket.com/albums/dd108/DrageSjel/?action=view&current=06-12-08_1729.flv (http://s225.photobucket.com/albums/dd108/DrageSjel/?action=view&current=06-12-08_1729.flv)

I am going to post a servo-driven Hydraulic muscle actuated arm assembly soon...

^_^

Ummmmm, NSFW??? LOL

sam
07-17-2008, 02:13 PM
Sam,

My idea involves useing the fluid stored in opposing muscles rather than carrying around a water tank. A Positive displacement pump is conveiniantly inserted into the fluid line connecting said actuators to one another. When you run the pump in one direction (obviously the pump in question must me reversable) the fluid is removed from one muscle allowing elongation/relaxation and forced into its opposed twin causing contraction there. When the pump is reversed the muscles actuate in the other direction. I'm using a Glycol solution (Pet safe anti-freeze) for my fluid.

My first prototype included a couple of syringe-type pistons attached to a servo to move the fluid but while it looked kinda steam-punkishly cool it is not as effecient as using the PD pump. I'm thinking about building a hex with this setup though with brass tubing and wood overlayed servos so that it looks straight out of a steam-punk novel. Crab-fu inspired of course.

Another benefit to these actuators is thier built in shock absorbtion and compliant nature. They act as natural suspension. Like your own muscles do.

Ha! I hadn't thought of that one! I have been thinking about air muscles, wich would be hard to move compressed air from one muscles to the opposite. It's a good idea! :happy:

And now if I want to build a mech using this I just lited my load of about 4 pounds! Yippi! :veryhappy:

Yes, the muscles give nice smooth movement to the limbs or whatever. Do you have web site for small pumps that could be used in this set-up?

MYKL
07-17-2008, 02:17 PM
Matt-

You can't see what I had my wife doing off screen to make that thing do that.

Hmm... Maybe you can hear it I don't use sound on any of my machines...

^_^

Sam-

I have been searching high and low for an affordable miniature PD pump. AFFORDABLE being the key word. I'm drawing up plans for a delrin version so I can see how much it would cost to make my own pump head. But, as you can see my production rate is rather slow... In the meantime I am going to use my Steampunk servo to drive the muscles.

All I need is for something to take off so I can sell it and live finantially free as a creator of many fine things.

^_^

Electricity
07-22-2008, 12:49 PM
All I need is for something to take off so I can sell it and live finantially free as a creator of many fine things.

^_^
Arn't we all... :p

I'd be interested to see what you come up with as far as a miniature pump. If you need any input, I'm more then willing to share the little knowledge I possess.
-e

sam
07-23-2008, 04:00 PM
Well, I received most of the pieces I need to make some air muscles for one of my projects (top secret :rolleyes: ). Now all I need is to go to the hardware store.

Sadly I can't use water in this project, it wouldn't be fast enough and wouldn't work like I want it too. Plus, as you mentionned the valves look expensive. But if you get a lot more strengh (I don't know what a lot is, so I can't weight out my options here :confused: ) it might be possible to replace air with water.

Tomorow I will put together a couple of muscles and try to test them to see their limits. ;) I will take pictures and stats and maybe a video and I will post them here ASAP.

See ya ;)

sthmck
07-24-2008, 08:09 AM
I just got a catalog from clippard a few days ago. I was really impressed by the quality of the catalog and the fact that it was free. (what this has to due with anything I dont know but I thought i would throw it out there for what it was worth) They have a lot of nice stuff for sale and it looks like a lot of it could be usable. Unfortunately the price for 20 valves that I was looking at is 400 dollars. I am going to keep looking for other solutions and if I suddenly come across some extra change I might get some of them.

MYKL
07-24-2008, 08:34 AM
I have some of the tiny pnumatic cylinders that Clippard produces. They are very good quality. I picked them up at an industrial surplus house here in town. You would do well to check the biggest town near you for whatever valves/cylinders you need for experimentation. Also check E-Bay, people sell these things in Lots.

Second hand stuff works fine for puttering...


^_^

sam
07-25-2008, 02:28 PM
Hey Ervybody!

I did my first tests today with the air muscles. I don't have anough time to do them well... I didn't put enough braided tube. So it didn't contract as much as it could have... But it was very interesting!!

And it's not that expensive to make.

BADfish10
07-29-2008, 05:32 PM
I have access to loads of the outside sleave but i am wondering
Could you tell me what type of Tube you use material wise and where you have found to source it ?
Although this is not what i am working on now i am sort of maaking some plans :)

Cheers

J

MYKL
07-30-2008, 08:36 AM
Look online for a latex tubing supplier. McMaster Carr has various sizes and you can get some pretty robust stuff from scuba gear suppliers. I have also found great success with the balloons that can be twisted into animals. The come in a few sizes, are made out of stronger latex, and are fairly easy to find. You should be able to find some at a local party supply or Target type store if the internet isn't option.

^_^

Electricity
07-30-2008, 01:51 PM
I had a thought a couple of days ago, and kept forgetting to post it.
Going back to the linking the opposite sided muscles (front and back of the leg) so that when one drains the other fills, what if you used a sealed cylinder with a solid divider in the middle that was attached to a worm gear. Each muscle would be attached to each side of said cylinder, turning the worm screw one way causes one muscle to contract and the other to expand, and vise versa.
A quick paint job-
http://i34.tinypic.com/14yaao1.jpg

MYKL
07-30-2008, 02:39 PM
The problem with this occours where the screw pierces the cylinder. It will be virtually impossable to get a good seal under pressure at that point. Try using a servo and two seperate cylinders or use a single dual acting cylinder with its standard set of seals. Keep in mind the volume differance on the rod side of the plunger...

The only way to make a truely efficient version of what you propose IMHO is to suspend a 'puck' inside of a completely sealed cylinder that would be moved from an external force (at this point only a magnetic coupling would be feasable) And actuate it via linear motor or servo motor.

I wonder, couldn't threads be machined on the interior and the 'puck' be the actual armature of an brushless electric motor? Sealing the threads is the wrench in the design again I think.

I've also thought about an uber-pressure peristaltic type pump but have not tried it yet.

The best bet is a very efficeint gearstyle PD pump. Ever heard of a Gerotor pump? Pretty neet-O...

^_^

ooops
07-30-2008, 04:43 PM
If you don't mind me jumping into the conversation;
How large or small are the "muscles" that are being fabricated?


Try using a servo and two separate cylinders or use a single dual acting cylinder with its standard set of seals.

Just a thought, and not exactly sure (meaning I haven't done any math to show this but I am pretty darn positive), but I would suspect that the energy required to inflate the muscle from the mechanical servo-plunger system would be greater than the energy to just actuate what ever you are moving with the servo. The one positive note would be if you were mechanically actuating a large plunger, you could inflate a number of small "muscles" simultaneously, again the total mechanical force of all the small muscles would be less than the mechanical force applied to the large plunger.
The little 12v "emergency auto style" pump would give you lots of pressure in a small package, and depending on how big your project is, an empty plastic Diet Pepsi bottle 12oz - 2 liters will hold an amazing amount of pressure. (any brand is fine, I just prefer D.P.) Obviously the bigger the bottle the higher the volume. Just pick up a rubber valve stem from the auto parts store, drill a hole in the cap smaller than the valve ... actual size will depend on the valve stem - use your best judgment. You will need to remove the valve core and come up with an adapter from the threaded part of the valve stem to what ever hose(plumbing) you are using and you have a low cost fairly high pressure and volume air system.

Note* if you over inflate the bottle you will know it immediately by the extremely loud boom:) I wouldn't think you would need to be anywhere near the burst pressure. I personally have only blown up one bottle that we were filling half full of water and shooting as a rocket. As I recall it was somewhere near the 180psi mark that it burst. That after repeated launches. I wasn't holding it at the time, but was near enough to get wet and suffer a loud ringing in my ears for a day or so.

MYKL
07-30-2008, 04:58 PM
I may have mislead.

All of the suggestions I've made are based upon using a fluid rather than a gas in the muscles. At the moment I am using distilled water 'cause its cheap and not messy. I have a bottle of a glycol solution that I plan to use as soon as I finish my 'Steampunk' version of a servo-driven pump (two small cylinders).

As far as my experiance and opinion goes, using a gas (air/methane/wuteva) in these things is at the moment an exercise in futility.

^_^

ooops
07-30-2008, 05:30 PM
I may have mislead.



Well, I have followed the thread since it began, but may have gotten lost in the conversation somewhere. Sorry if I misunderstood!
I am very interested in the DIY air muscle. And would love to participate... from afar.
My expectation is that the projects that would use them would be large enough to handle a pump of some sort and an air tank / fluid tank. Which was the main point of my last post ... cheap and light DIY tank:)

MYKL
07-31-2008, 08:49 AM
You were entirely right to point out the limitaions for what I had suggested had I been using air/gas. In the case of using such a medium I believe that your suggestion to store the air in a bottle makes a lot more sense than toting around a pump. There are some bottles on the Paintball fields that can handle 4500 PSI and are constructed from wound fibers so are fairly light. The valves are getting smaller too. I think the air version of the muscle would be fine where the object being actuated would have a fairly repeatable and known load. The programming and expected range of motion could be accomplished with preordained amounts of pressure. But when you are talking about a limb that may have to pick up an object or carry an ever changing load the compressable gas muscles will be very hard to control.

^_^

Electricity
07-31-2008, 11:46 AM
You were entirely right to point out the limitaions for what I had suggested had I been using air/gas. In the case of using such a medium I believe that your suggestion to store the air in a bottle makes a lot more sense than toting around a pump. There are some bottles on the Paintball fields that can handle 4500 PSI and are constructed from wound fibers so are fairly light. The valves are getting smaller too. I think the air version of the muscle would be fine where the object being actuated would have a fairly repeatable and known load. The programming and expected range of motion could be accomplished with preordained amounts of pressure. But when you are talking about a limb that may have to pick up an object or carry an ever changing load the compressable gas muscles will be very hard to control.

^_^
Yeah Compressed Air tanks have come a long way in recent years. You can pick up a 45CUx4500psi tank that weighs in just under 1.5lbs with the regulator now adays..
And i think, for me atleast, the interest in this form of locomotion is simply because its cool, not because its any better or worse then another way of doing it.

ooops
07-31-2008, 11:57 AM
But when you are talking about a limb that may have to pick up an object or carry an ever changing load the compressible gas muscles will be very hard to control.

^_^

You are 100% correct. Having years of first hand pneumatic experience, I can tell you it may be cool but it won't necessarily be practical.

I am intrigued with the DIY air/fluid muscles, and I am hoping to see more details at some point.

sam
07-31-2008, 07:34 PM
I don't know if I showed this to you guys yet but there is a great site ( http://www.instructables.com/id/How-to-make-air-muscles!/) That explains very well the project. I did one myself, but I haven't found the same results. My muscles doesn't contract has much. It can lift 50 pounds (more than 400 times it's own weight) about 1.6 inches up. It's about 12 inches long. It contracts about 13% of it's lenght. :cool: It could probably better though. :sad:

For my project it doesn't matter much how air is compressable. :happy: Good news of the day!

I was wondering about something tough. In a fire truck, tu push out the water it uses a big pump that pumps the water out?

Oh, I have a last questions for you guys. I'm looking into a way that I can control my muscles with more ease (tried to build valves myself... what a mess ;) ), I found this site : http://www.imagesco.com/catalog/airmuscle/AirMuscle.html? _____ Wich sells (you can scroll down a bit) a solenoid 3 way valve for 35$. Did any of you tried this or have experience with products that are similar? How well do you think it will work for the price considering I see an average of over 150 for 3 way solenoid valves. :confused:

Thanks :D

sthmck
07-31-2008, 08:43 PM
Oh, I have a last questions for you guys. I'm looking into a way that I can control my muscles with more ease (tried to build valves myself... what a mess ;) ), I found this site : http://www.imagesco.com/catalog/airmuscle/AirMuscle.html? _____ Wich sells (you can scroll down a bit) a solenoid 3 way valve for 35$. Did any of you tried this or have experience with products that are similar? How well do you think it will work for the price considering I see an average of over 150 for 3 way solenoid valves. :confused:

Thanks :D

Hey, I saw that vex has a pneumatic system that has solenoid air valves. I am not sure exactly how much they cost though, because the valves were offered as part of a kit. I am going to try to find out if they can be bought by themselves. I am also looking for them on ebay. Another option would be to check out www.clippard.com. Some one else mentioned them in an earlier post but they should be mentioned again. You can get a very nice catalog from them for free. It doesnt have any prices listed in it, but it you can find most of the stuff in the catalog online. I found a valve that I am going to try and they listed it at $20.

I have some good news. I just got the go ahead from one of my proffesors to start a research project that involves air muscles. I have now have to create a research proposale so that I can get some funding. Part of that includes making a parts list (shopping) Hopefully I will be able to start this project as soon as the fall semester begins. I will be sure to keep everyone updated.

ooops
08-01-2008, 09:04 AM
I don't know if I showed this to you guys yet but there is a great site ( http://www.instructables.com/id/How-to-make-air-muscles!/) That explains very well the project.

Perfect link ... exactly what I was asking for!

BADfish10
08-01-2008, 10:04 AM
Yeah Compressed Air tanks have come a long way in recent years. You can pick up a 45CUx4500psi tank that weighs in just under 1.5lbs with the regulator now adays..
And i think, for me atleast, the interest in this form of locomotion is simply because its cool, not because its any better or worse then another way of doing it.

I have just been looking at them from a paint ball supplier in the UK a 1.1L 4500psi 68ci £125
not to bad and if the weight is 0.7KG including the valve system.
but i have a qustion what would you use to fill the Bottle have you seen a small unit around what sort of hp is needed for the compressor to work?

if you have any input it would be very intresting

J

sam
08-01-2008, 10:46 AM
Easy! :veryhappy: Well, not quite. You can use a pump that can pump at that pressure. Air hog sells one that can pump up to 3300 PSI. Here's a link :http://www.airhog.com/af_pump.htm
There are also some hrydrolic pumps tht some companies sell that can easily go up to 4500 PSI. Just google it. Sorry I don't ahve a link :tongue:


Other than that, if you don't want to pump it yourself there are some compressors that can do the job but you need to send a request for the information and the price, so I assume that it's pretty expensive. :eek: Who would have thought?

That's the problem I saw with those tanks, there pretty expensive (not that much when you think of high-end servos) and then there are MORE fees for the compressor or hand pump. Otherwise I would love to use one for my project using those muscles, it would be much more convienient than that 20 pound 7 gallon air tank (125 PSI max). ;)

125 pounds looks expensive. Here it's about 125$. Maybe if the shipping isn't to expensive you can order it from here?

sam
08-01-2008, 10:51 AM
Perfect link ... exactly what I was asking for!

No problem.

I was lookingf at the link sthmck gave me. It looks very interesting indead! But after reading a bit I got confused : a 3 way solenoid valve, is it :

A) one port you plug into the air supply and the valve can direct the air in either port b or c or close the valve

OR

B) A valve that port A is the air supply, B to the actuator (air muscles in this case) and the valve C is for realssing the air from the air muscles.

I have the feeling that it can be both, but how do I know wich is wich when I buy a 3 way solenoid valve? :confused: Help?

Electricity
08-01-2008, 11:48 AM
No problem.

I was lookingf at the link sthmck gave me. It looks very interesting indead! But after reading a bit I got confused : a 3 way solenoid valve, is it :

A) one port you plug into the air supply and the valve can direct the air in either port b or c or close the valve

OR

B) A valve that port A is the air supply, B to the actuator (air muscles in this case) and the valve C is for realssing the air from the air muscles.

I have the feeling that it can be both, but how do I know wich is wich when I buy a 3 way solenoid valve? :confused: Help?
It depends on the valve, some are made for different things. Clippard has a pretty good explanation somewhere on their site.


As far as DIY air fills, the cheapest way is to buy a scuba tank or two, and build yourself a little fill station, you will get ~8 fills at ~3000 psi (with the last few fills being slightly less).
Generators and compressors are going to run you upwards of $4000 if not a lot more. Its a big investment. You could look for a closing paintball field/shop, and they might have an old compressor they'd sell for cheap.

ooops
08-01-2008, 12:10 PM
Page 1208 in the Grainger catalog has a very good technical terminology page. I couldn't find the page on their website, but here is the quick splanation:
2-way valve has two ports (on or off valve - garden hose comes to mind)
3-way valve has three ports with two flow paths - (1 in 2 out) there are normally passing and non-normally passing.
I assume these are (based on the picture of the manual 3-way) normally passing
So, when the valve is open it flows pressure in through the "in" port (port 1) to the "out" port (port 2) and blocks the exhaust port (port 3).
When you close the valve it would block the "in" port and allow the built up pressure behind port 2 to exhaust through port 3.
Hopefully that makes sense:)

sthmck
08-02-2008, 08:37 PM
Page 1208 in the Grainger catalog has a very good technical terminology page. I couldn't find the page on their website, but here is the quick splanation:
2-way valve has two ports (on or off valve - garden hose comes to mind)
3-way valve has three ports with two flow paths - (1 in 2 out) there are normally passing and non-normally passing.
I assume these are (based on the picture of the manual 3-way) normally passing
So, when the valve is open it flows pressure in through the "in" port (port 1) to the "out" port (port 2) and blocks the exhaust port (port 3).
When you close the valve it would block the "in" port and allow the built up pressure behind port 2 to exhaust through port 3.
Hopefully that makes sense:)

I was going to get three way valves but then I saw that you can get mechanical exhaust valves. I will probably design my muscle so it has the exhaust valve built in to it. I will probably get a few 10 mm valves from clippard and a few ehaust valves. I want to see how fast the the muscle will react before I go any further. I think the valves that I want to use may be to small but I will have to wait till they come in.

sam
08-03-2008, 09:51 AM
So, the exaust is to get out all the air from the muscle to the outside air. Right?

Me too I need to know how fast it can intake and exaust. :happy:

ooops
08-03-2008, 11:38 AM
So, the exaust is to get out all the air from the muscle to the outside air. Right?

Exactly:)

sam
08-03-2008, 11:48 AM
Ok, and how do I contrl this little bugger? It has two wires? Can I conrol it like a servo? Or just inverse the signals?

sthmck
08-03-2008, 03:30 PM
I am not positive, but I think that it just needs is a 6 volts power supply. My guess is that you could send a pulse to the valve and the duration of the pulse will determine how long the valve stays open or closed, depending on the type of valve obviously.

sam
08-03-2008, 04:25 PM
Well, ok. I'll look a bit more in those little things then I will buy myself 3. See how they work.

sthmck
08-03-2008, 09:08 PM
which ones are you planning on buying?

sam
08-04-2008, 07:38 AM
I was looking at this one : http://www.clippard.com/store/display_details.asp?sku=MME-32QES-W012

It let pass more than enough air.

There is also the 2013 series that are 3 way with 22 Scfp. But the low price (7$) makes me want to believe that it isn't actualy a valve? I think I will send an e-mail to clippard.

What do you guys think? Is this a good choice? Does it open/close fast enough?

ooops
08-04-2008, 08:08 AM
Ok, and how do I control this little bugger? It has two wires? Can I control it like a servo? Or just inverse the signals?

You most likely will want to operate them via relays: http://www.trossenrobotics.com/relays-switching-controls.aspx

I linked the relays-switching page since there are a number of different options of relays depending on what type of controller you are using.

I noticed that the one you linked is a 12v unit and although it has 3 watts of "consumption" I think the amp draw to actuate it is going to be much higher. Something to bear in mind when you are spec'ing relays.

BADfish10
08-04-2008, 09:31 AM
Well I have just ordered me a catalog from clippard will be interesting to look through!

What sort of PSI are you going to be running yours at and is there an expected cycle time you are aiming for?

I cant wait for that catalogue now

J

sthmck
08-04-2008, 09:42 AM
Yeah I was thinking I am probably going to have to use relays too. I was kinda thinking that maybe the valve that I was looking at wouldnt need one. I guess I will have to take a look at all the numbers to make sure. I am going to run some test to determine muscle and joint reaction times. I am thinking about setting up my system so that each joint has two muscles for each direction. One would be a big muscle that can handle a lot of weight and the other would be a small muscle that can actuate faster. Kinda like a fast twitch muscle. Still have a lot of stuff to do before I get to that point though.
I will take look at the valve you are talking about sam. 7 dollars does sound a little to good to be true, but maybe it will work.

sthmck
08-04-2008, 09:47 AM
Also I noticed everyone has been linking a video of an air muscle that some made and entered in that contest on instructables. I dont know if anyone else saw but there was another air muscle entry. I thought that it was cool because the guy used some brass or copper fittings on his air muscle. I was trying to figure out a way to close the ends of my muscles and I think that this might be a pretty good way of doing that. http://www.instructables.com/id/Pneumatic-Muscles/

sam
08-04-2008, 10:01 AM
Well I have just ordered me a catalog from clippard will be interesting to look through!

What sort of PSI are you going to be running yours at and is there an expected cycle time you are aiming for?

I cant wait for that catalogue now

J


Well, I read somewhere on the net about shadow air muscles (that's the same thing we are doing here about) That wear and tear much faster when used at 6.83 bar (DUH:rolleyes:... PS 150 PSI about) but it seemed like they could go for ever at 45 PSI (2 bar). For the test to see how much they contracted, I ran it @ 60 PSI. I have some heavy duty silicone so it looks like I will run it continuosly at 60 PSI. I even ran it @ 90 PSI for a while. Didn't see enough difference to consider it. :wink:


Sorry for my innosence in this domaine, but how would I use relays to make it work?

sam
08-04-2008, 10:23 AM
Also I noticed everyone has been linking a video of an air muscle that some made and entered in that contest on instructables. I dont know if anyone else saw but there was another air muscle entry. I thought that it was cool because the guy used some brass or copper fittings on his air muscle. I was trying to figure out a way to close the ends of my muscles and I think that this might be a pretty good way of doing that. http://www.instructables.com/id/Pneumatic-Muscles/

Yeah! It's cool the brass but I just noticed that he put the sleeve 1/2 inch less than the tubing (wich I am not doing.) I was putting about 2.5 times more sleeve than tubing, so it was not contracting like their muscles on the videos.

So I will try this afterward. The lasst test I did was with cheap long ballons (the ones you make animals with). I put one in the middle of the sleeving and put air in it. It contracted and lifted 20 pounds maybe 2 inches @ 40 PSI. Then it got holes and it was useless. But I had fun with it for a few minutes! :veryhappy: there was to much wear for the little latex.

Oh and here's how I connected my muscles to hold it togeter:

http://forums.trossenrobotics.com/gallery/files/1/5/4/9/picture_324_thumb.jpg (http://forums.trossenrobotics.com/gallery/showimage.php?i=608&c=3)

http://forums.trossenrobotics.com/gallery/files/1/5/4/9/picture_322_thumb.jpg (http://forums.trossenrobotics.com/gallery/showimage.php?i=607&c=3)


http://forums.trossenrobotics.com/gallery/files/1/5/4/9/picture_326_thumb.jpg (http://forums.trossenrobotics.com/gallery/showimage.php?i=610&c=3)

I used little metal srewy things. It works really well, you can find it evrywhere and it's cheap.

ooops
08-04-2008, 12:20 PM
Sorry for my innocence in this domain, but how would I use relays to make it work?

The two leads from the solenoid valve would be connected to the output (High voltage) side of the relay.
Controlling the relays will depend on what type of controller -

Computer controlled - http://www.trossenrobotics.com/store/p/3203-InterfaceKit-0-0-4.aspx

RC controlled - http://www.trossenrobotics.com/store/p/3206-Electronic-R-C-PWM-Switch-RC-110-.aspx

Those are just examples but hopefully you get the idea:)

Electricity
08-04-2008, 12:26 PM
Nice job Sam! I send some rep to yeah. I'd love to see a video of your muscles in action when you get them working well!

sthmck
08-04-2008, 12:39 PM
I agree looks like you have made some progress. Keep up the good work.

Adrenalynn
08-04-2008, 01:02 PM
The two leads from the solenoid valve would be connected to the output (High voltage) side of the relay.
Controlling the relays will depend on what type of controller -

Computer controlled - http://www.trossenrobotics.com/store/p/3203-InterfaceKit-0-0-4.aspx

RC controlled - http://www.trossenrobotics.com/store/p/3206-Electronic-R-C-PWM-Switch-RC-110-.aspx

Those are just examples but hopefully you get the idea:)

I know where to find some 8 channel relay boards at much higher DC current, serial or USB driven. ;)

sam
08-04-2008, 01:30 PM
I guess I will do a computer controlled one. I guess I will figure out how tocontrol them (actual code I mean :tongue:) when I get them.

Those air muscles are not the "finished" version. I dissambled them to try the ones with balloons. I can post pictures and videos Once I get a good version. Maybe tomorow. :happy:

Well, I almost finished another version.

Adrenalynn
08-04-2008, 01:46 PM
Make sure you're not drawing more DC current than those relays will handle. How much amperage are you drawing? Those relays are pretty small.

I just got done designing/prototyping/sending out for fab some 8 relay boards for Ooops and my projects.

sam
08-04-2008, 02:12 PM
It's supposed to be 12V 3 watt. Maybe it's more than the actual numbers, but I still have a pretty big margin of error with that relay.

For the people who have muscles, what percentage do they contract? i have just tested (again) mine and I get 15% contraction with 40 pounds. If you look on the shadow robotics site, it has more like 23%. just to bump it up 3-5% would already help a lot. :rolleyes:

sthmck
08-04-2008, 02:18 PM
What was the operating pressure that the shadow AM was using, and how much weight was it lifting?

sam
08-04-2008, 02:22 PM
Here is the chart : http://www.shadowrobot.com/airmuscles/techspec.shtml

I have a 20 mm air muscles. I have no idea what they do that makes such a difference. It's both @ 60 PSI (4 bar).

sthmck
08-04-2008, 02:28 PM
Im not sure either. When you were cutting your mesh and silicone did you make the mesh longer than the silicone? Im pretty sure that is how your supposed to do it. Its been a while since I made one, and I just read where someone said to do the opposite.

Adrenalynn
08-04-2008, 02:50 PM
Wow. Yeah, if it's 3 watts you have lots of headroom. 3/12 = 0.25A = 250mA (P=V*I therefore I=P/V)

I'd double check it though. That's not much for a pump motor...

sthmck
08-04-2008, 02:59 PM
I think that is for the valve, but I may be wrong.

Adrenalynn
08-04-2008, 03:02 PM
Ahhh, ok, my bad. That makes sense. I was solving the wrong problem. :)

sthmck
08-04-2008, 03:04 PM
right now the one i am looking at only uses .6 watts at 12 volts.

Adrenalynn
08-04-2008, 03:08 PM
0.05A - run that off of the static electricity in your socks. :)

There is a downside though - that might not have enough current draw to actuate a relay. That's more a job for discreet ICs.

sthmck
08-04-2008, 03:10 PM
I was wondering if it could be actuated using just the power coming from an io pin on something like a basic stamp.

sthmck
08-04-2008, 03:11 PM
ok wait that was stupid. basic stamp doesnt output 12 volts from its pins.

sam
08-04-2008, 03:13 PM
Im not sure either. When you were cutting your mesh and silicone did you make the mesh longer than the silicone? Im pretty sure that is how your supposed to do it. Its been a while since I made one, and I just read where someone said to do the opposite.

Yeah, I get a better percentage when I have about 2.5 times more mesh than silicone, but i still have aobut 8% to have the same as they have. Maybe I have to buy a bigger mesh sleeve. But I don't think it will do a difference because I tried with a 1/4 by 3/8 latex tubing instead and it gave the same results.

Yeah, it's only a valve, so open/close, no pumping thank god. :tongue:

Yeah I was thinking of doing a small thing to control it.

Adrenalynn
08-04-2008, 05:27 PM
You could just use something like a 4N25 optocoupler I'd think. Output from the stamp fires the LED through a pullup resistor, 12v goes into the sensor and the output goes to the device. Pin goes high, internal LED lights, resistance goes down on the photo cell and current flows. Solid state relay.

sam
08-04-2008, 09:36 PM
wow. Althought I read waht you said for a few minutes over and over, I didn't get a thing. :confused:

I'm thinking of making larger muscles and buying at the same time the valves.

On the shadow site, the 30 mm muscle can lift 70 kilos (155 pounds) with 27% contraction.

30 mm is the 1" nominal techflex sleeving, 3/4 " min ///// 1 3/16 Max. I'm wondering Wich size late tubing I should choose to go wich it. I have a choice between (Outside Diameter mentionned only) :
-11/16"---- 3/32 Wall ------ .5" expansion before the sleeves stops it.
-3/4" ---- 1/8 Wall ------ .44" Expansion before the sleeves stops it.
-7/8" ---- 1/16 Wall OR 1/8 wall ------ .3125" Expansion before the sleeves stops it.

In my past the maximum expansion I have tried with latex is 0.375" (3/8"). What do you guys think I should do to have a good expansion? :rolleyes: The best one... :veryhappy:

An advantage would be a smaller wall so air is more lifting and less spending at expanding thick latex. But, if too small, then it will wear and tear much faster. Then you have the expantion. The Latex only has a certain point before exploding. I don't want to pass this, but the further = the more contraction.

I will also need new options to hold the muscles together because with 150 pounds on it, it will have a rough way ahead of it. ;) Any ideas?

Please tell me what you think will hold up, wich is the best option in your opinions! :happy:

Good knight ;)

EDIT : Here is the site for a chart :

Tehcflex sleeving : http://www.techflex.com/prod_CCP.asp

ooops
08-05-2008, 11:46 AM
An advantage would be a smaller wall so air is more lifting and less spending at expanding thick latex. But, if too small, then it will wear and tear much faster. Then you have the expantion. The Latex only has a certain point before exploding. I don't want to pass this, but the further = the more contraction.

My first thought would be a thin bicycle inner tube cut to length. When cut to length it will be sorta curved, but that shouldn't be an issue for the lengths you are using. It should expand predictably using much lower pressure, which in my mind should/could add some longevity?

Electricity
08-05-2008, 01:35 PM
My first thought would be a thin bicycle inner tube cut to length. When cut to length it will be sorta curved, but that shouldn't be an issue for the lengths you are using. It should expand predictably using much lower pressure, which in my mind should/could add some longevity?
I would think because its so thin, it would counteract the lowered pressure, giving you the same durability, if not even less do to more environmental factors effecting it.

ooops
08-05-2008, 02:58 PM
My first thought would be a thin bicycle inner tube cut to length.

I guess I should have made the "thin" more clear ... thin like a skinny tube for a skinny tire. The wall thickness of the tubes will be pretty similiar for any bicycle tire. Sorry for the clear as mud discription:)

sam
08-05-2008, 03:47 PM
Hello!

i have already tried to make it work with a bicicle tire that I bought. It was leaking to much air and didn't seems to work propely. Like the tube didn't elongate so when it contracted it contract anywhere near 20%.

I made this chart (after spending a lot of time figuring stuff out on Exel ;) )

http://forums.trossenrobotics.com/gallery/images/file_bmp.gif (http://forums.trossenrobotics.com/gallery/showimage.php?i=614&c=3)

EDIT : You have to click on the image to see the chart.

I know it's written in french! Don't panic! :tongue: Y axis is the percentage of contraction of the muscles. The X axis is the weight it was lifting (in pounds). The red colum is the muscles I made with 3/8 ID X 1/2 OD silicone and the 3/4" max sleeve (silicone is 50A of roughness). The Yellow colum is 1/4 ID X 3/8 OD latex with the same sleeve (3/4" max) and 35A roughness. Usualy a bigger bladder means more power but I guess since the latex is softer and the sleeve gives it enough room to expand, it contracts almost the same thing as the other one wich is much bigger bladder.

After thinking a lot I think this option is the best :

-7/8" ---- 1/8 wall ------ .3125" Expansion before the sleeves stops it.

It will give me almost as much as the smaller one (in the chart) but less in case the bladder fatigues itself by working to much.

Electricity
08-05-2008, 05:22 PM
I was wondering what a poid was.. :p

sthmck
08-05-2008, 06:20 PM
I took two years of french in high school. Thought I didnt remember anything. I kinda looked at it and thought Im pretty dure that I am reading french here.

sthmck
08-05-2008, 08:07 PM
with that graph it looks like that heavier load seems to contract the same for both muscle types. I don't know if anyone has seen the festo muscles, but they are cool because the mesh is built into the rubber. They claim that it increases wear resistance by some impressive number. Probably does not that I can make muscles like that.

sam
08-06-2008, 06:55 AM
Yeah! I heard about the festo muscles! There pretty cool but I don't think I can do them :sad:

Did you see the arm taht they did with festo muscles? It looks so much like a real arm. It has 4 muscles to do the pec, 3 for the biceps. It's really nice.

MYKL
08-06-2008, 09:02 AM
The Festo version is very useful. There is a group, it may have been at MIT that made a roach-type configuration with smaller festo muscles. In fact it was watching the roach flop around under air power that made me start experimenting with fluids rather than gasses. Have you seen the giant ones they've set up to move whole pallets of crap around a warehouse?

Have any of you tried sealing your muscles from the outside with latex? That is one of the next steps I'm moving on to. I've had great success with the balloon bladder but like mentioned before the mesh will eventually wear it out. So why not make the muscle then soak/coat it in latex for a final sealing? I've envsioned making it look more like a real muscle this way by shaping and colouring the latex a bit.

How are the air valve assemblies coming? Any videos yet? My wife is going to be out of town for TWO WHOLE WEEKS and I should be able to finish a project or two during that hiatus. I'm hoping to do a servo driven muscle assembly in that time. My actuators are a lot smaller that what you all are looking at I think. My ultimate initial goal after all is to make a muscle driven steam-punk spider assembly.

I think its awesome to find a group of people pursueing this avenue of actuation. I've been futzing around with different forms of it for years. I hope we can break some plateus together.

All I need to make a dream muscle is a material that expands concurrantly with an increase/decrease in charge and returns to its original state when the current is removed. If you find it I will...

^_^

ooops
08-06-2008, 10:47 AM
My wife is going to be out of town for TWO WHOLE WEEKS and I should be able to finish a project or two during that hiatus.

LOL ... I wouldn't get anything done unless she took the kids!!!


I think its awesome to find a group of people pursueing this avenue of actuation. I've been futzing around with different forms of it for years. I hope we can break some plateus together.


I agree, it is very cool that others are interested in this! I just wish I had more time to mess with it at the moment!

Just curious has anyone tried inflating the tubing with the reinforced braiding built in? I think it is used on the kitchen faucets. It is mostly clear, but has some braid looking fibers built in. It may just burst instead of expand ...

I am excited to see pictures of your "steam punk"! I think I am envisioning it correctly, do you have one-way valves or something to hold the pressure?

sam
08-06-2008, 11:21 AM
Hey!

I agree it's fun not to be alone in this. So if someone needs help ;)

I have looked at those tubes. But the only ones I see are PVC with braided mesh. My PVC alone does not expand, so with a mesh I highly dought it. Maybe with very high pressure it would burst! :p

I would be interested in the festo muscles but I can't seem to find a way to make them and time is short for me. I has a month to finish my project. If I can find a way to make them before friday. I will buy the parts and try it out. Otherwise You're going to have to try something :cool:

I looked at his way of tiying the muscles end (http://www.instructables.com/id/Pneumatic-Muscles/ ). I am going to buy materials to make 7/8" OD bladder with 1 1/4" Mesh. I looked at the same things he used, but when it has to be one inch the price jump is astronomical.

Sam

Electricity
08-06-2008, 11:44 AM
Man.. I have all the time in the world, and no money to do anything cool. So I'm resigned to sit back, and watch you guys create!
Heres a thought that popped into my head. Going back to bleeding one muscle into another, to determine limb movement, what if you did a loop of some kind. i.e.
Pump->outside muscle->Valve of some kind-> Smaller inside muscle->back to begining.
That way, you pump to the first muscle, the leg extends, you open the valve between muscles, and the pressure equilizes between both muscles, but because the inside muscle is smaller/weaker, it expands fully, and the outside muscle deflates enough to release.
In a situation where you don't necessarily need a every ounce of strength you can get, this would work pretty well. Especially for something like legs, where you could then do if very fast, and create a running gait.. You'd want the more powerful muscle for lifting, and the weaker one simply to reset the position.
That way, you could still use the more powerful muscle for lowering if you needed the full strength, you would just slowly (in short bursts I guess) equalize the pressure..
Does that make sense? It seems to work pretty well in my head.

MYKL
08-06-2008, 11:56 AM
Sounds interesting. Try it. With a fluid the pressure will equalize like you said. If you used a finer bladder in one muscle over another it would probably assist in the effect you are going for.

There is no actual steam involved in mine but brass, copper and wood clad servo driven cylinders that hydraulicly energize biomemetic muscles of miniscule proportion. Some of the DOF on the body will be directly actuated directly by servos and the muscles on the thin limbs will be Servo-Hydraulic.

^_^

sam
08-06-2008, 01:04 PM
Indeed it sounds interesting. But honnestly I can't quite imagine it. It would be nice if you could try it ;) Parts from mcmaster.com are very cheap and youcan find almost evrything there.

I will upload a video a bit later on made with my latex air muscle. I just have to Take a couple of pictures of it (building it) then I will make a vid.

Ok, I have two questiion :

1-Do you think if I use a bigger bladder (latex tube) and a bigger braid, will the muscle be stronger?

2-The more I read on the subject the more I'm not sure what to use to make the muscles. Many different site tell you to take a different diameter of braided sleeving. It varies from 1/4" more than the bladder up to 1" more. I'm really not sure what to use. The one that says to use 1" sayd to cut the sleeving 1/2 inch less than the latex. I guess the sleeving just gets longer since it's a bigger diameter. What do you guys think?

Sam

Electricity
08-06-2008, 01:20 PM
With a fluid the pressure will equalize like you said. If you used a finer bladder in one muscle over another it would probably assist in the effect you are going for.
Thats what i was going for.. :)



Indeed it sounds interesting. But honnestly I can't quite imagine it. It would be nice if you could try it ;) Parts from mcmaster.com are very cheap and youcan find almost evrything there.

I will upload a video a bit later on made with my latex air muscle. I just have to Take a couple of pictures of it (building it) then I will make a vid.

Ok, I have two questiion :

1-Do you think if I use a bigger bladder (latex tube) and a bigger braid, will the muscle be stronger?

2-The more I read on the subject the more I'm not sure what to use to make the muscles. Many different site tell you to take a different diameter of braided sleeving. It varies from 1/4" more than the bladder up to 1" more. I'm really not sure what to use. The one that says to use 1" sayd to cut the sleeving 1/2 inch less than the latex. I guess the sleeving just gets longer since it's a bigger diameter. What do you guys think?

Sam
If you have the funds, you should try experimenting with different lengths and such and see what you come up with.

sam
08-06-2008, 02:51 PM
Well, I don't have a lot of funds. I will be buying 2 or 3 different mesh sizes to see wich one goes best. they also say that a mesh with a lighter patern will be more efficient. I will try the standard and the clean cut. No heavy duty.

Well I have a couple of picture for ya'll :veryhappy:

http://forums.trossenrobotics.com/gallery/files/1/5/4/9/picture_340_thumb.jpg (http://forums.trossenrobotics.com/gallery/showimage.php?i=615&c=3&userid=1549)

http://forums.trossenrobotics.com/gallery/files/1/5/4/9/picture_345_thumb.jpg (http://forums.trossenrobotics.com/gallery/showimage.php?i=616&c=3&userid=1549)

http://forums.trossenrobotics.com/gallery/files/1/5/4/9/picture_346_thumb.jpg (http://forums.trossenrobotics.com/gallery/showimage.php?i=617&c=3&userid=1549)

ooops
08-06-2008, 03:11 PM
Ok, I have two question :

1-Do you think if I use a bigger bladder (latex tube) and a bigger braid, will the muscle be stronger?

2-The more I read on the subject the more I'm not sure what to use to make the muscles. Many different site tell you to take a different diameter of braided sleeving. It varies from 1/4" more than the bladder up to 1" more. I'm really not sure what to use. The one that says to use 1" sayd to cut the sleeving 1/2 inch less than the latex. I guess the sleeving just gets longer since it's a bigger diameter. What do you guys think?

Sam

On the 1st question ... the force measured in PSI (literally how many pounds per sq inch) is exerted across the area of the inner surface of the membrane. So greater surface should = greater force, however some of the force is used to expand the membrane and keep it expanded what is left over is what you can harness for you actuator. So if you use a bigger bladder that has the same wall thickness then you should be able to harness more force. However if the wall thickness increases and more force is required to expand the tube then you could theoretically lose force. Same is true with the braid if it requires more force to contract that is force taken from the actuator. Bear in mind the bigger area also requires more gas to fill that area so you will have to pump more volume.

Matt
08-06-2008, 03:15 PM
So are any of these tutorials yet? Or are one of you thinking of making one? These would be great for showing others how to do it!

sam
08-06-2008, 03:29 PM
On the 1st question ... the force measured in PSI (literally how many pounds per sq inch) is exerted across the area of the inner surface of the membrane. So greater surface should = greater force, however some of the force is used to expand the membrane and keep it expanded what is left over is what you can harness for you actuator. So if you use a bigger bladder that has the same wall thickness then you should be able to harness more force. However if the wall thickness increases and more force is required to expand the tube then you could theoretically lose force. Same is true with the braid if it requires more force to contract that is force taken from the actuator. Bear in mind the bigger area also requires more gas to fill that area so you will have to pump more volume.

I just estiamted my power for my muscle and I arrive at about 100 pounds. It seems like it loses 50% of it's power? That can't be right. :confused:

I have .612" of diameter -.125 for the latex tube = (.5"/2 * .5/2) * PI * 10 inches long. * 60 PSI.

Oh, and Matt : I will start a tutorial on it as soon as I can. I already have all the pictures. I only need to figure out how the tutorials work I guess.

ooops
08-06-2008, 04:02 PM
Can you determine at what pressure it begins it's expansion. That should give you an idea of what force is being used to expand the membrane and what should be left over.
I see your math but haven't worked through it ... gotta go ... will check it out tomorrow.

Electricity
08-06-2008, 06:33 PM
I just estiamted my power for my muscle and I arrive at about 100 pounds. It seems like it loses 50% of it's power? That can't be right. :confused:

I have .612" of diameter -.125 for the latex tube = (.5"/2 * .5/2) * PI * 10 inches long. * 60 PSI.

Oh, and Matt : I will start a tutorial on it as soon as I can. I already have all the pictures. I only need to figure out how the tutorials work I guess.

If you don't mind me using your pictures, I'd be more then willing to write the tutorial.

sam
08-06-2008, 06:59 PM
I have already started a draft tutorial on the site. You can go see it and give some comments. It's not done yet. I will clarifie some stuff. You can point out what you want clarified too because something that I think is clear might be too vague for other people. This is for evryone :wink:.

I will also put some formulas in conserning compressed air relation ships. Volume calculations for the muscles and evrything i can think of. You can also tell me what you want that I ad. :happy:

I'm going to go see office space now. Bye bye ;)

Sam

sthmck
08-07-2008, 05:53 AM
Hey the whole reason, at least in my opinion for using a bigger muscle with thicker rubber tubing is strength. Also partially durability. The first thing is that the muscle can be used with higher air pressure. The second is that the thick membrane will not wear as quickly.
Correct me if I am wrong. Are you guys trying to measure the force of the muscle by taking the size of the tube and the PSI needed to make the tube fully expand? Im not sure exactly what you were saying ooops.
I'm not sure if you can measure the strength of the muscles the way that you guys are doing it. Its a good idea but has some flaws. If you use silicone instead of latex then you are going to have different numbers. Also if there are inconsistencies in the rubber tubing then your numbers could be off. Im also not even sure when you would determine when you have reached maximum contraction. This could vary between muscles, depending on the size of the mesh. The larger diameter mesh allows the rubber to expand more therefore contract more.

A easy way to measure the strength of a muscle would be with an old fashion spring scale. I have one somewhere, I will take a pic of it to show you what I mean. All you would have to do is hook you muscle up to the scale set a regulator to X PSI open the valve let it contract and see how much it pulls. You can even measure contraction length very easily. This way you don't have to worry about any inconsistencies in muscle construction, you are just worried about overall results. I will try to find the scale and then show you what I mean this weekend.

sam
08-07-2008, 06:46 AM
Yeah. I have tried with my tougher silicone tube and I managed a 90 PSI and it was roking well. the problem it that in the application I will be doing, I need all the force I can get. So, using 90 PSI means that I will finish my air supply faster, and I don't really want that.

Yes your right. We are trying to figure out the volume (square inches) and multiply it by the pressure. This should give us the theoretical force it should have. Than you know the ACTUAL force that the muscle can produce. With that you can calculate de efficienty of the muscle. For example in a combustion engine you put in so many gallons of gaz. Each gallon has a certain number of calories (energy). But some of that energy is lost in forms of heat or friction.

Now the same thing but with a muscle. I thought I read somewhere that a certain volume of air @ a certain pressure gave you the energy. Maybe were calculating it wrong. Anyone has an idea?

Well, I using the volume I had with 25 pounds on it. Should I do it with only 5 pounds to see almost full contraction? But then, if the muscles has load on it, it will but extra pressure on the air, so the air will go in the tube and be equal verywhere (in the tank too).

Indeed it would be much easier to calculate with a spring scale than trying to hang weights on it:p

Sam

sthmck
08-07-2008, 07:06 AM
you would have to calculate the volume at full contraction. Then you could get an idea. The problem is that the volume would change depending on the size of the load. A heavier load at 90 psi may not necessarily have the same volume as one with 90 psi with a lighter load. I guess what I am trying to say is that what you are trying to do is calculate for a fixed volume type of thing. An air cylinder can be done this way because even thought the height of the cylinder changes the diameter stays the same. You might be able to get a pretty good idea of the volume but it wont be exact.

sam
08-07-2008, 08:28 AM
Yeah, it's really hard to calculate the volume, just a rough estimate would have been nice.

I put up a video of muscle being tested. The muscle is lifting 28.2 pounds and contracts 20% of it's original lengh. It is made with 1/4 ID X 3/8 OD latex tube and 3/4 inch maximal diameter of the sleeve.


http://www.youtube.com/watch?v=v_J6cwucmIA

Sam

sam
08-07-2008, 11:10 AM
Sorry for the double post but :

I was looking at some valve. I need it to be able to open and close at least 400 times per minute. Is that feasable for a valve to do this? Do you know if they have specs about the opening and closing times of the valves?

Thanks, Sam

ooops
08-07-2008, 04:26 PM
Sorry for the delay in response... the problem with the math is that you are determining volume, what we need to determine is area. The pressure is measured by the sq inch that the force is acting on.
The volume is not really important. (area in sq inch * PSI = potential force)
So, large area with low pressure is the same as a small area with high pressure.
The force is used in expanding the membrane and in moving the payload. What isn't used expanding is used moving.
If you can determine at what pressure your membrane begins it's expansion, it should give you a clue as to what amount of force is required to expand the membrane, bear in mind that the actual force will be a dynamic moving number since the membrane will increase in area as it expands therefore changing the area that the increasing pressure is acting on.
Hopefully that sheds a little more light on it.

sam
08-07-2008, 05:08 PM
Ok, I understand that it's the contact area.

I tried to do the math again, but it just seems less realistic :

.5*PI=1.57
1.57*9 inches = 14.13
14.13*60 PSI (is this how you do it?) = 847.8 pounds.

Hum... That can't be possible:confused:

Ok... I understand the dynamic numbers.

When I test for when the membrane starts expanding, do I put it at it's maximum lengh? Do I put weight on it or just let it just like that? Thanks for explaning. I see a little clearer!

I appreciate it a lot :happy:

Sam

Electricity
08-07-2008, 07:28 PM
Sorry for the double post but :

I was looking at some valve. I need it to be able to open and close at least 400 times per minute. Is that feasable for a valve to do this? Do you know if they have specs about the opening and closing times of the valves?

Thanks, Sam
I would assume that would be no problem. The 3 way valves used in paintball are capable of 100cps+ (cycles per second), I'm guessing you should be good to go.

sam
08-07-2008, 09:54 PM
Yeah, that's what I guessed. Ego's shoot like over 30 balls per second.

But it seems I can't find any specs on it. I guess I will try to contact them to find out

sam
08-08-2008, 08:00 AM
I would assume that would be no problem. The 3 way valves used in paintball are capable of 100cps+ (cycles per second), I'm guessing you should be good to go.

AHHHHH!!! I thouhgt I read 100 cycles per minute. Damn! 100 cycles per second is fast!
Do you know if they are double piloted? That's what I was wondergin, because if I use electricity to move it in on direction, but it has a spring return, will the spring be strong enough/fast eenough to bring back the valve at the other position. Or should I try double piloted? Use like a motor driver to open and the same motor driver to close the valve?

Thanks

Sam

ooops
08-08-2008, 09:34 AM
Ok, I understand that it's the contact area.

I tried to do the math again, but it just seems less realistic :

.5*PI=1.57
1.57*9 inches = 14.13
14.13*60 PSI (is this how you do it?) = 847.8 pounds.

Hum... That can't be possible:confused:

Ok... I understand the dynamic numbers.

When I test for when the membrane starts expanding, do I put it at it's maximum lengh? Do I put weight on it or just let it just like that? Thanks for explaning. I see a little clearer!

I appreciate it a lot :happy:

Sam

That is the right math, and that is the right force. The tube itself is useing a great deal of force in expanding. What is left over is the force available to you.

Electricity
08-08-2008, 11:43 AM
AHHHHH!!! I thouhgt I read 100 cycles per minute. Damn! 100 cycles per second is fast!
Do you know if they are double piloted? That's what I was wondergin, because if I use electricity to move it in on direction, but it has a spring return, will the spring be strong enough/fast eenough to bring back the valve at the other position. Or should I try double piloted? Use like a motor driver to open and the same motor driver to close the valve?

Thanks

Sam
Heres a page describing all the various 'noids used in paintballage.
http://www.zdspb.com/tech/misc/solenoids_models.html
And the design of the various types.
http://www.zdspb.com/tech/misc/solenoids_design.html
It looks like the high end ones do use a spring return, so yeah, I'd guess springs are fast enough.. :)

sam
08-08-2008, 05:30 PM
That is the right math, and that is the right force. The tube itself is useing a great deal of force in expanding. What is left over is the force available to you.

Ok, so I will do some tests tomorow to see how much PSI it needs to start contracting.

@ Electricity : Ok, I will go tru all of the article a bit later, thanks for he information! Just adding a 2 way electricity instead of a string return cost 15 dollars more.

Sam

sam
08-11-2008, 10:34 PM
Hey guys,

Quick question here for those of you who know a lot in these waters : :veryhappy:

When it says 6.7SCFM @ 100 PSIg, is it proportional. Let's say I have 60 PSI, can I do 6.7/100 *60 and get SCFM when it' at 60 PSI?

Thanks a lot! :tongue:

MYKL
08-12-2008, 08:46 AM
Sam,

Is your plan to use valve regulated bursts of air to meter the muscles movment?
Like: unloaded it consistantly (theoretically) takes 100 bursts @ 100 PSI to move a muscle enough for the limb to move a eighth of an inch? Thats what I'm picturing in my head with what you are asking and purchasing here. Then you have a numerical value. You could use pressure sensors to adapt the burst count to adjust for increasing/decreasing loads.

With the air have you tried preloading the muscles at a higher pressure, keeping them right at the cusp of the amount it takes to initiate flexation? That way you don't waste time and energy building up pressure for each cycle.

I really hope you can make something happen for yourself with your setup. It is educational for us all.

^_^

sam
08-15-2008, 11:53 AM
Well, my plan isn't exacly that. But once I have all my materials, I might try something like this in the future. It would be interesting just to make an arm and try to put 5 pounds and it would sense the movment and try to keep it's original position. then add another 5 pounds, then it would contract more again.

I have ordered 3 3-way valves from clippard a couple days back. They should be here soon. I have also placed my order at McMaster, but they sent me an E-Mail telling me that they don't ship to Canada anymore since it's too complexe (and yet they still have a section where you can choose from a list of contries where Canada is in there... They should change that ASAP, very frustrating. I spent a LOT of time finding all the stuff/subtitutes of stuff I need on their site, I wouldn't have lost my time if I had known. And they told me they weren't processing my order after they took the credit card number... :sad::genmad:)

So I will buy the stuff, send it to my dad (in the USA) Then he will have to send them back here... So this puts my project on hold.

sam
08-15-2008, 03:36 PM
Now, I should receive my valves anyday now. I'm wondering how I can control them? They are 12 volts. Could I control them via an ABB with a Basic atom Pro? I don't know if the board can give 12 volts, so I might have to do a relay here? Does anyone know how I could program it to apply current? I'm really lost in this section. :confused:

Any help is appreciated :tongue:

Sam:veryhappy:

Adrenalynn
08-15-2008, 03:50 PM
You can't. The current draw would toast any microcontroller.

What you need is a serial relay board. I've built a handful of pretty large ones lately for controlling some really big motor controllers.

These will work: http://www.trossenrobotics.com/store/p/3203-InterfaceKit-0-0-4.aspx if you're not drawing more than 5A. They're kinda a pita if you're not going to have a PC connected because they're USB.

If you're needing some larger or with a lot more outputs, let me know. . Mine are also TTL serial if that's helpful - no PC requirement and further run distances. I have a few extras assuming I don't blow any up and Ooops doesn't blow any up. ;)

sam
08-16-2008, 08:22 AM
I don't even think I will ever go over 1 amp.


Ok, so I could control it with my computer. Let's say I wanted to controll the output of this relay with a PS2 controller or a potentiometer, I would need an analog input for the pot and a digital for the PS2 controller?

So I would take the 8/8/8 Phiget?

I could all control this from let's say a C/C++ program that I'm running on windows. Or, could I instead of using Digital or analog inputs, make a GUI on my computer and control the outputs by that?

WIch would be more simple for a beginner?

Adrenalynn
08-16-2008, 11:07 AM
A relay is on and off. Are you looking at a threshold on the pot?

Yes, you could write some code with a gui and control 'em that way. Mine, at least, are dirt simple serial bits. I haven't looked at the programming datasheet on the phidgets yet, but will if you'd like.

sam
08-16-2008, 02:56 PM
Well, I was looking at controlling the speed of the cycles of the solenoids with the imput of a GUI or the imput of a pot (the higher the voltage, the faster it opens and close).

ScuD
08-16-2008, 04:17 PM
You may want to scratch relays if you're going to do lots of switching. Especially when driving inductive loads (eg. the solenoids, massively inductive) you'll be pulling sparks on the relay contacts, wearing them out in no time.

You'll want to use either solid-state relays ($) or look into switching them with transistors / fets / optocouplers.

I may have missed the link on the solenoids, I'll check the previous posts and see what I can come up with.

/edit: ok I didn't find a direct link to the ones you've bought, but I found some clippard solenoid valves and they all had 3W or lower ratings.

So I made a lil schematic on what you could try:
http://users.pandora.be/svendecock/robot/parallel.PNG
Couldn't find a "real" parallel port connector immediately in the libraries; anyway... it's a 25-pin connector, pinout can be found here (http://www.dpie.com/pcmcia/spp100_pinout.gif).
Pins 2 to 9 are the databits, the ones you can set (and in some cases, read) using your own software.
Pins 18 through 25 are ground, as is the shroud.

I took a BC116 as the transistor, they can sink 600mA continuous, more than plenty for the 3W solenoids. Don't forget the diodes! Any standard diode will do, just make sure they're the right way around.

With 8 transistors, 8 diodes, a couple of connectors and some protoboard you can roll.


one more /edit:
If you want serial output, it's possible, but would be somewhat more complicated. I could whip something up if you'd like though.

sam
08-16-2008, 05:57 PM
Ok, I kinda understand the diagram. I have a few other questions now.

I assume the X's are the + and - sides of one solenoid?

So when the transistor open, the valve has current? And the diode obligates the current to pass thru the valve instead of beside?

Ok, i'm starting to get the picture. I will read a bit more on how I could control it with a prallele port. Any good sites someone knows taht has information that vagely concerns that? :happy:

Thanks a lot for the diagram, it help clear my mind a bit!

EDIT : In case of future reference, the model of valve I have are these : http://www.clippard.com/store/display_details.asp?sku=MME-3QDS-W012

ScuD
08-17-2008, 04:12 AM
The X'es are indeed the connectors for the solenoids. The diodes serve as flyback diodes, they are meant to reduce induction currents so the transistors don't get damaged.
In fact, they serve the same purpose as the diodes on the H-bridge schematics i've posted in the tutorial draft.

sam
08-17-2008, 11:27 AM
So, let's say I could use a Basic Stamp 2 to control the valves? I would use the same set up that scud did. I would send one pin on high to open one of the valves. It wouldn't put any pressure on pin or microcontroller, right?

I would only have to program it once and then I could bring it where I want and it would also work there. I will check up how I can connect the BS2 to my PC.

ScuD
08-17-2008, 12:02 PM
Yep, just make sure to use a transistor and diode as in the schematic, you can hook it up to the Stamp perfectly.

sam
08-17-2008, 12:54 PM
Ok, I have no idea how that I can connect my Basic stamp to my computer using a prallel port (I don't have a RS232 port, and my USB to serial converter does not work any more).

I have been looking on google to find some kind of site to help me out. So far no luck. :sad:

If any one could helpl me on this it would be much appreciated! :veryhappy:

thanks,
Sam

EDIT : Ok, after finding my long lost BS manual, I see that Pin 1 (SOUT) should be connected to pin 3 on DB25 (serial connector with 25 pin, right?)
-Pin 2 of the BS2 (SIN) should be connected to pin 2 of DB25
-Pin 3 (ATN) is connected to pin DB 20.
-Pin 4 (VSS) connected to DB7

So I just need to connect those 4 pins and I can program it via my PC?

I will go buy some connectors to connect it to my PC.

4mem8
08-17-2008, 01:54 PM
It may be safer to buy a serial cable or USB depending on what board you have for your BS2.Or are you trying to connect it directly to your BS2?? I hope you have a board, It's a lot safer for you. Voltage regulation and all that.

sam
08-17-2008, 02:07 PM
Well, I was think I'd build a small regulated 5V current. Install and solder a 25 pin connector to the board (or soderless breadboard). Connect the 4 pins that I need connected. I think that's it, I have a connection right? (well I would connect it with a 25 pin cable).

Should buy a board for the BS2 and then work from there? Is what I'm trying to do dangerous?

Advice? Warnings? :cool:

Sam

Adrenalynn
08-17-2008, 02:15 PM
I thought you didn't have a serial port on your PC? How does that help you?

The pin-out that ScuD gave you is PARALLEL, not serial. Don't confuse them or not-so-fun things will happen.

sam
08-17-2008, 02:19 PM
Ok, sorry.

I have a 25 pin connector (I assume it's parallel port, right?), I have no serial 9 pin connectors.

How do I know the difference between a parallel and a serial? Parallel have 25 pins and serial 9?

Adrenalynn
08-17-2008, 02:23 PM
Old serial ports had 25 pins. The main difference on a PC is that a serial port is male and a parallel port is female on the PC-side. [NOT universal on non-PC components!]

I can practically guarantee you don't have a 25pin serial port unless you knowingly put it in there (or are running a 386 or older.. ;) )

ScuD's circuit was meant to turn your parallel port into a simple Digital to Analog device and control your device from the PC, not from the BS. You'll likely be using digital out pins on the BS to interface to it, NOT its serial port (unless you want to blow it or the circuit up. ;) )

sam
08-17-2008, 02:36 PM
Ah ok, so I have a parallel port then :tongue:

What I want is just to connect by BS2 to my computer to program it. Since I don't have a serial port or a USB to serial calbe, I want to use the parallel port (wich, correct me if I'm wrong, is possible). In my Basic stamp syntax and reference manual they say that the first 4 pins are used to connect to a parallel port for programming. Then I would use somewhat the same setting as scud showed to control the solenoids with a BS2 (just use high/low settings for on and off).

Could I be able to do this from scratch/worth the trouble?

Sam

Adrenalynn
08-17-2008, 03:04 PM
Could you post where it says it's programmed over the parallel port? The names of the pins are serial (SIN = Serial In, SOUT = Serial out, etc.)

sam
08-17-2008, 03:31 PM
Yeah, I'm starting to dought that. The only thing I read is : Pin 1 : SOUT. Serial out : connects to PC serial port RX pin (DB9 pin 2 / DB25 pin 3) for programing.


I Guess it's for the old 25 pin serial ports :sad:

Maybe I will have to find another way.

Adrenalynn
08-17-2008, 04:04 PM
The "other way" to program it is hooking it up to a serial port. Or, in a pinch, you could hook it up to a serial port. I might also try hooking it up to a serial port. ;)

The SOUT on the BASIC Stamp goes to the Serial Data In (Pin 2 on a DB9) on the PC, the SIN on the BASIC stamp goes to the Serial Data Out (Pin 3 on a DB9) on the PC. The latter is the important pin for programming the Stamp.

MYKL
08-18-2008, 09:04 AM
Do any of you read Popular Science?

The newest has an article about a 'Motorcycle that you wear'. It is just an idea that a young designer has but it incoroprates the 'Pnumatic Muscle' actuators...

http://www.popsci.com/cars/article/2008-05/motorcycle-you-can-wear

I have dreamt of something like this but with a hybrid powersource, hydraulic muscles and four wheels.

^_^

sam
08-19-2008, 07:49 AM
The "other way" to program it is hooking it up to a serial port. Or, in a pinch, you could hook it up to a serial port. I might also try hooking it up to a serial port. ;)

The SOUT on the BASIC Stamp goes to the Serial Data In (Pin 2 on a DB9) on the PC, the SIN on the BASIC stamp goes to the Serial Data Out (Pin 3 on a DB9) on the PC. The latter is the important pin for programming the Stamp.

Oh! I could use a serial port! :cool: Ok, I will try another time with a home built board to connect to my computer. So, I will need to build a current regulator (with a simple resistor and zener diode should do?). Buy some NPN transsitors with enough current to use the solenoids. I will incorporate a LED in this to tell me when it's on and off.


Do any of you read Popular Science?

The newest has an article about a 'Motorcycle that you wear'. It is just an idea that a young designer has but it incoroprates the 'Pnumatic Muscle' actuators...

http://www.popsci.com/cars/article/2008-05/motorcycle-you-can-wear

I have dreamt of something like this but with a hybrid powersource, hydraulic muscles and four wheels.

^_^

Well, I don't :sad: .

It's weird, they say it has 36 penumatic muscles, but they don't specifie what they do. I thought they ran the car, but they say it has a battery.

sthmck
08-19-2008, 07:54 AM
My guess is that they have something to do with controlling the position of the two outside wheel arm things. Not sure what to call them. At least that is what it looks like in the pics. It kind of a cool idea but I think it will be about as useful as a segway. It is cool that air muscles are used in it though.

MYKL
08-19-2008, 08:47 AM
Well, I don't :sad: .

It's weird, they say it has 36 penumatic muscles, but they don't specifie what they do. I thought they ran the car, but they say it has a battery.

I have the printed article. It show more detailed renderings of the unit. It appears that the designer/dreamer uses the muscles like sthmck and you guessed. They are used to manipulate the 'arms', the helmet and the spine. The wheels are supposedly driven by speacial Lithium Batts and Capacitors. I like the idea of useing capacitors to deliver boosts of power. I don't know that I'd drive it but it's cool. I still think things like this would be more feasable with a non-compressing agent at work. The wife is gone for two weeks and her sister has moved out so I should have time to experiment a bit this week or next. I hope to join you with some pictures and practiced thoughts soon. Keep it up! When you and Adrenalynn get those Air Muscles under control I think it'll be a thing of awe to watch...

^_^

sam
08-19-2008, 09:46 AM
On the other hand, if you have a compressed gas in it, if (read when) you crash, it will give a *little* spring motion to slow you down. I think something like water wouldn't move at all since it's non compressable. Unless you have a sensor and start pumping the water out :tongue:.

I'll whip up the circuit ths afternoon and see if I can connect it with my USB to serial converter (cross your fingers :wink:). I'll take picture and post them here as well as results!

sthmck
08-19-2008, 10:16 AM
On the other hand, if you have a compressed gas in it, if (read when) you crash, it will give a *little* spring motion to slow you down. I think something like water wouldn't move at all since it's non compressable. Unless you have a sensor and start pumping the water out :tongue:.


Nah it should still move. Since the tube is rubber it will still flex. The only way it wouldn't move that much is if the muscle had fully contracted.

sam
08-19-2008, 10:38 AM
I thought that let's say the pump just filled the muscles with water and you try to pull it when the pump does not move, the water won't compresse and wont be able to get out (since the pump is stopped). Will the water still manage to get out?

Well, good news for me : My dad just received the package in California! So, very hopefully, I should receive my package on friday. That would be great! I still have no idea for the 3 way valves. They didn't tell me a tracking number or anything... I will send them an E-Mail I guess.

I'm having lunch and then going to the electric store (wich is really far in bike :genmad:)

MYKL
08-19-2008, 12:06 PM
On the other hand, if you have a compressed gas in it, if (read when) you crash, it will give a *little* spring motion to slow you down. I think something like water wouldn't move at all since it's non compressable. Unless you have a sensor and start pumping the water out :tongue:.

I'll whip up the circuit ths afternoon and see if I can connect it with my USB to serial converter (cross your fingers :wink:). I'll take picture and post them here as well as results!

The fluid is non-compressable but the muscles are very compliant. I've found that you don't want to push these things all the way to capacity with fluid because they will tear apart. The muscles are awesome shock absorbers with the fluid inside and move with exponentially more force than with air. (What that exponant is is beyond me.)


I thought that let's say the pump just filled the muscles with water and you try to pull it when the pump does not move, the water won't compresse and wont be able to get out (since the pump is stopped). Will the water still manage to get out?

Well, good news for me : My dad just received the package in California! So, very hopefully, I should receive my package on friday. That would be great! I still have no idea for the 3 way valves. They didn't tell me a tracking number or anything... I will send them an E-Mail I guess.

I'm having lunch and then going to the electric store (wich is really far in bike :genmad:)

If the muscle is at the bursting point when you shock it the fluid will make its way out. ^_^

No one there to bum a ride from? It would suck to break your new toys (or yourself) whilst toteing all that stuff home...

Adrenalynn
08-19-2008, 02:23 PM
I think something like water wouldn't move at all since it's non compressable.

The physics geek in me has to point out that, thankfully, water is compressible. If it weren't, there wouldn't be a California or the East Coast. In fact, we'd lose more than 2 million square miles...

Water has a compressibility factor of around 3.4 x 10^-5 at room temps: 1 ie. ~1psi will reduce the unit volume by 0.0000034 of the original volume

sam
08-19-2008, 02:42 PM
You have good points everyone.

So!! I have good news! I just received my valves!! :tongue: I took the time to take pictures for you guys. They are with my WebCam, but it's the only camera I have, so I don't have much choice.

Here is the package :

http://forums.trossenrobotics.com/gallery/files/1/5/4/9/picture_354_thumb.jpg (http://forums.trossenrobotics.com/gallery/showimage.php?i=670&c=3)

Here is the valve :

http://forums.trossenrobotics.com/gallery/files/1/5/4/9/picture_355_thumb.jpg (http://forums.trossenrobotics.com/gallery/showimage.php?i=671&c=3)

Here is the connection to my air tank :

http://forums.trossenrobotics.com/gallery/files/1/5/4/9/picture_357_thumb.jpg (http://forums.trossenrobotics.com/gallery/showimage.php?i=672&c=3)

Here is the exaust :

http://forums.trossenrobotics.com/gallery/files/1/5/4/9/picture_358_thumb.jpg (http://forums.trossenrobotics.com/gallery/showimage.php?i=673&c=3)

Here is the connection to my muscle :

http://forums.trossenrobotics.com/gallery/files/1/5/4/9/picture_359_thumb.jpg (http://forums.trossenrobotics.com/gallery/showimage.php?i=674&c=3)

Well, I'm going to go see tropic thunder!!

I had a little trouble, I will explain later.

MYKL
08-20-2008, 08:43 AM
Those look fairly small compared to what (I hope) are your finger tips. What is the envelope on those? Are you going to build a single header for all of these valves?

What about propylene glycol Adrenelyn?

I am planing to use it for one of its already accepted applications:
Propylene glycol is used:

As a moisturizer (http://en.wikipedia.org/wiki/Moisturizer) in medicines (http://en.wikipedia.org/wiki/Medication), cosmetics (http://en.wikipedia.org/wiki/Cosmetics), food (http://en.wikipedia.org/wiki/Food), toothpaste (http://en.wikipedia.org/wiki/Toothpaste), mouth wash (http://en.wikipedia.org/wiki/Mouth_wash), and tobacco (http://en.wikipedia.org/wiki/Tobacco) products
In electronic cigarettes (http://en.wikipedia.org/wiki/E-cigarette) to make the produced vapor better resemble cigarette smoke
As a medical and sexual lubricant (http://en.wikipedia.org/wiki/Lubricant) (A.K.A. "personal lubricant (http://en.wikipedia.org/wiki/Personal_lubricant)")
As an emulsification (http://en.wikipedia.org/wiki/Emulsification) agent in Angostura (http://en.wikipedia.org/wiki/Angostura_bitters) and orange bitters (http://en.wikipedia.org/wiki/Bitters)
As a solvent (http://en.wikipedia.org/wiki/Solvent) for food colors (http://en.wikipedia.org/wiki/Food_coloring) and flavorings (http://en.wikipedia.org/wiki/Flavoring)
As a humectant (http://en.wikipedia.org/wiki/Humectant) food additive (http://en.wikipedia.org/wiki/Food_additive), labeled as E number (http://en.wikipedia.org/wiki/E_number) E1520
As a carrier in fragrance oils (http://en.wikipedia.org/wiki/Fragrance_oil)
As a less-toxic (http://en.wikipedia.org/wiki/Poison) antifreeze (http://en.wikipedia.org/wiki/Antifreeze)
As a solvent (http://en.wikipedia.org/wiki/Solvent) used in mixing photographic chemicals, such as film developers
In smoke machines (http://en.wikipedia.org/wiki/Smoke_machine) to make artificial smoke (http://en.wikipedia.org/wiki/Smoke) for use in firefighters (http://en.wikipedia.org/wiki/Firefighter)' training and theatrical (http://en.wikipedia.org/wiki/Theater) productions
In hand sanitizers, antibacterial (http://en.wikipedia.org/wiki/Antiseptic) lotions, and saline solutions (http://en.wikipedia.org/wiki/Saline_solution)
In cryonics (http://en.wikipedia.org/wiki/Cryonics)
As a working fluid in hydraulic presses
As a coolant (http://en.wikipedia.org/wiki/Coolant) in liquid cooling systems
To regulate humidity in a cigar humidor (http://en.wikipedia.org/wiki/Humidor)
As the killing and preserving agent in pitfall traps, usually used to capture ground beetles (http://en.wikipedia.org/wiki/Ground_beetles)
To treat livestock ketosis (http://en.wikipedia.org/wiki/Ketosis)
As the main ingredient in deodorant (http://en.wikipedia.org/wiki/Deodorant) sticks.
(From Wikipedia) Hey! in this case, One can use the bots blood to make smoke too! ^_^

I guess what you mean about compressing water is that the weight of the water compresses a large volume of it into the trenches... Right?

sam
08-20-2008, 08:51 AM
I'm sooooo happy!! Filled with joy!:veryhappy:

I did the circuit you guys showed me on a breadboard and I connected it to my computer and my USB to serial adapter worked!

I just finished intalling and soldering my BS2 board. Connected it to the computer and programed the valves! It was so easy! Just saying HIGH 7, Pause 250, LOW 7. I just used the simple circuit Scud showed me. Had a little trouble but it was solved very fast (I was putting the pin number instead of the I/O number, Pin 5 = 0 when you write the code).

I tried up to 50 ms, woah those valves are fast! :eek:

I have a couple pictures for you guys :

Here is my setup to control the valves with the basic stamp 2, 3 diodes, 3 transistors and the serial port :

http://forums.trossenrobotics.com/gallery/files/1/5/4/9/picture_362_thumb.jpg (http://forums.trossenrobotics.com/gallery/showimage.php?i=676&c=3)

Here is a close up :

http://forums.trossenrobotics.com/gallery/files/1/5/4/9/picture_363_thumb.jpg (http://forums.trossenrobotics.com/gallery/showimage.php?i=677&c=3)

Clsoe up with valve attached :

http://forums.trossenrobotics.com/gallery/files/1/5/4/9/picture_364_thumb.jpg (http://forums.trossenrobotics.com/gallery/showimage.php?i=678&c=3)

Under the board (sorry for the fuzz) :

http://forums.trossenrobotics.com/gallery/files/1/5/4/9/picture_365_thumb.jpg (http://forums.trossenrobotics.com/gallery/showimage.php?i=679&c=3)

Well, these are all the photos! It worked, I'm at awe. Each time I did this kind of thing with microcontrollers, it didn't work or I spent hours troubleshooting. This one work very well.

I really want to thank evryone that helped me! :veryhappy: You were really nice. Maybe I will do a tutorial for the basic stmap if you guys want me too, just get the info in one post will be usefull.

Thanks

Sam

sam
08-20-2008, 09:00 AM
Those look fairly small compared to what (I hope) are your finger tips. What is the envelope on those? Are you going to build a single header for all of these valves?



I guess what you mean about compressing water is that the weight of the water compresses a large volume of it into the trenches... Right?

Yeah, I was suprised too, there pretty small. But as along as they can pass the 6.7 SCFM they promised, it's fine :wink::tongue: Simgle header? I don't understand.

I assumed too that it was because the weight of the water was pressing on the water below it.

Oh, and what is propylene glycol, is it to reaplace the water?

MYKL
08-20-2008, 12:14 PM
As a working fluid in hydraulic presses (and/or muscles). Yup Yup

A single header to direct the air flow from each valve to the appropriate area of the bot. It could serve as a structural element for the bot as well.

^_^

Adrenalynn
08-20-2008, 01:07 PM
Those look fairly small compared to what (I hope) are your finger tips. What is the envelope on those? Are you going to build a single header for all of these valves?

What about propylene glycol Adrenelyn?

I guess what you mean about compressing water is that the weight of the water compresses a large volume of it into the trenches... Right?

I didn't find any quick numbers on the compressibility of propylene glycol, only methods of making it more compressible. ;)

Not just compressing a large volume of it in the trenches, but over the entire ocean. The more pressure, the more it compresses, of course. The weight of water on water is what it is, and that grows as more of it is there.

The upshot is that if water weren't compressible, the entire ocean level would be 30m+ higher.

Just my cheapie little press in the garage should be able to compress water by ~0.07%.

For the application you guys are doing, it shouldn't matter - you're talking hundred PSI not tens of thousands of PSI. It's just a technical gotcha, one of those ones the physics geek in me feels compelled to comment on. ;)

MYKL
08-20-2008, 01:38 PM
As long as you continue proving that you truly are a geek and not a dork... Nerd even, but the world is too glutted with dorks.

^_^

ScuD
08-20-2008, 01:40 PM
Great to hear it works fine Sam!

Nice looking solenoids, too, seems like you'll get a nice system built up.
Looking forward to seeing more of it!


I have a few solenoid valves laying around, as well as about 7 small (read 1 to 2cm stroke) pistons in various versions (double piston, single piston, spring return etc) but up to this point I haven't come up with any project for them yet.

ooops
08-20-2008, 01:53 PM
Congratulations!!!
It is always a great feeling when everything works like you hope!

MYKL
08-20-2008, 02:36 PM
Festo has a few assemblies using air actuated muscles.

I recently found this video:

YouTube - better robot muscles

Awesome...

^_^

sam
08-21-2008, 10:08 AM
Yeah, I saw that video a couple weeks ago. It's very impressive to see that they manage to do pretty precise movement with air muscles.

sam
08-23-2008, 10:17 AM
UPDATE :

Ok, Well everything is working as planed but I have a problem. I want to control a variable in the program with an annalog input (a potentiometer). Well, waht I wawnt to control is the pause time.

So I could write PAUSE X.

Then, could I define the x with the input of the potentiometer? Do any of your guys know how I could do this?

Thanks a lot,

Sam

Adrenalynn
08-23-2008, 10:58 AM
What microcontroller are you using?

Sounds similar to my servo controller. I control servo position with a microcontroller. I take a pot in, convert the analog to a serial command for a servo controller, smooth the output, then ship it over wireless to the servo controller.

sam
08-23-2008, 12:01 PM
I am using the basic stamp 2.

I don't know how to use analog input (in the code) to change a variable. And how do I put limits on the variation of the output? Like the speed of pause.

sam
08-25-2008, 02:25 PM
WOOHOOT!

I just managed to control the time variable with the potentiometer using the RCTIME statment! I'm on fire! Well, I'm pretty happy. all the electronics have been done to control the valves :happy:

See ya.

ooops
08-25-2008, 02:32 PM
WOOHOOT!

I just managed to control the time variable with the potentiometer using the RCTIME statment! I'm on fire! Well, I'm pretty happy. all the electronics have been done to control the valves :happy:

See ya.

Sam, great news Congratulations!!!
Video to follow??????????

Adrenalynn
08-25-2008, 02:37 PM
I had a post in another browser window documenting RCTIME and POT commands. I'm sorry - I have about 50 browser windows open (literally), my apologies that one snuck past... :(

Grats!

sthmck
08-25-2008, 04:01 PM
Great work. Are you actually rotating a joint right now?

sam
08-25-2008, 04:20 PM
Well, not yet, I'm just controlling the air valves with no air passing thru them. You can hear the thunk when it opens/closes. :happy: I still haven't received my stuff from McMaster. Maybe it got stuck in the customs. Who knows. :genmad:

After reading my BS2 manual and a lot of sites I managed to understand and after a few tries made it work. Thanks anyway Adrenalynn :veryhappy:.

Well, I would do a video, but I don't have any sound and as I mantionned, it's the only thing it does now :tongue:. I could show you the final board with the electric tape on it :tongue:

sthmck
08-25-2008, 04:32 PM
You could do captions lol. It could go like "click, now the valve is open" "click now the valve is closed" It would be a sensation in the hobbyist community. It will probably make hundreds at the box office.

MYKL
08-26-2008, 08:36 AM
Congoratz on the control sir.

How many cycles per second can you glean from your new nervous system?

sam
08-26-2008, 02:03 PM
I don't know the exact time since it's just a variable, but feeling with my hands tell me it's at least 10 cycles per second.

I feel captions won't do the trick. anyway my computer is too old to go in windows movie maker. :wink:

sthmck
08-26-2008, 09:48 PM
Any idea how much air flows through the valve per cycle? If so could you use that to control the length of the contraction like the steps in a stepper motor sort of?

sam
08-27-2008, 06:36 AM
Yeah, I don't know that. I could probably estimate a number. They say that the valve can pass 6.7 [email protected] PSI. but alas I don't know how much time it takes to fully open, I think the response time is 20 ms. Does that mean it takes 20 ms from the time it receives to signal to the time it's fully open?

I'm not even sure for the 20 ms.

sthmck
08-27-2008, 07:16 AM
I think you are right about the response time of 20 ms being the time it takes to fully open the valve.

sam
08-28-2008, 04:34 PM
Hello to all!

I have just received my pieces from taht my father sent me. I did my first muscle of too (next one coming tonight). I manage to pull off a wooping 31% contraction with 28.3 pounds attached to it.:eek: Much better than the measly 18-20% I got with my other muscles! The tests were with 60 PSI of course. Not bad, Ièm getting in the high percentages. Also, the muscle looks really great when it contracts, all nice a buldging! :veryhappy:

I will test the second muscle this evening. I tried to take pictures but my computer freezed and I had to restart it 2 times because of the damn Logitech program that's too hard for my computer :sad: That's sad, pore computer...

Sam

BADfish10
08-29-2008, 05:59 AM
Hello to all!

I have just received my pieces from taht my father sent me. I did my first muscle of too (next one coming tonight). I manage to pull off a wooping 31% contraction with 28.3 pounds attached to it.:eek: Much better than the measly 18-20% I got with my other muscles! The tests were with 60 PSI of course. Not bad, Ièm getting in the high percentages. Also, the muscle looks really great when it contracts, all nice a buldging! :veryhappy:

I will test the second muscle this evening. I tried to take pictures but my computer freezed and I had to restart it 2 times because of the damn Logitech program that's too hard for my computer :sad: That's sad, pore computer...

Sam

Sounds great

Cant wait to see them in action
good numbers you are getting with your setup!

J

MYKL
08-29-2008, 08:41 AM
Very encouraging SAM.

Set up a joint and knock our socks off! What colour of sleveing did you go with? I know McMaster has a variety to choose from. I live about 15 miles form a McMaster so I can will call. Its a great place to get prototype stuff. You can get almost everything they have from direct sources (save some $) once you are happy with your project and want to start making anything in numbers.

^_^

sam
08-29-2008, 09:14 AM
Yep, they have practicly evrything one would need in terms of mechanical stuff. Too bad they don't ship to Canada, it's a real bummer. :genmad: Oh, I just took old plain black sleeving. I find it clasic, maches with evrything :wink:

I will try to post a video of it when I have time. Might take a while.

So I attached the valve and made one muscle work. It goes pretty good actualy. It takes longer to empty the muscle than to fill it, so the muscle is fully contracted after 5 shots and won't decontract. But if I have a muscle on the other side that pull too, it will empty much faster I think. So I will try to set this up today. It's one of my last days without work.


Build FRENZY!

Sam

sthmck
08-29-2008, 12:25 PM
clippard has exhaust valves that you could attach right to the muscle. They activate when there is a preassure differintial greater than 5 psi. That should make your muscles deflate much faster.

sthmck
08-29-2008, 01:54 PM
By pressure differential I mean that when the regular valve starts to vent air it creates a pressure differential between the air line and the muscle. The exhaust valve then opens and lets all the air out really fast.

MYKL
09-02-2008, 08:59 AM
I think what Sam may be refering to is the fact that these 'muscles' tend to retain some of thier contracted shape 'till a load pulls them into their preferrred stretched out position. The braided sleaving and the latex core will pull a bit on its own unless a load stretches it to full extention. This is where you want these actuators to be because this is where the most powerful part of their stroke usually is. So the opposing/antagonistic approach is the most mechanically efficient set up for these actuators. You could use a spring loaded approach but then the actuator looses the amount of power it needs to expend to overcome the spring AND move the load.

sthmck
09-02-2008, 09:21 AM
yeah i see what you mean. I thought he was talking more about air still being in the muscle. I haven't started building my arm yet but my plan is to design it so that I can vary the initial length of the muscle. If one side is at full contraction and the other side is not I want to be able to adjust the muscle so when the opposing muscle starts to contract it will be able to take advantage of the muscle properties that you just mentioned.

MYKL
09-02-2008, 03:22 PM
That sounds perfect. When you get the controls down and write some code for it maybe I can get your help with controling the liquid version...

^_^

sam
09-02-2008, 03:50 PM
Yeah, that's basicly it. I think I will just use the power of the other muscle to pull the first one.

But I have one question for sthmck, what do you mean by "I haven't started building my arm yet but my plan is to design it so that I can vary the initial length of the muscle. If one side is at full contraction and the other side is not I want to be able to adjust the muscle so when the opposing muscle starts to contract it will be able to take advantage of the muscle properties that you just mentioned."

What can you take advantage of? How?

I'll proobably buy a strech sensor on my next oder.

ooops
09-03-2008, 07:55 AM
Wow, Sam congratulations on your success so far!!!
I am looking forward to pictures/video when you get a chance:)

sthmck
09-03-2008, 12:54 PM
Yeah, that's basicly it. I think I will just use the power of the other muscle to pull the first one.

But I have one question for sthmck, what do you mean by "I haven't started building my arm yet but my plan is to design it so that I can vary the initial length of the muscle. If one side is at full contraction and the other side is not I want to be able to adjust the muscle so when the opposing muscle starts to contract it will be able to take advantage of the muscle properties that you just mentioned."

What can you take advantage of? How?

I'll probably buy a strech sensor on my next oder.

Ok so my statement about making the intial length of the muscle variable is not correct. What I should have said is that I plan in making the point at which the muscles attaches to the arm variable.

sam
09-03-2008, 04:33 PM
Wow, Sam congratulations on your success so far!!!
I am looking forward to pictures/video when you get a chance:)

Thanks a lot! I'm really happy with how they work. If anyone has questions I would be happy to explain further.

@sthmck : Ok, so the attach to the muscle can be longer (so th elatex doesn't change) or shorter. But why would you need it to change?

To keep the muscle fully extented even if the arm is lower or higher?

sthmck
09-03-2008, 04:56 PM
Yeah it kinda depends on how the joint position is when it starts out. We called call it a home position. For example if the one muscle could be attached like it was in a fully contracted position, while the other one is in an extended position. If one muscle is a little longer than the other then all I have to do is move the attachment point. Also I could pretension a muscle before any air is put into the system.