PDA

View Full Version : [Question(s)] Arms: SG5-UT vs SES ?



Adrenalynn
04-10-2008, 02:51 AM
Since Trossen sells the bits for constructing an equivalent to the SES arm, and the kit for the SG5, I feel fairly safe asking...

Any thoughts on one vs the other? I suspect the SG5 gripper is much superior, but I do, in theory, like the design of the SES arm. I like how trivial it would be to change the lengths of the segments, and even number there-of. I also think I like the counter-balancing springs.

I'm being pretty indecisive, so any input would be most welcome!

LinuxGuy
04-10-2008, 05:57 AM
Since Trossen sells the bits for constructing an equivalent to the SES arm, and the kit for the SG5, I feel fairly safe asking...
I don't think you have to worry about talking about products here that Trossen Robotics does not sell - it could indeed end up with them carrying a new product.


Any thoughts on one vs the other? I suspect the SG5 gripper is much superior, but I do, in theory, like the design of the SES arm. I like how trivial it would be to change the lengths of the segments, and even number there-of. I also think I like the counter-balancing springs.
It doesn't look like you can expand or build on the SG5 arm, as far as I can tell. However, you can definitely build on and change an SES arm any way you want to. :happy: You can construct an SES based arm to be exactly what you want it to be. My own feelings on this are to build the sort of SES arm you want with the kind of servos you want according to your application. You can use the tubing to increase the reach of an SES arm.

Of course, you can also change the servos in the SG5 arm to increase torque and holding power, but as far as I can tell, that is the only kind of change you can make to it. I would go with the flexibility of the SES and get exactly the arm you want.

Yes, without a doubt, I am completely in favor of going with the SES over any other solution I have seen to date. The only other thing I might consider for an arm would be the Bioloid kit. I know for a fact that the SES brackets and components are very strong and have been using the SES exclusively for almost two years now. :happy:

8-Dale

LinuxGuy
04-10-2008, 06:07 AM
One last note on this. I have designed a gripper using the SES that has 5DOF of flexibility. I just need to get some dimensions on some of the other Lynxmotion components (like robot hands, Little Grip, etc) so I can complete my 3D model of the new gripper. I don't think you will find a more flexible gripper that can be reconfigured any way you want at any time. This new gripper I designed makes use of the double ASB-04 joint I designed for an SES biped leg, based on what has been done by Matt Bauer on his Robonova double knee joint. I also have an SES leg design that uses the double knee joint now. How much more flexible can something be, than the SES?

8-Dale

Adrenalynn
04-10-2008, 12:33 PM
Thanks for your input, Robotguy!

I concur that it would appear the SES is more flexible and modular. Not having seen your design yet, I think the gripper on the SG5 appears more useful than that of the SES package. Of course, there's probably no practical reason why they can't be mated together - the SES arm and the SG5's gripper. I haven't put a tremendous amount of thought into that yet...

I'm really getting to a point where I need to stop thinking about all the bits and pieces I want this project to employ, and start building something. There's a chicken-and-egg thing going on here where I can't code anything until I build the embedded hardware, but I can't build the embedded hardware until I get the arm working, and I can't really figure out if this whole mad plan can even be made to come to fruition until I build the hardware, write the code, put the arm together, and start trying to calibrate the whole mess.

I had hoped to go in stages, but it's beginning to appear like I need to make a leap of faith and simultaneously pull the trigger on everything I think I'm going to need all at once...

LinuxGuy
04-10-2008, 01:00 PM
I concur that it would appear the SES is more flexible and modular. Not having seen your design yet, I think the gripper on the SG5 appears more useful than that of the SES package. Of course, there's probably no practical reason why they can't be mated together - the SES arm and the SG5's gripper. I haven't put a tremendous amount of thought into that yet...
Anything can be put together with anything, given the right brackets and such.. :happy: I guess it's time to put up a picture of the 3D Model I have for the new gripper I designed. I don't have a model for the "fingers" of the gripper that attach to the final servos though, so I will just have to create a new part.


I'm really getting to a point where I need to stop thinking about all the bits and pieces I want this project to employ, and start building something. There's a chicken-and-egg thing going on here where I can't code anything until I build the embedded hardware, but I can't build the embedded hardware until I get the arm working, and I can't really figure out if this whole mad plan can even be made to come to fruition until I build the hardware, write the code, put the arm together, and start trying to calibrate the whole mess.
I think you are grossly over complicating this and just making yourself crazy over it. The first step is design and build something to control. The second step is to write some software to control it. The third step is debug any problems found between step 1 and 2. Those are pretty easy steps to handle, and it breaks things up nicely into smaller chunks you can actually get your head around. Don't even try to see the entire project at once - it will just give you a huge headache. :veryhappy:

When all else fails, just build it and see what happens. :veryhappy: Don't worry about the software until you have something to control.

8-Dale

Adrenalynn
04-10-2008, 01:53 PM
I've never been able to design that way. In past lives, I was a software engineer. I did hardware to support cool software. ;) Then I graduated to an Architect, and that's when you have to start worrying about how all the pieces are going to fit together... These days, I'm the CTO, and I have to worry about not only how all the software/firmware/hardware is going to fit together, but also how all the _people_ will fit together - and how I do it all with no budget and insane time constraints.

Some times I long for the days of 8 bit CISC assembly language and good adult beverages...:veryhappy:

Semicton
04-10-2008, 03:45 PM
I know the feeling Adrenalynn I'm no expert software devloper, and it doesn't help that I am new to the whole idea of automating a robot either lol. I've been playing with a SES using phidgets and a lynx 6 for almost two years now. I'm a little bit different than you however. I get all excited and then start brain storming and then i get so overwhelmed that i burn out and put my project down for months at a time lol.

I like to use pc based robotics boards like phidgets becuase I am a c# developer, but everytime I think about kinematics, I get a little discouraged and then fire up the lynx 6 arm and use the RIOS software package for the ssc-32. If I could get my head around kinematics, I'm sure I would be having alot more fun.

Anyway,of the two arms I have, I like the SES best. I have extra tubing to change the length and bought a few other parts on impulse. The lynx 6 is a nice robot arm but I was surprised how flimsy it was when i set it up, probably because I was used to my SES. I haven't actually built or used a SG5, but I'm guessing it is about as configurable as the Lynx 6 (servos only unless you have a tool shop in your basement hehe)

Cheers :)

Matt
04-10-2008, 09:05 PM
Hey Guys,

I'm going to throw a bit of a wrench in the discussion :D I would HIGHLY recommend seeing if you can bump your budget up a bit more and just get the new AX-12 Smart Arm (http://www.crustcrawler.com/products/smartarm/index.php?prod=12). It's SO much more advanced you are getting your extra expense back ten fold. Just compare the size, strength, sturdiness to the old models not to mention that you have feedback on the Bioloid actuators. We don't carry the arm yet unfortunately, but we are in love with it here at TR. It's worth adding to your list for consideration.

Adrenalynn
04-10-2008, 09:40 PM
Hi Matt,

Thanks for chiming in!

I had looked at it, and I agree that if I were shooting for a general purpose arm, that'd be the way to go, beyond question.

I have a specific project in mind that doesn't need that accuracy and strength. Honestly, I have a Bridgeport in the garage, and a decent (12Gx60") brake. If I were going that route, I'd probably be looking at large steppers or commercial servos, and rolling my own. With a specific project and wanting both rapid prototyping and to sink my budget into the control mechanisms, the lower-end arms are more attractive to me today.

Of course, who knows what the future may bring?

Matt
04-10-2008, 09:48 PM
Okay, well with that in mind, the crust crawler is a more solid arm when sticking to the old school servos. We have the assembled one we are giving away in the next competition, I don't know if we have any in stock other than that and I think CrustCrawler is out of them, but I could be wrong about that... if you find out, let us know here in the thread.

cheers,

Adrenalynn
04-10-2008, 09:54 PM
Well, shoot, that just tosses a wrench in the works!

Matt
04-11-2008, 11:17 AM
Good news, I looked and we do have 2 left :) Sorry about that!

Adrenalynn
04-11-2008, 12:30 PM
Darn. I really wanted to like the SES, because it was sooo simple to machine any custom arm from it. Looking at the SG5, it appears that it's a little more work, but also re-inventable. More complex bends, more milling on the arms, as well as doubled-up identical pieces. But still workable.

I think, based on the strength of your recommendation (you've touched both), I'm leaning towards saying "sign me up!"

I like Lynx's servo controller. It seems to be one of the most mature out there, with good software support. I'm leaning towards their controller. Any input?

Adrenalynn
04-11-2008, 01:13 PM
Just to put some initial closure to this topic, after chatting with Dave, I pulled the trigger on a combination of the basic SG5-UT and the SSC controller.

Thanks for your insight, Dave (and Matt)!

LinuxGuy
04-11-2008, 02:15 PM
I like Lynx's servo controller. It seems to be one of the most mature out there, with good software support. I'm leaning towards their controller. Any input?
Even if you only get the SSC-32 based on price alone, just get it. :) I have looked at other servo controllers, a LOT of them in fact, and I just don't think there is a better value anywhere. My only problem with the SSC-32 (http://www.lynxmotion.com/Product.aspx?productID=395&CategoryID=52) is what I consider flimsy screw terminals for the power connections - I don't think it's reasonable to try plugging 16 awg stranded wire into them.

8-Dale

Adrenalynn
04-11-2008, 02:22 PM
Thanks, RobotGuy! I didn't mean to downplay your insights earlier. I did end-up with ordering the SSC, and awaiting it with bated breath. [Don't mind the smell. ;)]

Semicton
04-13-2008, 12:41 AM
Hey Guys,

I'm going to throw a bit of a wrench in the discussion :D I would HIGHLY recommend seeing if you can bump your budget up a bit more and just get the new AX-12 Smart Arm (http://www.crustcrawler.com/products/smartarm/index.php?prod=12). It's SO much more advanced you are getting your extra expense back ten fold. Just compare the size, strength, sturdiness to the old models not to mention that you have feedback on the Bioloid actuators. We don't carry the arm yet unfortunately, but we are in love with it here at TR. It's worth adding to your list for consideration.

Darn, wish I knew about this before I bought the lynx6. Where has this robotic arm been hiding all my life lol? Ok, Man, I have to have this.... Maybe in a few months.