Finding a lost airplane? (I hope they do find it)...... But that's a very good question. I would think the opposite of avoidance being detection like a person entering and area? I know their will be other creative ideas chiming in
Proximity detection, i.e. people/animals moving past it.
Distance measuring/mapping (although this get's complicated and is something I've not yet tried to tackle)
I have been thinking about putting ping sensors on all sides of my bot, and when it detects something (person, animal, other bot) it would turn the camera to look and try to identify, rather than just trying to avoid.
Hey Alan, I was thinking along similar lines... Only I was going to use "focused" PIR detectors for the same purpose... Whatever side the robot picks up motion on it will rotate to face the direction of movement...
All great ideas. I was going to omit the ping from this design, but I like the idea of the sensor detecting movement and looking in that direction.
I'm currently looking for anyone interested in helping with the programming in EZB. I will be too busy printing, molding, running pieces to make my deadline (this is for a Kickstarter campaign). If you are interested in working on it on please contact me privately. Probably the programming could be started now and fill in the values when the prototype is finished. I can offer a trade, programming for one of the kits.
Yes. I've turned down a film to get it done and launch the Kickstarter. I'm spending hours and hours getting all the parts reworked. In reality this is V2.0. With some updating some design programs I use, I able to really push in some great details. Also reworked all the mechanical portions. Its a much better product now (although not quite done about two more weeks). Great learning platform. Thank goodness for all the great customers over at the B9 builders, I was able to get most of my financing from this last batch of robot parts. How about that? Robots supporting robots?!
Navigation , relationship to the robots location in a room you already have measured out for your robot. Or maybe to determine the size of a room in a proxy centimeters.
If you can get some scripting in line to record distance then reference it on a table then a robot can crudely map a room when it enters it then switch to distance for the robot check and updates aprox position in that map. Distance sensors can be utilized for motion if they are checked often enough but the robot would need to be stationary at that moment to determine the sudden change in distance readings. A distance sensor faced down can assist in determining the difference between carpet or hard wood as carpet will be closer to the sensor.
It doesn't need to be stationary for proximity to be detected, you can easily detect a big jump which would detect movement of something else.
@Fxrtst, feel free to send me over some info on this if you want to (my email is in my user details). While I can't promise anything as I am very pushed for time at the moment and under a lot of pressure as far as work goes, it's usually feast or famine, currently it's feast so if the famine comes I'd be able to jump on board. Also feel free so ask for any guidance if needed, I'm always more than happy to help when I can.
Great Rich. Film work is the same way. I get three calls in a day or none for a month. LOL. I will send you over a few pics. Any help would be appreciated.
Also because I'm in a time crunch if anyone else would like to jump on board, feel free to contact me.
I'm probably going to need all the help I can get to get this done in time for shooting the video for Kickstarter. I'll let you know. Thanks for the offer!
Oh and @Richard, don't sell yourself short Think logically and follow the logic and you'll be writing awesome scripts. The only difference between me and anyone else who can follow the logic in a script is I know pretty much all script syntax and controls by memory.
DJ has done an awesome job with EZ-Script and ARC so it really is very easy to do most things (so easy I often find myself over complicating things only to find a simpler method after I've finished).
The hardest part of programming using ARC is the typing
As I said, I'm insanely busy with work at the moment so may not have time to do much (making hay while the sun shines and all that, I need to earn lots more money to afford a 3D printer so I can get all these ideas in my head out of my head lol) however am always on hand for quick and simple guidance or advice
@Fxrtst one of my goals for my B9 is to have him extend his claw and shake hands when he detects a face in front of him. I am thinking that it would be a good idea for him to ping first and make sure that the person is not too close before he sticks his arm out and clocks someone.
Here is the first test of the EZ:1 wide beam sonar array using 4 low cost (less than £10 for the four) automobile ultrasonic transducers. It has 4 sense zones where the beams just slightly overlap each other so most front objects (even slim chair legs) can be detected also giving their position. In this photo a 25mm wide piece of plastic trunking is place 60cm away from the inner right sense zone (its right because the robots leg section is upside down for the tests).
As you can see, only that particular sense zone detects the trunking, all other zones show "clear" which means that there are no obstacles (within 2.5m range) of their respective sense zones.
Here I come close to the sonar array (0.5m) and all zones pick me up (I have a pretty wide chest!). You can also see the new "sense field" monitor (forth line at the bottom of the display) that works with on 3 distance zones which cover 360 degrees around the robot body, and detects all objects and humans/animals etc, the two systems working together make a really nice sensor array for the robot.
@Fxrtst If I can be of any help to you on your project please let me know? I would be happy to help.
@fxrtst, After reading again your original posts, I think you are looking for sensors to detect people etc around the robot?
On my EZ robots I use miniature Panasonic PIR motion sensors, they have a detect range of 5m with a beam angle of 100 degrees, here is a link for them.
I have sent some of these out to Dave and Mel, but I am not sure if they have tried them yet?
This photo shows how I use them, one goes in each left and right ear lobe, when they detect moving heat radiation the blue LED rings glow. The front facing one you will see in the nose area and its a different type that has a tighter beam angle of 38 degrees.
I currently use 3 on my robots one each side and one in the front. I also developed a special algorithm (on an 8pin PIC) to make them more intelligent. I needed to do this as I found that if the main controller (EZ-B etc) just looks out for the 3 digital inputs, it can get very confused if there are a few people moving around its sphere of view! So what the algorithm does is monitor the 3 inputs over a set period of time and calculates where (which side or front) the most movement is, this is the most interesting side to be looking. How this works is say you had one person on the left side and two people on the right, then the PIC would flag to the EZ-B that the right side was more interesting etc and your script could send the robots head to look at this (more activity) side first.
If you need any more info on this, just let me know.
@Toymaker , no I haven't used these yet but plan on making them my choice for movement detection. This feature is a little down the list of priorities. I just haven been able to get the needed time to finish other parts of my B9. I'm a little conserved though about the "confusing" issue you mention. I'm not sure I have the talent and know how to incorporate anything like an algorithm program like you did and incorporate it into ARC. confused Heck, I'm still struggling with a ramping script to help my Sabertooth / Kangaroo X2 combo start and stop smoothly. Thanks to Rich I was able to ramp up nicely but having trouble getting it to ramp down. Anyway that's for another thread.
Personally with the little I know about sensors I think there may be a better choice then the ping sensor for this application.
I brought up this discussion as I needed to make a design decision, based on the space I have left. As this project is "kit" based, customers will be able add items on their own. I'm trying to keep my price point down by not including any hardware (servos, EZB or sensors), but to give customers optional places in design, to place a ping or PIR or other sensors. But the 3d printed prototype for Kickstarter should have as much hardware as I can cram in there.
@Tony , thanks for all the links and info. You certainly will be quarried many times in the next few weeks, as I get the roll out going.
@Josh , I'm going to keep it all private for now. But, I'm sure I've discussed it here before.
@Dave , Yeah I was hoping the ping would work, but I just don't know enough. I'm hoping when this is released people will cloud their scripts and whole projects, then others will build on that.
I am hoping that @Will's Kickstarter project is a scaled down B9 Robot, if it is I would certainly would purchase one! @Will, if I can be of any technical assistance on your project, please do not hesitate to ask?
@Dave, you could do the simple PIR algorithm in an ARC script, I just like to do "pre-processing" with support microcontrollers so that the EZ-B can do other more important things.
I was privileged enough to see Will's project. You guys will be amazed. Real Hollywood grade stuff. I'm not really surprised really though. I expect stuff like this to come out of a mind like Will's. He didn't win that award he's holding in his avatar for nothing. It's kind of a big deal. You guys gotta stay tuned to this one.
@ Tony, i actually have a 80% complete B9 in the computer, just waiting to get out at any scale! But have you seen the new Mobeus B9, its beautiful! And thanks so much for the offer on any info/advice/help. I'm going to need it!
NEW B9 1/6th scale
@Dave, Thanks again at taking a look. Its shaping up nicely. Hopefully something fun to stick a Ezb v 4 into. Hope those ship REAL soon! Hint hint!
@will, if you do need any tech help on hardware or programming for your project, you can email me direct at toymaker(dot)tony@gmail(dot)com, as I have said I would be happy to help.
Wow, I like the Mobeus B9, I wish I was a better modelmaker, if I was I would get one but I am useless at assembling model kits!
@Will , Wow, your going to be able to print a B9 at any size? That's mind blowing! I imagine needed and unavailable parts could be printed the same way. A person could print out a power pac, neck bracket, finger lights or ear posts! Your new project is just screaming for a V4 EZB.
I have one of these new Mobeus B9 models now waiting to be put together. Sadly my big B9 comes first and I don't see myself having time for the little model for a long time.
I'm going to hijack this thread a little here (sorry Will) ; @Tony, could you take a look at my other Kangaroo thread? If you see any ideas on programming ramping my Kangaroo X2 and have a little time I could use a little advice in that thread. No worries if you don't have the time. OK, Hijack over.
@Will , speaking of ramping, is this a problem with the motors on your new project? Without giving too much away are your motors starting and stopping smooth enough for your distinct taste?
@Dave Everything is direct drive ( that is all the axis are connected directly to servos). I'm not using any exotic motors, so that the average person can get off the shelf components to assemble it. With servo speed, I can keep some parts from tearing apart by jerking around from the servo, by basically damping the speed. I am looking forward to 180 steps for the servos! Nice smooth movement. If the front becomes a little too heavy I will counter with springs on the opposite side , a great trick from my animatronic days.
Yes A b9 could be printed, they even have some very cool rubberly type filaments called ninja. This stuff is a close as you can get to rubber. Its actually a urethane and needs a special extruder. But exciting to print in all these cool materials. If you have a cool $330,000 laying around, you can get the new Objet. You can print in many types of materials at the same time. So you could print a pair of glasses one material for the frames and the clear lenses at the same time. Its getting really very cool. Here is a link :
Objet Connex3
@Will, can I ask your advice, from your experience of animatronics and springs etc can you help me with a problem that I have with the EZ:1 arm which has backlash and flex issues in the arm assembly which can be seen in this video.
Do you think counter springs may work on this type of problem? I want to try to reduce the flexing etc.
@Tony, every time I see this video I'm amazed. We're so lucky you have a robotics person of your caliber a member of our group.
I don't even to pretend that I'm even close to an animation expert like Will but I do have a lot of building experience. Here's what I see (and I'm sure you do too); The length of your arms and their weight is causing it to bob at the joints whenever they move. I know of several ways to stop this. The surest way is to beef up the joints with heaver brackets and larger motors. Of course I know that's not an option because of your design and cost restraints. Perhaps at least a different type of shoulder joint should be considered. In a different thread one of our forum members uses a Vex Turntable bearing kit for a shoulder joint. It looks like a good answer to robot shoulder joints because it's small and strong. Look here:
Other then that I think your only other choice to stabilize the arm joints are springs or small gas struts. Struts are intriguing and really do a nice job. I have a pair on my B9 to assist with the weight when he bends over at the hips. I was amazed how nice it works and when at rest they hold position and weight with no pressure on the motor shaft. The down side is you need it sized properly so the weight of the part your supporting or gravity will assist the motor. They also look cool (think of C3P0). Springs are your other answer but I have little experience here. I know they have been used forever with success to assist motors. The other Tony that built his Lexi 3000 used them in his arm design. They really worked nice for him and he even installed them in a small tube to help them work and keep them from view. I was really impressed with his solution and how it turned out with the springs being hidden in the tubes.
@Tony, I'd say Dave is correct. The wobble comes from joints or motors that are not tight. Springs could take out alot of the wobble and dampen the motion. Think articulated desk lamp.
Hey Dave, I agree the Vex turntables are an awesome cheap but strong shoulder joint... You can even use 2 motors (that's what I have done) to double your lift load...
Two wire 393 motors... you'll need a few more parts like a couple of 2 or 3 inch shafts, shaft locking collars and 2 x 12 tooth metal pinion gears... however, I would buy the clawbot kit too (instead of the individual parts) because you get way more parts, including 4 x 393 motors , nuts and bolts, shafts, collars, metal and lots more... You'll still have to buy the turntable and 12 tooth metal pinion gears separately (they don't come with the clawbot kit)...
Finding a lost airplane? (I hope they do find it)...... But that's a very good question. I would think the opposite of avoidance being detection like a person entering and area? I know their will be other creative ideas chiming in
Proximity detection, i.e. people/animals moving past it. Distance measuring/mapping (although this get's complicated and is something I've not yet tried to tackle)
I have been thinking about putting ping sensors on all sides of my bot, and when it detects something (person, animal, other bot) it would turn the camera to look and try to identify, rather than just trying to avoid.
Alan
Hey Alan, I was thinking along similar lines... Only I was going to use "focused" PIR detectors for the same purpose... Whatever side the robot picks up motion on it will rotate to face the direction of movement...
One of my posts, I think it was my introduction to scripting, covers moving towards or away from detection which may be useful for that
All great ideas. I was going to omit the ping from this design, but I like the idea of the sensor detecting movement and looking in that direction.
I'm currently looking for anyone interested in helping with the programming in EZB. I will be too busy printing, molding, running pieces to make my deadline (this is for a Kickstarter campaign). If you are interested in working on it on please contact me privately. Probably the programming could be started now and fill in the values when the prototype is finished. I can offer a trade, programming for one of the kits.
Thanks in advance,
Will
I know we chatted once before about your kit. Is this the same one and can you refresh my memory?
Yes. I've turned down a film to get it done and launch the Kickstarter. I'm spending hours and hours getting all the parts reworked. In reality this is V2.0. With some updating some design programs I use, I able to really push in some great details. Also reworked all the mechanical portions. Its a much better product now (although not quite done about two more weeks). Great learning platform. Thank goodness for all the great customers over at the B9 builders, I was able to get most of my financing from this last batch of robot parts. How about that? Robots supporting robots?!
Navigation , relationship to the robots location in a room you already have measured out for your robot. Or maybe to determine the size of a room in a proxy centimeters.
If you can get some scripting in line to record distance then reference it on a table then a robot can crudely map a room when it enters it then switch to distance for the robot check and updates aprox position in that map. Distance sensors can be utilized for motion if they are checked often enough but the robot would need to be stationary at that moment to determine the sudden change in distance readings. A distance sensor faced down can assist in determining the difference between carpet or hard wood as carpet will be closer to the sensor.
...another great idea
It doesn't need to be stationary for proximity to be detected, you can easily detect a big jump which would detect movement of something else.
@Fxrtst, feel free to send me over some info on this if you want to (my email is in my user details). While I can't promise anything as I am very pushed for time at the moment and under a lot of pressure as far as work goes, it's usually feast or famine, currently it's feast so if the famine comes I'd be able to jump on board. Also feel free so ask for any guidance if needed, I'm always more than happy to help when I can.
Great Rich. Film work is the same way. I get three calls in a day or none for a month. LOL. I will send you over a few pics. Any help would be appreciated.
Also because I'm in a time crunch if anyone else would like to jump on board, feel free to contact me.
@fxrtst... I would love to help and although I'm decent at programming with ARC, it would be stupid to use me over Rich....
I'm probably going to need all the help I can get to get this done in time for shooting the video for Kickstarter. I'll let you know. Thanks for the offer!
Cheers and good luck...
OK Rich sent you a package of stuff.
Perfect, got it, I'll have a look through shortly
I was gonna use it directly in front of my doombot's claw, amongst other areas, so it opens and closes when an object is in front of it
Oh and @Richard, don't sell yourself short
Think logically and follow the logic and you'll be writing awesome scripts. The only difference between me and anyone else who can follow the logic in a script is I know pretty much all script syntax and controls by memory.
DJ has done an awesome job with EZ-Script and ARC so it really is very easy to do most things (so easy I often find myself over complicating things only to find a simpler method after I've finished).
The hardest part of programming using ARC is the typing
As I said, I'm insanely busy with work at the moment so may not have time to do much (making hay while the sun shines and all that, I need to earn lots more money to afford a 3D printer so I can get all these ideas in my head out of my head lol) however am always on hand for quick and simple guidance or advice
Yea on my omnibot I has distance sensors in the claw to tell when a item like a cup or can of coke is inside , triggering it to vlose.
So what is it you are working on?
@Fxrtst one of my goals for my B9 is to have him extend his claw and shake hands when he detects a face in front of him. I am thinking that it would be a good idea for him to ping first and make sure that the person is not too close before he sticks his arm out and clocks someone.
Here is the first test of the EZ:1 wide beam sonar array using 4 low cost (less than £10 for the four) automobile ultrasonic transducers. It has 4 sense zones where the beams just slightly overlap each other so most front objects (even slim chair legs) can be detected also giving their position. In this photo a 25mm wide piece of plastic trunking is place 60cm away from the inner right sense zone (its right because the robots leg section is upside down for the tests).
As you can see, only that particular sense zone detects the trunking, all other zones show "clear" which means that there are no obstacles (within 2.5m range) of their respective sense zones.
Here I come close to the sonar array (0.5m) and all zones pick me up (I have a pretty wide chest!). You can also see the new "sense field" monitor (forth line at the bottom of the display) that works with on 3 distance zones which cover 360 degrees around the robot body, and detects all objects and humans/animals etc, the two systems working together make a really nice sensor array for the robot.
@Fxrtst If I can be of any help to you on your project please let me know? I would be happy to help.
Tony
I see Paris, I see France, I see EZ:1's underpants LOL
Sorry, childish I know
@ Justin, LOL. @Tony, thanks very much. I'm sure there will be some upcoming questions. @Josh, was that question directed toward me, or Doombot?
@fxrtst, After reading again your original posts, I think you are looking for sensors to detect people etc around the robot?
On my EZ robots I use miniature Panasonic PIR motion sensors, they have a detect range of 5m with a beam angle of 100 degrees, here is a link for them.
uk.farnell.com/panasonic-ew/amn31112j/sensor-motion-5m-100-82-white/dp/1373711
I have sent some of these out to Dave and Mel, but I am not sure if they have tried them yet?
This photo shows how I use them, one goes in each left and right ear lobe, when they detect moving heat radiation the blue LED rings glow. The front facing one you will see in the nose area and its a different type that has a tighter beam angle of 38 degrees.
uk.farnell.com/panasonic-ew/amn33112j/sensor-motion-5m-38-22-white/dp/1373715
I currently use 3 on my robots one each side and one in the front. I also developed a special algorithm (on an 8pin PIC) to make them more intelligent. I needed to do this as I found that if the main controller (EZ-B etc) just looks out for the 3 digital inputs, it can get very confused if there are a few people moving around its sphere of view! So what the algorithm does is monitor the 3 inputs over a set period of time and calculates where (which side or front) the most movement is, this is the most interesting side to be looking. How this works is say you had one person on the left side and two people on the right, then the PIC would flag to the EZ-B that the right side was more interesting etc and your script could send the robots head to look at this (more activity) side first.
If you need any more info on this, just let me know.
Tony
@Tony... you have some serious design, electronics and programming skills thanks for helping us out here...
@fx , it was directed to you
@Toymaker , no I haven't used these yet but plan on making them my choice for movement detection. This feature is a little down the list of priorities. I just haven been able to get the needed time to finish other parts of my B9. I'm a little conserved though about the "confusing" issue you mention. I'm not sure I have the talent and know how to incorporate anything like an algorithm program like you did and incorporate it into ARC. confused Heck, I'm still struggling with a ramping script to help my Sabertooth / Kangaroo X2 combo start and stop smoothly. Thanks to Rich I was able to ramp up nicely but having trouble getting it to ramp down. Anyway that's for another thread.
Personally with the little I know about sensors I think there may be a better choice then the ping sensor for this application.
I brought up this discussion as I needed to make a design decision, based on the space I have left. As this project is "kit" based, customers will be able add items on their own. I'm trying to keep my price point down by not including any hardware (servos, EZB or sensors), but to give customers optional places in design, to place a ping or PIR or other sensors. But the 3d printed prototype for Kickstarter should have as much hardware as I can cram in there.
@Tony , thanks for all the links and info. You certainly will be quarried many times in the next few weeks, as I get the roll out going.
@Josh , I'm going to keep it all private for now. But, I'm sure I've discussed it here before.
@Dave , Yeah I was hoping the ping would work, but I just don't know enough. I'm hoping when this is released people will cloud their scripts and whole projects, then others will build on that.
I'm really looking forward to seeing your magic. If it's anything like the rest of your work it's going to be amazing!
I am hoping that @Will's Kickstarter project is a scaled down B9 Robot, if it is I would certainly would purchase one! @Will, if I can be of any technical assistance on your project, please do not hesitate to ask?
@Dave, you could do the simple PIR algorithm in an ARC script, I just like to do "pre-processing" with support microcontrollers so that the EZ-B can do other more important things.
@Richard, thanks for the kind words.
Tony
I was privileged enough to see Will's project. You guys will be amazed. Real Hollywood grade stuff. I'm not really surprised really though. I expect stuff like this to come out of a mind like Will's. He didn't win that award he's holding in his avatar for nothing. It's kind of a big deal.
You guys gotta stay tuned to this one.
@ Tony, i actually have a 80% complete B9 in the computer, just waiting to get out at any scale! But have you seen the new Mobeus B9, its beautiful! And thanks so much for the offer on any info/advice/help. I'm going to need it! NEW B9 1/6th scale
@Dave, Thanks again at taking a look. Its shaping up nicely. Hopefully something fun to stick a Ezb v 4 into. Hope those ship REAL soon! Hint hint!
@will, if you do need any tech help on hardware or programming for your project, you can email me direct at toymaker(dot)tony@gmail(dot)com, as I have said I would be happy to help.
Wow, I like the Mobeus B9, I wish I was a better modelmaker, if I was I would get one but I am useless at assembling model kits!
@Will , Wow, your going to be able to print a B9 at any size? That's mind blowing! I imagine needed and unavailable parts could be printed the same way. A person could print out a power pac, neck bracket, finger lights or ear posts! Your new project is just screaming for a V4 EZB.
I have one of these new Mobeus B9 models now waiting to be put together. Sadly my big B9 comes first and I don't see myself having time for the little model for a long time.
I'm going to hijack this thread a little here (sorry Will) ; @Tony, could you take a look at my other Kangaroo thread? If you see any ideas on programming ramping my Kangaroo X2 and have a little time I could use a little advice in that thread. No worries if you don't have the time. OK, Hijack over.
@Will , speaking of ramping, is this a problem with the motors on your new project? Without giving too much away are your motors starting and stopping smooth enough for your distinct taste?
@Dave Everything is direct drive ( that is all the axis are connected directly to servos). I'm not using any exotic motors, so that the average person can get off the shelf components to assemble it. With servo speed, I can keep some parts from tearing apart by jerking around from the servo, by basically damping the speed. I am looking forward to 180 steps for the servos! Nice smooth movement. If the front becomes a little too heavy I will counter with springs on the opposite side , a great trick from my animatronic days.
Yes A b9 could be printed, they even have some very cool rubberly type filaments called ninja. This stuff is a close as you can get to rubber. Its actually a urethane and needs a special extruder. But exciting to print in all these cool materials. If you have a cool $330,000 laying around, you can get the new Objet. You can print in many types of materials at the same time. So you could print a pair of glasses one material for the frames and the clear lenses at the same time. Its getting really very cool. Here is a link : Objet Connex3
@Will, can I ask your advice, from your experience of animatronics and springs etc can you help me with a problem that I have with the EZ:1 arm which has backlash and flex issues in the arm assembly which can be seen in this video.
Do you think counter springs may work on this type of problem? I want to try to reduce the flexing etc.
Thanks in advance for any help.
Tony
@Tony, every time I see this video I'm amazed. We're so lucky you have a robotics person of your caliber a member of our group.
I don't even to pretend that I'm even close to an animation expert like Will but I do have a lot of building experience. Here's what I see (and I'm sure you do too); The length of your arms and their weight is causing it to bob at the joints whenever they move. I know of several ways to stop this. The surest way is to beef up the joints with heaver brackets and larger motors. Of course I know that's not an option because of your design and cost restraints. Perhaps at least a different type of shoulder joint should be considered. In a different thread one of our forum members uses a Vex Turntable bearing kit for a shoulder joint. It looks like a good answer to robot shoulder joints because it's small and strong. Look here:
Pics of Vex Turntables
Vex Turntables Web site
Other then that I think your only other choice to stabilize the arm joints are springs or small gas struts. Struts are intriguing and really do a nice job. I have a pair on my B9 to assist with the weight when he bends over at the hips. I was amazed how nice it works and when at rest they hold position and weight with no pressure on the motor shaft. The down side is you need it sized properly so the weight of the part your supporting or gravity will assist the motor. They also look cool (think of C3P0). Springs are your other answer but I have little experience here. I know they have been used forever with success to assist motors. The other Tony that built his Lexi 3000 used them in his arm design. They really worked nice for him and he even installed them in a small tube to help them work and keep them from view. I was really impressed with his solution and how it turned out with the springs being hidden in the tubes.
Good luck! Your work and talent is amazing!
@Tony, I'd say Dave is correct. The wobble comes from joints or motors that are not tight. Springs could take out alot of the wobble and dampen the motion. Think articulated desk lamp.
Hey Dave, I agree the Vex turntables are an awesome cheap but strong shoulder joint... You can even use 2 motors (that's what I have done) to double your lift load...
I'm going to get one right now.
@Richard, What motor are you using to move this turntable set? Did you get one of Vex's motors and shaft kits?
Two wire 393 motors... you'll need a few more parts like a couple of 2 or 3 inch shafts, shaft locking collars and 2 x 12 tooth metal pinion gears... however, I would buy the clawbot kit too (instead of the individual parts) because you get way more parts, including 4 x 393 motors , nuts and bolts, shafts, collars, metal and lots more... You'll still have to buy the turntable and 12 tooth metal pinion gears separately (they don't come with the clawbot kit)...