Hi all, I need help with this project please. I bought a Lynx 6 Robotic Arm that came with the SSC-32 servo controller board. The board has 2 servo channels which are not used for the arm. I want to mount the arm on a platform that will lift the entire arm up (elevator style almost) and also the platform will be able to slide the arm side to side. So aside from its normal movement, the arm will be raised from the ground (up/down) and side to side (right/left) on the platform. Here’s the fun part, I’m not sure what I should use to elevate the arm, servo motors or a regular motors. If I use servo motors I can use the 2 channels that are unused on the SSC-32 but they would have to be modified for continuous rotation and for torque. Which ever way I go, the platform’s movement must be precise. If the platform moves up/down or right/left it must stop consistent in a given spot. I would like to use the RIOS Arm Control Software (preferred) or the SSC-32 Sequencer. Can someone tell me what I need to buy and what configuration to go with?
I’m a little confused.
You have the SSC-32 that is capable of operating 32 servos and you have all but 2 being used? That’s an arm with 30 servos!
Also, you mentioned you want the platform to move: You mean as if it were on train tracks with the ability to rotate while running back and forth?
No Mike, I think he is talking about using the 2 remaining channels from RIOS. It uses 6 for the arm and 2 left over. I know there are many unused channels on the SSC-32 when operating RIOS. It was a grandfathered limitation. We may address it at some point in the future.
I’m not sure how to address the other questions though…
Well to start how big of a platform are you talking about?
For a suggestion… vxb.com sells several linear rails sets of varying length and diameter for what seems like reasonable pricing. then you need a linear actuator or lead screws and ball nuts to make the platform move. the linear actuator might be a good option as they usually have position feedback built in and require a h-bridge type control to operate. You could probably gut a digital servo for the electronics and use it to operate the actuator. Some of these firgelli actuators are in the $50 USD price range and do 18 inches at 40lbs.
edit: on second look the firgelli actuators don’t appear to have pot feedback like many others and rely on limit switches just to limit travel. so while they are cheap you would need to get position feedback either by hacking them to add a pot or using some other sensor type. we have no idea how precision your position must be so I’ll just stop now. 
First, thank you for your replies. The platform will be made by having two shafts running parallel from one another and two linear bearing sliding units (this will control the left/right movement). I plan to attach a timing belt to the sliding unit with a timing pulley idler on the opposite end from the motor/servo drive end. The two sliding units will be attached together with a flat aluminum piece that will also act as a flat mounting surface for the Robotic Arm. A rough estimate on weight is about 7 to 8 pounds with Arm mounted on platform. I will have two liner sliding units, one on each side of the platform (this will control the up/down movement). The side units will have a timing belt attached with a timing pulley idler on top, a timing pulley attached to a horizontal shaft at the bottom in such a way that when the shaft turns, it will move both sides up/down at the same time. The more I look at it, I’m thinking about going with torque down motors for both, the left/right and up/down movement. I guess I would need an encoder to track the motors movements for accuracy. Can I still use the RIOS Control Software (preferred) or the SSC-32 Sequencer to control everything? If so what motor controller do I need so that I can attach the encoders into the equation? This robot will be used to remove DVD movies from a shelf so the accuracy for the up/down right/left movements can be given leeway by adding more space between DVD movies.
position sensing seems like it will be your biggest hurdle. If you can get something that either has a potentiometer output or produces a proportional voltage then you can use a standard servo controller board with your SSC-32 output to operate the motor. the most direct option, although not necessarily the cheapest, is a linear position sensor. Something like this or this are direct solutions but are in the $200usd range for 18-20inches (45-50mm) of travel. accuracy is pretty good and they would be pretty easy to implement if the price doesn’t floor you. more complicated solutions might be to use a mechanical gearing down from your motor to a multi-turn pot (perhaps on the idler sprocket end), or ir position sensors could be arranged between the two tracks to report the shuttle position. again either of those solutions could be used to report position to a servo board and directly control the motor using pwm input from the ssc32.
I have a 500 and a 1000 count optical encoder that I can attach to the moving shafts (which ever will work better). Will this work? If so, what other hardware would I need to buy to interface everything with? Any specific motor controller, I’m new in Robotics.
ok I am cogitating an idea here so it might be a bit rough.
you said 7-8 lbs in the vertical axis. it might save you a lot of headache and burnt out motors if you counterbalance this with a big spring or easier still a weight over a pulley like the old fashioned window counterbalances.
anyway, I hate to reference servocity but take a look at the robotzone servo power gearboxes. consider that if they are getting 90 deg of rotation with a 5:1 gear reduction then the servo must already have been modified for continuous rotation. if you can replace the potentiometer in those with a 10-turn multiturn pot like a bourns 3501 series, which will require some monkey bussiness because the pot shaft is only 0.5" long, then you get 900 degrees of rotation.
so 900/360 = 2.5 complete rotations. let’s say your tracks are 30" long, circumference = 2xPIxR so 30 / 2PI = 4.7 and divide that by 2.5 = a required sprocket radius of about 1.9 inches.
If their hub sprockets will stack on top of the large gear (rather than the power arm) because getting a 4"+ pitch diameter is really easy that way. I would esitmate about $125usd or so for each track doing it this way.
the problem with an optical encoder is you need something to read it and then implement the servo control loop based on its position. this is why I keep hitting on the standard servo controller board idea because it implements the full servo pid as well as a driver for the motor, and natively understands the ssc-32 output pusles. the limitation is you need to report a position in a format the controller board understands, namely a proportional voltage from either a potentiometer wiper or an appropriately scaled analog voltage from some other type of position sensor.
with that in mind, if these are absolute position optical encoders with parallel binary output then you could probably wire up a D/A converter to translate the digital word to an analog voltage to drive the servo board.
If you want to move away from your initial desire to use the two remaining RIOS+SSC-32 servo channels implement this some other way then there are a lot of other options. you could use the basic atom pro and a bot board to write some software to implement the pid using a motor controller but it’s several more expensive parts. you can even read the pulse inputs from the ssc-32 so that would let you stay with using RIOS+ssc-32 idea. I don’t know that I would say it was an easy thing to do though for someone new to the scene. it’s sort of like testing the water with both feet if you know what I mean.
What do you think about this gear head servo? I did a search on the forum and found that it’s a new product (not sure when it will be available). lynxmotion.net/phpbb/viewtopic.php?t=2211
This is something new that Lynxmotion is working on. They have a few specific applications in mind but it should integrate well with the rest of the SES system as well. It has lots of torque, and the control they propose integrating with it has some excellent features that will put the final product in a class almost by itself. I am not certain though how you would apply this particular gearhead servo to the system you were asking about. How were you thinking of using it?
My biggest weakness at the moment is programming. I haven’t worked with microcontrollers (to a programming level) in 10 years. I’m sure things have changed somewhat. If I can interface this gear head servo to the SSC-32 Controller and be able to use the RIOS software, I’m sure I can make the mechanics side of the equation work. I can change the design of the platform that the Robotic Arm will mount to accommodate this new servo. I’m not sure if this answers your question or not. Do you know when this servo is due of on the market?
Not to throw a fly in the buttermilk, but the project demensions have not been stated so far (is horizontal and vertical movement going to be 4" or 40’?). Specifically what is the gizmo expected to do. The arm you have is probably not best suited for the job you want to do. How many DVDs are involved? I’d generally ignore “future state” servo talk and stick to products currently available. Your project is probably fairly easy to make, but more info is needed.
Ok sure, let me be more specific with the project dimensions an functionality. The Robotic Arm will travel 6 feet horizontally and 4 feet vertically, and the only weight the Arm will handle will be the weight of a DVD in its case. A DVD shelf will be built to the wall and the Robotic Arm and platform with be attach to the front of the DVD shelf. I plan to install glass or Plexiglas sliding doors on the front of the entire thing (so it’s important the every aspect of the robot looks clean). It will have and opening (the size of a DVD standing upright) located to the bottom right side of the shelf. This is the location the DVDs will be set by the Arm as a delivery or as a pickup location to have the Robot shelf the DVD again. This area will be considered the Robot’s “home positionâ€. At this location the servos will have no force applied to them. The platform the Arm will ride on (the vertical travel for the arm) will weigh about 7 to 8 pounds with the Arm installed. This is where I need 1 of 2 high torque servo/motor to run the vertical. To cut down on weight, I changed the way the arm travels horizontal. The Arm will travel on a small light weight cart that will be guided in a track. The cart will have the Arm mounted and the second high torque servo/motor. The cart’s weight is about 4 to 5 pounds. I want to control the two high torque servos with the SSC-32 and the RIOS software.
this sounds a lot like a commercial application.
at 6 feet you are more than likely going to need that encoder you mentioned as most linear potentiometer type devices will be cost prohibitive. your best bet is probably the encoder based feedback version of the openservo design one of their guys was talking about in another thread recently.
from past experience I suspect you will need to solve a tendancy for the vertical slides to bind due to the horizontal track not being evenly loaded (i.e. the shuttle is to one side or the other of center.) You might want to investigate using twin lead screws on the ends and stepper drives, or having a gear rack on each end and a driven horizontal shaft with gears on each end. OTOH maybe you’ve built it all mechanically already, it works great, and you are just looking for a drive system. good luck to you.
Not at all Eddie, it’s just that a have collected a lot a DVDs over the years and now I have to find a better way to organize the mess. Apart from taking forever to find a given movie I want, I sometimes forget the ones I own and buy them over again. I find that this Robot will help me keep organized and decorate my media room wall nicely. Tell me what you think about this idea, what if I use regular servos to trigger a switch. I have two high torque motors that can handle the weight easy, I can setup the motors to turn one direction when a switch is triggered, have it stop on a dime when the switch is not triggered, and have it move the opposite direction when another switch is triggered. I would set a switch on one side, a servo in the middle with a small arm or paddle (to better trip the switches), and another switch on the other side. I can do this for both the vertical and the horizontal.
If I were designing this system, I’d first determine if a servo based setup would have the positional accuracy to do what is desired. As to servo positional capability, I’ve found that a standard servo has a positional capability of about 400 descrete positions in its 180 deg rotation (or about 340 deg rotatation of the servo pot). This would mean that you could have no more than 400 dvds in a row, and maybe less. You would have to fabricate some gear reduction such that the 6 feet of travel would result in 340 deg rotation of the pot. Substituting a precision 10 turn pot might add some extra position range, but I’ve not seen anybody testing such a setup. Your platform needs to roll instead of slide to minimize friction issues. Check the sliding door and drawer sections of the home improvement stores for various track roller setups. If you are interested doing your own testing, I suggest you get one of the below servos and modify it for continous rotation as shown. Then you can use the servo for various test to determine servo suitability for the project. There are probably many other methods positioning the platform at the desired dvd if a servo based setup is not practical. The success of the project depends on how much R&D effort you want to put into it.
hm, i haver read this post and im wondering if he would be limited to just pot feedback, he could always find a clever way to make it so he doesn
't have to limit himself to pot feedback
he could set up a ype of swith that depresses when it passes a little notch in the track its on, this way hes wouldn’t have to get all complicated with pot feedback, the micro just counts how many times the switch has been depressed for a number, 1-w/e, you wouldn’t really be limited, this is of course a suggestion if I had missed if he said he wanted pot feedback 
I infer that the basic requirements are to construct the gizmo as easily and cheaply as possible, by using a Lynx 6 Robotic Arm that came with a SSC-32 servo controller board, and additional parts as necessary.
I have seen a project where someone took a pan and tilt servo, attached a laser, and programmed it to point to locations on a wall full of parts bins to search for parts.
As for the DVDs, you could group them in 5’s or 10’s and ther servo would not need to be so precise. you could have the servo point to one group of 10 DVD’s and you know what you are looking for is in that group.