Hi, I am new to this site and I am looking at using the AL5D kit for building an automated tester. My company makes keypads, rotary switches and encoders. The keypads typically have an actuation force of about 500 grams. My question is, will the arm be strong enough to push down (-90° from the ground) at this force? I have no experience with these robotic arms but after receiving a quote from a robotic arm supplier in excess of $30K I decided to look into doing this myself. I am an electrical engineer with plenty of microcontroller, hardware and programming experience. Any info would be greatly appreciated. Thank You
For starters, if you use a powerful servo such as the one pictured below, it should work. You might even be able to use a less poweful servo. I am not certain about the base with negative forces though. The servos have the power but flex in the arm and base is an un known for me. Usually the arms lift instead of push.
Thanks. I posted in the arm section too. I guess my thought was that if the arm has a 13oz lift capacity then it should be able to push down with the same force if not more because it would benefit from the weight of the arm. Just thinking out loud.
Psst- I moved your post into the arms section; this is a relatively small forum and most people read everything, so there’s no need to post multiple times. 
There is a big difference between an arm that can push down with 500g or more force, and an arm that can push at exactly 500g force. So it depends on what you are requiring. There is a big leap from a $400.00 arm and a $30k arm. Hobby servos are great, but I seriously doubt they will operate for a year doing repetitive tasks 24/7. Not sure how long it would operate before needing maintenance, or servo replacement. The good news is it’s not overly expensive to try it.
Knowing the exact force would be great but is not required. Thought about using some of the FRS to try and measure force. I would think that 500-600g would be about the maximum force I would need to actuate a button. I figured I could try this for under $1000 and if it works great. If not at least it would be fun to try. I guess if maintenance became an issue I could always try and beef up the components. Thanks.
Rather then use an arm, why not use a servo-driven X-Y drive to position a carriage over the desired key (or other actuator), and use a solenoid or perhaps a DC servo motor (stepper?) to provide the key-press force. What were they doing in the $30K 'bot?
A servo or stepper driven X-Y drive can run continuously 24/7, this arm and it’s R/C servos are “Hobby rated”, and not intended for this strenuous duty.
Alan KM6VV
Servo driven x-y would be my first choice for button presses but I am trying to combine button presses and rotating knobs into one assembly. Basically I need to press buttons, rotate 2 knobs on the top surface of the assembly and rotate 1 knob on the side of the assembly. The 30K robot included a software package for about 18k. Would it be possible to use this kit with better servos?
An even better solution is to electronically switch the keypads, and knobs, kind of the same way a microcontroller turns LEDs on or off. An interface board of some sort would need to be made to replace the keypad, but it would be much better than using a robot arm to physically press buttons and rotate knobs.
You’d be hard pressed to find R/C servos with sufficient resolution and range to match typical DC servos. Check out these servo products:
If you can move a carriage to the desired position, then having two “end effectors” operate from the carriage might be a solution. Drop the one you need to use down onto the DUI, and press buttons, or rotate knobs as needed.
A knob on the side? That’s going to be a chore! That suggests another X-Y carriage and an actuator. I suppose you could rotate the platform that you have secured the DUI (device under test) to in order to access the additional knob, but that’s extra!
Alan KM6VV
Can’t replace the keypad with anything. The goal is to test finished assemblies before they go to the customer. I am trying to remove a human operator from the tester.
I didn’t want to suggest that you replace the keypad (DUI). You’ll have to have a fixture that can position the finished assembly to be tested.
Also, note that a lot of the automated test fixtures I see choose to use pneumatics rather then solenoids and servos to “press keys”, rotate knobs and the like.
Alan KM6VV
Alan frenzel was replying to Mike. 
OK, that makes more sense!
A quote in the reply could be helpful.
Alan KM6VV
Oh ok. It all make sense now. The arm kit could be made to do the job but not for industrial use like what you need. You need something that can run all day long and not miss a beat. Some kind of arm would be needed to simulate button pushes but you need something stronger than a hobby servo driven machine in my opinion. I’m not sure what the key pads are installed in or if they are going down an assembly line loose? I regret to say that I really don’t know what to tell you.
If they are loose, the pads would need to be put in to some sort of clamping fixture to make sure each one has the same x,y locations or a camera would need to be used and that’s a whole different ball of wax.
There are so many variables.
You need an X-Y style drive like this one I’m building to route PCBs.
PCB_Router.jpeg
http://www.marconettengineering.com/PCB_Router.jpeg
You could build one from surplus linear bearings and stepper motors. I think I’d use air to actuate the “button pushers”. On a pick-and-place robot I worked on, it had a pneumatic “rotary motor” to rotate a small table 180 degrees. Something like that would work to turn a knob.
Alan KM6VV
Very nice Alan! looks great! What camera did you use to take that picture?
Thanks Mike!
I used my old Nikon 880 Coolpix that I got for Christmas probably 5 years or so ago.
Alan KM6VV
I think one could make a working x-y table prototype using servos fairly inexpensively. A button pushing setup does not require a lot of precision compared to milling setups. I think simple tables have been made using sliding drawer rollers and such. A button pushing servo could be placed on an overhead frame that swings up and back out of the way when changing the keypads. The amount of travel of the button pusher would be small, so a mechanical advantage could be used such that very strong servo would not be needed. How long the setup would last would depend on the design and duty cycle. If the prototype setup seems workable, the servos could possibly be modified to extend their service life or made easy to change out for new ones.
Thank you all for your input. Ithink the general consensus is that the arm could be programmed to do the task but reliability is a huge issue. I think I will look into a few more options.
Thanks Again