Opinions on micro servo robotic arm using cables

I just made a simple robotic arm using… Popsicle sticks/hotglue. It’s actually pretty dang sweet, except i have 1 small problem.
The weight of the additional servos and the arm is heavy on the base servos, it cant pick up anything heavier than a keychain-laser pointer. Though it’s accurate enough to pick up a screw.

So i was thinking of using cables, much like how the brakes on a bike work, or the throttle on an engine, and some nice bearings on the joints.

What are your opinions on using cables on the joints of a micro servo scale robotic arm?
Feel free to give in some designs or concepts. Though it’s all on a budget, i got plenty of servos, but limited funs on building components. I can pick up the ball bearing spinners for $2 for 4 of them, other than that, i’m using popsicle sticks and hotglue. I got plenty of cable though.

Cables might work, although you’d have the overhead of the weight. I’d look at pushrods, there a few LM leg designs with pushrods.

Our Intuitive Surgical 'bot uses tons of cables, both in the “master” arms and the “slave” arms!

Alan KM6VV

That’s strange as I started a simple arm last summer made from craft sticks (like large popsicle sticks), popsicle sticks, hot glue, paperclips, small plastic bottle caps, rubber bands, dental floss, and servos. Now with the availabity of $5 servos and $10 servo control chips, I may pick back up on it. For a small light weight arm, the servo weight needs to be kept as near the base as possible and off the arm if possible. In my design, all the servos were to sit on the rotating base. The upper arm was directly connected to a servo on the base. The forearm was operated via the dental floss, and the gripper was also to be operated via dental floss. The traditional setup with a servo at each joint makes for a heavy arm. I’ve seen some interesting light weight designs. If I come across them again I’ll post a link.

That’s exactly the problem i’m having, the servos at the joints do make for a heavy arm.
I reduced the length by 1 inch, and repositioned a servo so it sits more on the center of gravity. But it’s still a tiny bit to heavy, i can hear the servos working when the arm is fully extended, though that’s not a real surprise, or uncommon for servos.

Using cables and tubes like a bikes breaking/gear shifting system, would take the weight of the servo off the arm, but the rigidness of the cables may be just as bad, unless i find some that are small enough, I was thinking of the cable system for the throttles of gas powered RC airplanes, it should be small and light enough for this arm.

I have to use micro servos as my robot is way to small for normal servos, the weight of normal servos could potentially tip the whole robot forward, if i were to pick up anything.
I’m going to check out possible cable possibilities later today, And i’ll see about uploading some pictures of my robots current construction.

Just with cables, you can have most of the servos away from the arm, allowing you to make a smaller, lighter arm.

BTW i have an off topic question… How many times can i write to the BAP-28-M, i know alot of rom/flash based components have a limited number of write cycles. How many does the BAP have? i’ve probably written to the thing 50 times now, which is me being conservative, but if i knew just how many it had, i could write to it more often.

Another question.

The BB2 power input says 7.2-9v, will i be risking damage if i use 10.8v?
i normally use 8.4v (71.2v AAA) rechargable batteries, leftovers from an old RC helicopter i crashed in 5 minutes.
But i also have 10.8v (9
1.2v) AAA packs too, i havent used them, i thought i’d ask first, i’m trying to modify them to remove 2 cells, with limited success. I may have to get a 9v voltage regulator, i have one, but it seems… off.

I’d use my 5v li-ion battery, but i dont think 5v would be enough.

I have experience using cables for the gripper open/close on our older L5 arms. I used a single pushrod configuration with a spring. The spring tried to open the gripper and the cable cosed it. The strength of the closing was limited directly by the plastic tubing. The cable was plenty strong though.

For your use you would need to have a push / pull configuration. The strength limit is still there though. There is also a limit to the range each axis is capable of reaching. I’m just not sure how to overcome these restrictions. You will need a lot of custom miniature components with precision tolerances to make anything useful in my opinion. Good luck with it!

I’m guessing the Atom can be programmed thousands of times.

It’s not a problem to power the Bot Board from up to 12 vdc. But you must watch the current draw from the 5vdc side. If it gets too high the regulator will heat up. The regulator does not have a heat sink, so it can’t be used to it’s full spec of 500mA.

As far as a source of cables, do you have a friend who plays electric guitar? Electric bass is even better.

My son is the bassist for “Barricades” here in the US, and gives me all his guitar strings when he replaces them. They make great cables and flexible actuators!

Check out my project at lynxmotion.net/viewtopic.php?f=20&t=5948. These pictures show a guitar string cable. Cable guides can be made made from inepxensive poly tubing from a hardware or pet store.
djhttp://www.lynxmotion.net/download/file.php?id=295&mode=view

My robot, with popsicle stick arm.
Sorry, my camera is really bad. REALLY bad, it’s one of those $20 walmart cameras.
Pff 3.1MP camera… camera phone takes better pictures.


That is the cutest little rover ever! :smiley:

Why not just scale the servos down from the shoulder up? Use a 485 for the shoulder, a 225 for the elbow, and an 85mg for the wrist…

Fishing line makes good cable. I’ve used it to work the grippers in on an arm I built a long time ago.

I used long timing belts and pulleys driven by geared stepper motors in that arm. All METAL.

This is not my robot, but the Lab Volt Servo robot is done with the timing belts I’ve been talking about. Also notice the placement of the servos in the base, to reduce the weight in the actual arm.

theoldrobots.com/Lab-Volt-5100.html

Alan KM6VV

Ya i was thinking of using a spring and fishing line on the gripper, and a push solid metal rod (heavy solid wire) on the sholder (top servo) and placing the servo closer to the base.

It was wireless-bluetooth, but the bluetooth adapters both on my PC and the uart adapter was crap.
It might work better after i get a 5v->3.3v ttl converter and a 3.3v regulator.

Else i’ll get This

I agree that putting the elbow servo at the base of the upper arm and using stiff wire (big paperclips would do) to move the end of the forearm would be ideal. Below is part of my craft stick arm pieced back together to show how it was made.

http://web.comporium.net/~shb/pix/arm1.jpg

I’m thinking of using standard servos for the 2 base servos. If i do that, than it wouldn’t matter if i servos at the joints of the rest of the arm. I was thinking of a couple nice big tower-pro MG995’s which have a stall torque of 11kg/cm, at 6v ('ll be using 5v). Though any standard servo would be fine.
The tower pros are like $10 each so 2 of those wont exactly break the bank. as for my current arm, it’s relatively accurate, once i can get a terminal program with 14 macro buttons.
I saw xans rover with an arm, i like his idea of a toggled arm/controls, i think i’m going to do that, since it should be really easy to impliment.
I just got to find out how to add a toggle switch inside the Atom studio (shouldn’t take to long to figure out, though you guys can help)

Edit’ed out slightly messy toggle command.

Edit: i added my toggle command, works beautifully. Copy/pasted xans sound beep code too, so it beeps when i switch between arm and rover mode. Sadly my arm controls aren’t as smooth or as advance as xans, but neither is my arm.

Gotta find a way to better power the arm, these servos are making the BB2’s regulator a bit hot. Rover is to small to use multiple battery packs.
Damn where can i find a 5v 3-5A voltage regulator! I only need 2 of them.

That’s a cool way to do it. Looks like you need to keep the arm oriented so gravity is holding pressure on the line, but that’s ok.

Assuming you are still using the 10vdc battery you should be able to find a suitable regulator. We used one from Castle Creations that worked well. The Bot Board regulator is definitely not the way to go. It will cause problems with resetting the processor.

using an 8.4v battery pack.
I want to use my 9.6v battery pack or even my 10.8v battery pack, but the BB2 manual said 7.2v->9v so i don’t want to be using 5 regulators for 1 battery you know?

Can i stack voltage regulators?
If i use two 1A 5v voltage regulators in parallel will it create 2A?
What about a 1A and a 1.5A = 2.5A?

http://img291.imageshack.us/img291/4390/parallele.jpg

I could probably get my hands on a few more 1A-1.5A 5v regulators. I have a box full of broken stuff from burned out motherboards, computer power supplies, radios, all kinds of stuff. I won’t have a problem using multiple regulators in parallel if you can confirm it’s ok. I have a few random diodes (not sure what kind) if it would help.

Oh, 1 more thing, If i use a 5v regulator, in parallel with a 9v regulator, will i get 14v back?

I’m not aware of any 3 pin regulators that can be stacked for more power. Sorry

Dang, ok thanks. I’ll have to use another regulator separately then.

Well the second dedicated regulator works fine. The BB2’s regulator is still over heating.
Well maybe not overheating, but it’s hot to the touch. But all it’s powering is the BB2, the Atom, and a RS-232 to TTL converter.. I’m surprised it’s still getting hot.

I’m thinking of removing the BB2’s regulator, soldering on a nice custom heatsink, then putting it back.
Soldering is not one of those things i have any trouble with. Trust me, i’m good.

Well for my dedicated voltage regulators, i added a tiny 5v 0.2A fan. And i mean tiny! 3cm x 3cm x 1cm.
And i didn’t have a second heatsink, so i coiled up some heavy guage solid wire, and soldered it onto the second regulator, i have an idea how to improve upon that too.

Another $150 before this thing is really usable.

Edit:

If i hear the servo’s working, like trying to keep the arm steady, while the arm is fully extended, is that bad? Or are servos designed to work like that?

I want to try to pick up heavier and heavier things, but i don’t want to burn out a servo. (i have extras, but it means i’d have to un-hotglue a bunch of stuff and rebuild part of the arm from scratch)

Dang i wish i could afford those fancy brackets.

Actually the forearm is movement is controlled in both directions. The green bottle caps have a thin slit cut in them such that the slit faces the arm support wire opposite the shoulder servo horn, under which the floss runs. One piece of floss with a knot in the end is pushed into the slit on the servo horn cap, then rotated one turn around the cap counter clockwise, routed under the arm support wire, then up to the elbo cap, where it makes a counter clockwise wrap around that cap and into that cap’s slit. A second piece of floss is used the same way, except the floss is wraped around the caps clockwise. Now when the servo rotates it has positive control of the forearm in both directions. The gripper servo was mounted behind the forearm servo and also had a bottle cap and floss actuator. I used dental floss as it is very thin, slick, and strong with no stretching.

Radio Shack has heat sinks that fit the regulator chips. For inexpensive low budget tinkering, el cheapo servos are available like below. They fit in well with the affordable stick arms.

hobbypartz.com/kahaoubrmo14.html

hobbypartz.com/topromisesg9.html

hobbycity.com/hobbycity/stor … &sortlist=

hobbycity.com/hobbycity/stor … oduct=9549

web4robot.com/Flexel.html