I’m sorry if these are real basic questions… I’ve just set up my L6 arm with the RIOS SSC-32 V1.03 and done the preliminary testing.
Is there a preferred routing method for the cables to allow the arm free movement but not getting them all tangled up?
After doing the Step 7 through Step 12 setup as stated in the RIOS manual, the zero points are not in center. Is there a way to readjust the zero to center? For instance, after adjusting the base so it rotates a full 90 left and a full 90 right, and then sliding the slider to a zero value in the ‘Pos Deg’ window, the base is facing about 5 degrees to the right. This is similar for the other servos.
Is it normal for the arm to shake/stutter when slewing to a position? Is there a way to reduce/eliminate this?
I’m still working on the project, so nothing’s finalized yet, but as of now, I’ve got mine running up the forearm, then down along the side of the shoulder. I’m using some stiff wire attached to the assembly screws as standoffs/looms.
As to keeping everything routed: Nylon cable ties. Lots of them. They’re cheap and plentiful, and available in a wide variety of snazzy colors to complement those of your arm.
With power removed from the servos, I start at the gripper and work through the full range of motion to see where I’ll need the most wire, then I add a bit. Always leave a bit of a loop when you’re routing servo cables; if you constrain them to always making a sharp bend in the same spot, they’ll fail on you (go ahead - ask me how I know this). I’ve got mine bundled into a single “umbilical” type of harness, with the individual servos’ cables breaking out of the bundle wherever they pass close to their servos. Just add a zip-tie where you want it to leave the bundle, and everything stays neat.
Servo wires are usually about the same length, and they’ll have runs of different lengths to get back to your controller, but there are a number of solutions to this:
You could use pre-made extensions that just plug into each other. The extensions tend to come in standard lengths as well, so you’ll have some extra length to bundle up somewhere along the way.
You could make your own custom-length extensions and plug or solder them to the ends of your servo wires. That way, you’ll have plugs that end when they’re supposed to. It will probably look neater than the above, and you won’t have all that extra wire bundled up at your controller.
You could pick up a spool of servo wire and replace the cabling on your servos with appropriately-sized runs, terminating at your controller board without extra length. This would probably yield the cleanest-looking results, as you wouldn’t have little bits of heat-shrink tubing and/or plugs all up and down your wiring where the splices are.
I’m currently using option #1, as it’s still in the works. Once I get the specifics nailed down, I’ve got a spool of leftover servo wire that’s just waiting to clean up the runs.
As far as I know (and I have a pretty decent familiarity with servos) there’s no way to mechanically adjust the 0 point to the exact center.
You’ll have to rely on software for that.
You’ll have to change the “offset value” of each servo to have a 1.5ms pulse go to the center position.
I believe that the Visual Sequencer has this ability, but I’m not sure about the RIOS software, since I don’t have an arm.
(A robotic one, that is.)
I’m not sure, but I believe that changing the offset value hampers the rotation of the servo.
Let’s say that you have a 180 degree servo that’s 5 degrees off to the right.
Thus, your offset value would be -5 degrees.
I believe that the offset value simply tell the program to add or subtract the offset number (in microseconds, not degrees) from whatever your position that you tell it to go to.
In essence, to the right, you’ll be able to rotate 85 degrees, and to the left, you’ll be able to rotate 95 degrees.
With luck, I haven’t messed up in my explanation, or (as usual) just outright lied, yet…
Yes, it’s normal…
But, not for the reason’s that you suspect…
It’s normal for over-burdened, under-paid servos to shake/stutter/overheat and die.
Shaking/stuttering results from weak servos.
The best, and – in my opinion – only way to remedy this, is to buy stronger servos.
There are alternate methods, but I wouldn’t recommend them.
You can try spring setups to reduce the strain on them, but they tend to restrict the full motion of the servo.
You’re welcome to try this and any other options that you can come up with, but I’d advise you to skip straight to getting better servos (at least for the problem joints), and save yourself a lot of hassle.
I had this problem in the base servo of my Lynx 6. If the base servo #0 is shaking or jerky then the whole unit shakes. I replaced the stock base with the Metal Arm Base Price $49.95] Model Number [MAB-01]. I replaced the base servo HS-422 with a HS-645MG Price $39.99] Model Number [S4-01] which has about twice the torque. Everything seems to be working great now.
Shaking and stuttering, I have found that moving a Servo Past the desired point and returning a couple of degrees will stop the stuttering, unless of course it is the weight of the arm causing it, in which case balancing springs and weights would help.
Thanks, everyone for the replies. I have been able to successfully route the cables without causing any motion hindrance. It looks like the best solution for the weak servo/stuttering problem is to get better servos.
We should try and make sure it’s not something else. Does the stuttering happen only when it’s moving, or even when it’s standing still? Are you using a regulated supply? It should be 6vdc and at least 2 amps. If the supply is week, or defective it can cause the servos to act funny. If you are powering the SSC-32 logic from the same servo supply it is possible to have the processor brown out or reset. This can be fixed by powering the logic from a separate 9vdc battery and removing the VS=VL jumper. If the stuttering only happens when it is moving, there are adjustments that can be made within RIOS that can fix this.
