Boy…
And here I thought that I was addicted to commas!
I’m assuming that you’re asking why we couldn’t just use any old pot, eh?
Well, the answer to that would be price and ease.
It’s a lot easier to rip the pots out of the servo as well, since we’ll be needing the control boards, anyhow.
The H-bridge is just the set of transistors (or FETs) that turns on and off the motor in a given direction.
It’s accuracy isn’t really questioned.
The accuracy of the PID controller portion of the PID is what one needs to worry about, as well as the linearity of the potentiometer.
The potentiometer is almost always attached “directly to the joint”.
In the best case scenario, it’s attached with a 1:1 ratio set of gears to the final output shaft.
In the worst case scenario, the potentiometer itself is the output shaft.
(Really bad for durability.)
Uh…
I don’t understand.
A servo has an internal pot.
There isn’t an external one.
wht i was asking about H-bridge was with a pot can it be made to move to a certain position and hold, can it become a servo ?? thts sounds better,if not wht else would u need and why?
A Period tends to help at the end of your sentences so we can see where one thought ends and another begins…
You seem to have asked whether a potentiometer and an H-bridge can make a servo hold a certain position.
Nope.
A potentiometer is a tool that is used, in this instance, to tell where the servo is.
An H-Bridge can be used to control the direction and speed of a motor.
However, alone, neither would function.
A servo is a closed-loop device.
You tell it where to go and it goes there and monitors itself, correcting at any time that it finds itself not where it was told to be.
The brains behind that is the PID (proportional integral derivative) controller.
A remote microcontroller of some sort communicates the desired position to the PID controller in the servo.
The PID controller reads the value off of the potentiometer to determine where the servo is.
The PID controller then uses the H-bridge to get the servo to the new position.
Allong the way, the PID controller keeps asking the potentiometer, “are we there yet?”
You could control a servo with an external potentiometer and a 555 timer to do the pulse width modulation, but that’s not got anything to do with the internals of a servo, which needs to have a PID controller, an H-bridge, and a potentiometer.
That’s the idea.
Although you need a “PID controller”, not a “PID”.
“PID” is just a software tuning method used to get a device’s output to a desired state (a desired position, in our case) as efficiently as possible.
Ideally, one wants the output to be in the necessary state as soon as it’s told to be.
This isn’t possible, though, in our case, because a motor moved too fast will miss the target.
So, a PID algorithm is just a way to balance speed vs. accuracy, in our case.
A “PID controller” is just a piece of hardware that physically employs the PID algorithm.
In a digital servo, it’s a microcontroller.
In an analog servo…
um…
::scratches head::
I don’t know what does it in an analog servo.
I haven’t gotten the chance to open one up, yet.
As far as I can tell (the page isn’t very descriptive) those are ordinary permanent magnet motors.
If so, then they’d theoretically work.
If you had a powerful enough H-Bridge, then you’d be able to make a servo out of it with a PID controller and a feedback pot.
A lot of large motors (such as some car starter motors) are series/shunt wound motors.
In other words, they have electromagnets (coils) inside instead of permanent magnets.
If that motor is one, then no, I don’t think it would be so simple.
By the way… Jim/zoom are we talking about MOSFETs or JFETS?
As far as I know, MOSFETS would be hard to interface, since they’re voltage-controlled (bipolars are current-controlled).
Maybe those logic-level ones would work…
In any case, I believe that there’s a large amount of current even for thos to turn them on and off quickly.
Maybe the answer to that would be to use those transistors to drive the MOSFETs…
I haven’t yet researched JFETs, though…
That’s on my To Learn list.
Thanks nick
thts wht i needed to know . now to do some leg work .
ohhh and i think wht zoom was talking about was like removin the motor from a servo and some how running the new (gearhead)shaft all the way through . so as to use the pot and pid but not the outputgear or shaft,am i right zoom ??
slick idea,problem ! outputshaft =180 degrees=10 turns of the pot.
how many turns does the motorshaft turn . i think u might get away with
using the same gear ratio and placing it(the reduction gear) after it comes out of the servo.
I’m currently working on the inverse kinematics problem for Arm 5 (turret, shoulder, elbow, wrist, wrist rotation) and I’ve been running into a bit of trouble. If I may ask (and this question can go to anyone else as well), what method/algorithm did you use?
The Open Servo folks are doing just that, but they are actually replacing the electronics of a servo with electronics and software of their own design. I am hoping to use at least some Open Servos
[