Modify a Futaba S3001 servo for continuous rotation

In this instructable I well show you how to modify a Futaba S3001 dual ball bearing servo for continuous rotation. Why you might ask, you can get already modified servos from Parralax? Two reasons, one I like to tinker with things and two my local hobby shop had a box of these for $15 each so I probably saved a couple of dollars by not having to pay for shipping.

Step 1 Materials and tools

Materials:
1 - Futaba S3001 servo

Tools:
- #1 Phillips screwdriver
- Dremal with a cutoff wheel
- Needle nose pliers

Comments

[Hoolaput]

awesome this is one of the simplest servo mods i've seen one question does it work for all servos???

[Ohm]

The principal is the same but each servo is different so the stop mechanism well be different for each.

[Hoolaput]

how much does these things normally cost?

[Gelfling6]

Standard servos, anywhere from $12.95 (USD) to $50.95 (USD) (depends on gearing, torque, size.) I've gotten Futaba S3004's (standard torque, ball-bearing) for $14.95 (USD) at a local hobby store. Parallax charges $12.99 (USD) each, equivilent to the S3001's, potentiometers are changed to a small, PC mount 270-degree turn 5K potentiometer, which you can adjust with a small flat-blade screwdriver through a hole drilled through the side.

[mrgreencool]

I Don't think it will work with servo mg995 (steel gear), i did remove the blocker...but once i remove it, when it turn the rheostat in the servo will follow the cycle and break the wire...and walaaaa i need to buy a new one!!!...-_- ....rheostat can move 180 deg only...

[Gelfling6]

As others have said, you need to remove the connection between the final drive gear, and the potentiometer (rheostat, refers to wire-wound, mainly for AC voltage, but you've got the right idea.) This will entail removing the potentiometer completely, or the inner workings, so it can rotate without hitting stops. (and replacing the wiring to it, with a pair of 2.2K resistors, in series, but the center connection to the 'wiper' wire for the pot.

[manmelvin]

Can the servo direction be reversed? I just did it with one of mine and it spins fine one way but just buzzes when I try and reverse the direction.

[Gelfling6]

As the other person below mentioned, One missing step. The potentiometer (under the final drive gear) needs to be replaced with a pair of 2.2K resistors to simulate the potentiometer being centered. I've modified Futaba 3004's (same as the 3001's, but ball-bearing on the final drive gear.), by "S" bending the leads to the potentiometer, and sticking it out the side of the case, so I can tweak it by hand. It sounds like, if you didn't replace or center the potentiometers, they're still in one direction (not at center).. The dremel is a bit overkill.. Most of the time, a pair of diagonal cutters and a file (or careful whittling with a utility knife) to get rid of the stop point.

[dwalton]

Nice job on the pictures! This is very well done.

Unfortunately there is one more step to creating a continuous rotation servo. You need to hold the potentiometer in a centered position or replace it with a resistor divider. I learned this the hard way. If the potentiometer is not centered, the servo will turn with a 1.5 ms pulse when it should be still and will turn faster in one direction than the other.  If the potentiometer is turned too far, your servo will only turn in one direction, no matter how wide the input pulse is.

The Parallax continuous rotation servos have moved the potentiometer so you can access it from the side of the servo without having to disassemble it. This allows you to send a 1.5 ms pulse and adjust the potentiometer with a screwdriver so that the servo is still.

I think there is a simple solution to the problem.  I have been able to drive a 1.5 ms pulse to the servo with the gears exposed (like in step 3) and turn the potentiometer by hand until the servo was still.  I then reassembled the potentiometer and tested to make sure the servo was still with a 1.5 ms pulse.  There is still a chance the potentiometer could get vibrated or bumped into a non-centered position, but so far it has worked.