Continuous rotation servo mods
#1
Thread Starter
My Feedback: (2)
Continuous rotation servo mods
Hi Guys,
http://www.seattlerobotics.org/guide/servohack.html
I need to modify another servo for use in a blimp's prize drop mechanism. Prior to this I'd use a servo with a mechanically disconnected pot, but the trim seems to be changing by several click flight to flight. Having accidentally dropped a few items at the wrong time, I need to address it for our next game.
I'm going to modify a 148 with the instructions in the above link, but one question comes to mind. The article states early on that a pair of 2.2K resistors are to be used in place of the pot, but one can go as high as a pair of 3.3k resistors so long as the values are equal... Not fully understanding the circuit's function here, will a pair of 3.3K resistors make the servo more or less sensitive, or have no effect? Ideally, I think I'm looking for a less sensitive servo, with a larger range of pulse widths that result in zero rotation.
Any thoughts?
Kelly
http://www.seattlerobotics.org/guide/servohack.html
I need to modify another servo for use in a blimp's prize drop mechanism. Prior to this I'd use a servo with a mechanically disconnected pot, but the trim seems to be changing by several click flight to flight. Having accidentally dropped a few items at the wrong time, I need to address it for our next game.
I'm going to modify a 148 with the instructions in the above link, but one question comes to mind. The article states early on that a pair of 2.2K resistors are to be used in place of the pot, but one can go as high as a pair of 3.3k resistors so long as the values are equal... Not fully understanding the circuit's function here, will a pair of 3.3K resistors make the servo more or less sensitive, or have no effect? Ideally, I think I'm looking for a less sensitive servo, with a larger range of pulse widths that result in zero rotation.
Any thoughts?
Kelly
#2
Senior Member
RE: Continuous rotation servo mods
The servos work by generating a pluse internally and matching it with the one sent from the receiver. The speed is determined by how fast the stick moves and how fast the motor responds. As soon as the pluses match, motion stops. The value of the pot replacement is not important as long as the two sides are equal, tricking the servo electronics to think the pot is centered.
Now, it just dawned on me, no matter where the arm is when you power on, that is the defacto center. Also as there is no feedback from position, there is no holding current, so you could manualy move the motor with power on and there would be no correction as there would be nothing to feed in the need for correction. That would mean that your drop could drift open long befor the signal was given to activate the servo.
The previous paragraph is just my theory based on how the servos work, I haven't tried it so it should be viewed as a theory that needs proving at this point. Something to think about.
I would guess that you could use something link a retract servo with more reliable results. Or a wench servo from a sailboat.
Don
Now, it just dawned on me, no matter where the arm is when you power on, that is the defacto center. Also as there is no feedback from position, there is no holding current, so you could manualy move the motor with power on and there would be no correction as there would be nothing to feed in the need for correction. That would mean that your drop could drift open long befor the signal was given to activate the servo.
The previous paragraph is just my theory based on how the servos work, I haven't tried it so it should be viewed as a theory that needs proving at this point. Something to think about.
I would guess that you could use something link a retract servo with more reliable results. Or a wench servo from a sailboat.
Don
#3
Thread Starter
My Feedback: (2)
RE: Continuous rotation servo mods
The prize drop works by hooking a long spring up to the servo arm, aligned with the rotational axis of the servo arm. Right aileron stick causes it to unwind and drop the prizes off the end, while left stick is reload for the next flight. Its remarkably simple but there really is something fishy going on with this one recently... Up close, I can easily trim the servo to sit sill. A few clicks in either direction causes it to rotate slowly. From a distance away in the arena it seems to take a new center position though, which causes it to unwind slowly at the old trim. The distance thing might just be coincidence though. The radio is a JR 7ch PCM, so I don't think the radio is at fault.
I'm stuck for time since the next game is tomorrow night, so I'll try one with the resistances recommended on the web page, and we'll see how it goes...
Thanks,
Kelly
I'm stuck for time since the next game is tomorrow night, so I'll try one with the resistances recommended on the web page, and we'll see how it goes...
Thanks,
Kelly
#4
Senior Member
RE: Continuous rotation servo mods
My theory may be proving right. Try just moving the servo arm by hand after triming it to center. See if that becomes your new center after it has moved. I think that because you are removing the positional feedback, IE the pot, and replacing it with two fixed resistors, wherever it is when it is powered on is it center. Your actual position at center will rely on a very stable pulse width from the transmitter, coupled with a very stable receiver decode. If you transmitter has expodential, try dialing in like 95% expo so the transmitter arm needs to move almost full stroke before there is any movement. See if that stablizes things.
Curious problem. Let us know how it works out. Also, how about some photos of the drop mechanics and the drop in action. Sounds like something fun to do.
Don
Curious problem. Let us know how it works out. Also, how about some photos of the drop mechanics and the drop in action. Sounds like something fun to do.
Don
#5
RE: Continuous rotation servo mods
sorry to but in, I would like to modify a servo like this also, but I have a few questions.
1-does the servo have to be a 148 or can it be a standard futaba s3003 or s3004
2-where would I be able to get resistors
Thanks in advance,
Austin
1-does the servo have to be a 148 or can it be a standard futaba s3003 or s3004
2-where would I be able to get resistors
Thanks in advance,
Austin
#7
Senior Member
RE: Continuous rotation servo mods
Measure the pot, outside pin to outside pin for the total resistance required if it is not stamped on the case. Your two resistors should be equal resistance and 1/2 the resistance of the pot. If you want the best performance, get 1% tollerance resistors from an Electronic supply store. THe radio shack ones will most likely be 20% which means that a 3K resistor coulld be anywhere from 3600 ohms to 2400 ohms. The 1% ones will be around $0.50 each so cost isn't a factor. The other option is to take a good Ohm meter in and select a matched pair. If they are mismatched in value, you will have to use a lot of trim to get the motor to stop turning.
Don
Don
#8
Senior Member
RE: Continuous rotation servo mods
How about using a dc motor to drive the drop mech and use a regular unmodified servo to switch the power to the motor Direction would be controled by the switch or switches one on each side of the servo arm The KISS logic
#9
Senior Member
RE: Continuous rotation servo mods
How about using a dc motor to drive the drop mech and use a regular unmodified servo to switch the power to the motor Direction would be controled by the switch or switches one on each side of the servo arm The KISS logic