Hi all,
me and a colleague are trying to make a 6DOF robot arm. So far so good, we have the electronics sorted out, the designs printed and assembled but the servos are vibrating when we execute "counter-gravity" movements, for example on the third axis: moving the axis towards the ground works like a charm, but raising the arm causes the servo to start vibrating when we stop the movement, the same type of vibrations you get when you power the motors and try to move them by hand. It's kinda like the inertia it's causing it but this is just my educated guess.
It sounds like you're overloading the servos. It may be the power supply is not adequate or it may be simply that what you are trying to get them hold up is too heavy for them. Both servos are cheap clones rather than original TowerPro servos. Unfortunately there are many very poor quality clones around.
How are we supposed to compare that rating (which I didn't see) to anything, without knowing anything about your robot?
The Universal Robot (UR3) that we have at work does not use wimpy servos to move the joints. It uses some really beefy ones - 10 times as large as yours - but still can't carry much weight. While it is aluminum, and your apparently is plastic, there is still far more torque available from the UR3's servos than from your servos.
groundFungus:
Is your power supply adequate to drive all 6 servos. Even when standing still the servos could be drawing considerable current to maintain position.
The power supply is a 300w mATX i got from an old PC, should i try a bigger one?
PaulS:
Where did you see a torque rating?
The Amazon page says 12kg/cm at 6 volt. Considering the longest arm is 30 cm, the resulting torque should be 12/30 = 0.4 kg, right? We haven't weighted our prints but considering that the ABS is 1.07 g/cm³ we didn't do further calculations and tried. We're considering making a less heavier prints though
Then, how did you decide how much torque you needed?
You understand that the torque needed to move an arm requires moving the arm AND everything outboard of the arm - the other arms, the joints, the servos, etc. - right?
The power supply is a 300w mATX i got from an old PC, should i try a bigger one?
Are you powering the servos off the 5V rail? What is the current rating for 5V? Below is the data sheet for the bigger servos. It lists the operating and stall current for them. Does the 5V have enough current to start all 4 (or 6) servos at once?
PaulS:
Then, how did you decide how much torque you needed?
You understand that the torque needed to move an arm requires moving the arm AND everything outboard of the arm - the other arms, the joints, the servos, etc. - right?
Of course i do, maybe we were too optimistic... I also couldn't afford to drop several hundreds of cash on the motors, unluckily
groundFungus:
Are you powering the servos off the 5V rail? What is the current rating for 5V? Below is the data sheet for the bigger servos. It lists the operating and stall current for them. Does the 5V have enough current to start all 4 (or 6) servos at once?
And on 5V you only have about 9kgf.cm torque to play with (assuming those servos are up to the TowerPro spec). And bear in mind that's the STALL torque. That will burn out the servos if it's sustained for any length of time. Practical long-term operating torque is generally around half that value.
slipstick:
And on 5V you only have about 9kgf.cm torque to play with (assuming those servos are up to the TowerPro spec). And bear in mind that's the STALL torque. That will burn out the servos if it's sustained for any length of time. Practical long-term operating torque is generally around half that value.
Steve
I get it, we haven't considered that. What should I do?
