I have a 100W 9VDC supply that I want to power two HSR 5980SG’s with. Is 9V too much; will it fry the motor or cause major malfunctions?
I have looked at adjustable voltage regulators if this is not an option. These servo motors will be required to reach stalling torque, so we’re talking about a lot of current draw. I also need all the torque I can get, so I would aim for the 7.4V mark (stalling @5.2A according to the spec sheet).
Sorry if I’m not being clear enough. Let me know if theres anything I need to elaborate on.
9vdc is way too much. I’m sure the servo will not survive that. I do not recommend designing a device on the stall torque value. A servo at stall does draw a lot of current, but this also means it will heat up and quickly. There are voltage regulators that can deliver 5 amps. lynxmotion.com/Product.aspx? … egoryID=48
Good luck with the project.
Thanks. About my stalling torque comment, these will be used on a 2-joint horizontal (planar motion) robotic arm for gripping objects against a wall and applying a specified force on them. I am using sensors on each link to measure the force. How would you suggest accomplishing force control with servo motors?
My thinking on the subject was…Once the arm contacts the object and presses it against the wall, it will continue to tell the motors to rotate, applying torque and hence force on the object. Once the required force is reached, the motors would cut off, maintaining the force due to “holding torque”. This would not require large currents for a long timespan and hopefully not overheat things. Does this make any sense, am I going about it all wrong?
Not sure exactly what you are wanting to do, but I think I got the gist of it. I would use a single pressure sensor such as an FSR. lynxmotion.com/Product.aspx? … tegoryID=8
And install it where the object is touched. This would be easier to implement and would probably be more reliable than incorporating sensors into the mechanical linkage.
As for the arm itself. Perhaps a single lever with the servo pushing on it to make a mechanical advantage. This would enable a lot more pressure to be applied by the servo. It may limit the range of motion though.
I would think if you try using anything and a servo, you are likely to get hysteresis.
The servo presses the FSR against the object
The FSR registers the ‘correct’ force, causing the servo to ‘creep’ back
The object pushes back* on the FSR / arm, eventually pushing back on the arm enough to cause the FSR to re-trigger the servo to ‘creep’ forward again.
Goto 1.
I am assuming the object is elastic / flexible somehow, because there is no logical reason you would push something in-elastic.
I would think you want to design around a stall speed, just not the max stall speed. I believe you can use a power transistor to deliver a specified amount of current by adjusting the voltage on one of the legs. (I can never remember which one is which)
In this case, you use a variable pot or something, and feed a motor (either in the servo or using a tamiya gear box or something), and then you adjust the pot, hence adjusting the amps the motor is drawing, hence adjusting how much power the motor can apply. After the circuit is adjusted, you can either glue to the pot or replace it with the closest value 1% or something, and then everytime you command the motor on using that gate, it will excert your known force continuously.
I understand your concern, but I haven’t experienced this with our Little Gripper / FSR combination used with RIOS. It’s a closed loop constant force servo driven solution. However due to the large forces frumaster is wanting maybe a gear motor would work better. Dunno…
So you’re saying I can control the amount of torque (subsequently force) by canging the supply voltage? I really need to apply a range of forces, not just one specifically. What about this…the arm first contacts an object against a surface and is instructed to move past the point of contact. The servo will continually apply torque to try and reach the unreachable position. That torque can be controlled by controlling supply voltage and you can reach any desired force under stalling torque.
Would that work? Of course then the problem is controlling the supply voltage (by computer program). Can you PWM a set supply voltage to give a lower average voltage, thereby reducing the power delivered?
You may want to look at using something like a linear actuator with a pressure adjustable limit switch to stop it at the desired pressure. It would be a simple mechanical setup operated by an h-bridge. When the limit switch opens, the linear actuator stops moving, requiring no holding current while still applying the desired force.
Oh, I believe you! I am basing this on my (limited) experience with the 5990TG. Because the 5980SG is similar (I think?) to the 5990TG, I am assuming that the 5980SG will have the same jittering problems that the 5990TG has.
No, I was saying you should be able to control the torque of a motor by controlling the amperage it can draw.
Again, you want to control the amperage, not the voltage. The voltage should be whatever the internal logic to the servo can handle safely, say 6V. The transistor on the power line will control how much amperage that the servo can draw. You could control the transistor by using a digital potentiometer or a DAC.
And even after you have current limited your servo, you have to make sure that you are not stressing the servo. I don’t believe that the 5980SG has any safeties built into the controller, which means if you use it too long at too high current you could risk damaging it. As I believe others have said, motors aren’t designed to operate at stall torque for more then ‘peak’ (very short) durations.
The 5980SG should have the latest version of the HMI software which detects if the motor is at high current for too long and cuts off the power. The servo has to be powered off to reset. Current in this case is the applied current and is actually based on the proportion of on time of the motor current and is not really measured. Holding the servo off target position in stall will trigger this. 10 degrees off position will give maximum torque(current) on a 5990, so I guess the 5890 is similar. There is a dead band of about 1 degree, and the torque appears to increase linearly between these points.
Operation of the servo at 9V would give 1.5 times the power dissapation in the motor before it cut out. Not a good idea on expensive servos.
If you want to play around with torque and current I would agree the gearhead servo is more appropriate. The motors tend to be larger and able to dissapate more power. Maybe with an openservo driver. You can provide the motor and processor supply seperate.
Controlling the voltage to the 5980SG digital servo is not recommended, since the same supply is used for motor and for the processor. Same goes for controlling the current, since the voltage will also drop when the current limit is reached.
Is this really true? The servomotor will apply different amounts of torque depending on the “angle off” (from about 1-10 degrees)? So would 1-10 degrees not be considered stalling? When would the motor cut off, at past 10 degrees? Would you recommend using this to control the motor torque? It would be in a feedback loop with FSRs.
What about using the servos arms with spring loaded bumpers? This would allow the servomotor to reach its position while the springs apply force as the bumpers are compressed. The servos would not have to go past the object, they would just have to hold position.
Would there even be a difference as far as the servos are concerned in the first option (brumper without springs against FSRs, told to rotate past contact point) and the second (bumpers with springs pressed against FSRs, position reached)? Torque is torque, right? Won’t either way draw the same amount of current with torque being equal? Or is the difference in operating procedure significant for the servomotors. How does the lynxmotion gripper do it?
Sorry, I have lots of questions but I will really appreciate the help. Last one I promise…How exact can you get the positioning of a servomotor? For instance, if I have a 40% duty cycle PWM control signal, then change it to 40.0001%, will the servomotor respond accordingly?
The gripper uses RIOS to monitor the pressure applied to the FSR while closing. When that amount reaches a predefined amount the servo is stopped right where it is. Normally the “load” button is clicked to lock the step. The condition for the step has been reached so no further attention is given to it. This means you can remove the object and the gripper will not change position.
The last question, you are using the wrong terminology or mindset. Think in terms of uS not thousandths of degrees. These servos respond to pulses in the range from 600 to 2400uS where 1500 is centered. The servo has a deadband which is listed in us. So if a servo has a deadband of 3uS it will not respond to changes that are within 3uS of the current position. I think the 5990 has a 1uS deadband. BTW the deadband could be looked at in degrees if you really wanted, but it’s just adding to the math. 2400-600=1800 180°/1800=0.1° So the deadband could be looked at as a tenth of a degree.
That torque can be controlled by controlling supply voltage and you can reach any desired force under stalling torque.
These servos respond to pulses in the range from 600 to 2400uS where 1500 is centered. The servo has a deadband which is listed in us. So if a servo has a deadband of 3uS it will not respond to changes that are within 3uS of the current position. I think the 5990 has a 1uS deadband. BTW the deadband could be looked at in degrees if you really wanted, but it’s just adding to the math. 2400-600=1800 180°/1800=0.1° So the deadband could be looked at as a tenth of a degree.