|
mglavin -> RE: 4.8v last longer than 6.0v? Really? (8/12/2005 2:53:05 PM)
|
OK, I feel asleep reading as John suspected, BUT I woke up with these thoughts.
Our servos use permanent magnet dc motors.
While servo motors offer constant resistance by nature the application of our servos places them in dynamic load scenarios.
Torque is proportional to the current, more current equals more torque. Cored motored servos with like power rating to coreless motored servos will likely consume more power to perform like work. It’s a simple design consideration; coreless motors use a larger “lever” in short. The circumference of the coreless motor is essentially the “lever” while a poled motors lever is the armature or rotating mass within the magnetic field (coreless motors are inverse, the magnets are fixed and the core rotates around the magnets).
At rest, motor not rotating, current is equal to the voltage at the motor terminals Vt divided by the resistance R of the winding. As the motor starts to rotate it becomes a generator and produces a back emf (voltage) that has a polarity opposite the terminal voltage. The current then becomes Vt minus Vemf divided by R. The faster it rotates the greater the back emf and the lower the current. Therefore a motor at no load draws relatively low current.
As we load the motor it slows down, the back emf drops, the current goes up to create more torque to overcome the load. If we stall the servo, the back emf goes to zero, the current goes back to Vt divided by R and if you’re lucky you won’t burn up the servo. Digitals do not pull max current regardless of the load. As the load increases so does the need for more current.
The output circuit of the servo amp is essentially four switches in an H configuration. Two switches are closed to drive the motor one way and the other two to reverse it. This is the same for analog and digital servos. At large error signals the switches are closed continuously until the servo gets to the commanded position and then all switches turn off. A large error signal equates to large difference between the commanded position and the actual position, you’d experience this by moving the stick to an extreme. One could surmise that analog and digital servo amps driving the same motor will draw the same current.
The difference is realized with small error signals. Both Analog and digitals amps pulse the switches at small error signals in order to slow the servo down so it stops without overshooting. The so-called digitals pulse more often with shorter pulses but the net result is they are closed for a greater proportion of the time and therefore the average current is higher and there is more torque at small error signals. This accounts for increased holding power too.
The limit on how much current you can put into the motor is a function of the windings and we are getting close to the point where they will burn up if you put too much voltage on them which would in turn put too much current through them either at stall or during oscillations where it is continuously reversing itself and never builds up a back emf. So the plateau is really a cliff where it burns up if you increase the current.
In an ideal system design our servos would drive to the commanded position and it would null out and draw no current. Then it wouldn’t matter if you had 4.8 or 6 volts except it would get there faster with 6V. In the real world the servo might not get to the commanded position because as it gets close it starts pulsing, the current goes down and the torque goes down so it can’t overcome the load. If you have more voltage, 6V instead of 4.8V the current is higher, the torque is higher and you will get closer to the commanded position. Digitals will perform better because they have a higher average current under these conditions than analogs. So 6V will perform better under heavy loads but at 4.8V the servo will perform adequately.
The problem I realized with today’s servos is we don’t really know what the voltage is at the servos due to all the drops due to high current through inadequate connectors, switches and ancillary devices. Many modelers are seeing less than 4V with a 5 cell packs under extreme conditions like starting six servos in a snap roll.
Theoretically, if the efficiency doesn't change, it should take the same energy to do the same amount of work. Power is the rate at which energy is converted to work. In our cars the efficiency goes down at high power so we use more gas to move a given distance. We could get to grandmas with one horsepower if we had enough time. If we had a motor rotating with a given load, like a prop, the 6V system would turn the prop faster. So when the servo is moving it is moving faster. It is not so clear to me when we are talking about holding, no movement, against a force. No work is getting done but it is consuming energy. Let's assume we have servos with enough power to do the job. That means they will never be completely stalled during flight. However, as mentioned, they might not always get exactly to the commanded position so there will be some error. The higher voltage will produce more current, providing more torque, so it will drive closer to the commanded position. When it gets closer the pulses get shorter, and the current goes down. I believe it takes less average current to hold the same load with 6V’s. I believe many assume our servos offer a constant resistive load and divide the voltage by the same R and conclude you'd use more current with higher voltage. This may be a reasonable assumption because so much of our capacity is used with the servos essentially at null but with vibration and buffeting that make them oscillate. The higher voltage system will draw more current under these conditions and may even contribute to the oscillations if the servos are optimized for lower voltages.
A quick test @ 4.8V/6.0V validates the current is the same to hold a given force. The difference is the five cell pack will achieve a greater maximum force. It gets back to the basics I guess, more current, more torque.
|
|
|
|
