1
\$\begingroup\$

I am looking at trying to position control a voice coil actuator (something like these). On the end of the voice coil actuator will be a small load (let's just imagine a 10 oz cube), screwed onto the end of the shaft.

Here is where I come from. When position controlling a spring-return solenoid with a feedback sensor, it's relatively simple: to move the solenoid out to position X, you control the power based on the feedback signal's value. The more power you input, the further out the shaft moves.

I am wondering, how does one actually conduct position control on a voice coil actuator? In other words, to hold the shaft static at position X, does it require a constant voltage/current? Or, can you remove the input voltage/current, and have the shaft not move?

\$\endgroup\$
  • \$\begingroup\$ That depends if there are external forces or damping; for example levitating something against gravity will require continued actuation; doing it without any physical contact friction to provide damping makes it a moderately tricky feedback control problem. That version at least gets written up often... \$\endgroup\$ – Chris Stratton Jul 9 at 21:22
  • \$\begingroup\$ More power means the shaft moves more quickly. More energy means the shaft extends further. \$\endgroup\$ – Andy aka Jul 10 at 9:14
2
\$\begingroup\$

I am wondering, how does one actually conduct position control on a voice coil actuator? In other words, to hold the shaft static at position X, does it require a constant voltage/current?

A constant voltage or current will hold the shaft at a constant position provided the load, angle, and temperature don't change. To control in that fashion is generally called "open loop" and can work well enough for some applications.

For precise position control some form of position feedback is required to adjust the output so that the position error is reduced. There is no feedback built into the units in your link.

Or, can you remove the input voltage/current, and have the shaft not move?

There is no mention of a shaft brake on the datasheet either. The brake, if supplied, would require additional wiring as you couldn't expect the brake to release on low voltage operation (for a small movement).


Response to comment:

If a brake were to be included, what would be the brake's technical name?

I'd be looking for "holding brake" or similar.

Would it be a linear magnetic brake?

No. Linear magnetic brakes apply a decelerating force proportional to the velocity of a conductive plate through a slot in a magnet. At zero velocity the have no holding force. You'll need a mechanical holding brake.

Also, to interpret what you're saying in the 2nd half of your post... If the brake were added, it would prevent motion at low voltages because the power input to the VCA would have to be enough to generate a magnetic field that could overcome the magnetic field from a brake

Let's say we had a 12 V actuator with a 12 V brake. The brake requires some voltage - say 8 V - to release it fully. If the brake is sharing the actuator wires then the actuator won't move well or at all until the voltage reaches 8 V.

\$\endgroup\$
  • \$\begingroup\$ if a brake were to be included, what would be the brake's technical name? Would it be a linear magnetic brake? Also, to interpret what you're saying in the 2nd half of your post... If the brake were added, it would prevent motion at low voltages because the power input to the VCA would have to be enough to generate a magnetic field that could overcome the magnetic field from a brake? \$\endgroup\$ – Intrastellar Explorer Jul 13 at 20:45
  • \$\begingroup\$ See the update. \$\endgroup\$ – Transistor Jul 14 at 9:03
2
\$\begingroup\$

There are many ways to control a voice coil positioner, here are some of the analog systems:

  1. You can connect the shaft to a linear differential transformer (LVDT). This is probably the simplest high accuracy analog position control you can get at reasonable cost.
  2. You can connect the shaft to a torque bar with a strain gauge or piezoelectric sensor attached. Neither is easy to interface.
  3. Potentiometric linear sensor. I've seen many DIY implementations of this, and they are quite easy to implement. Probably the cheapest implementation you can do.
  4. You can get magnetic and capacitive linear position sensors, but they are problematic to design with. This article explains some of the choices.

All of the analog systems are easily included in an analog feedback loop for closed loop position control.

For digital sensors the choices are more difficult and typically more costly, except for AMS magnetic sensors I've used. Probably for your application you could use the AS5306 with Quad outputs that would easily accommodate a closed loop position system with an MCU maintaining the position counts.

\$\endgroup\$
  • 1
    \$\begingroup\$ thank you for sharing all of those informational articles!! I will be returning to this answer for years when I want to remember how to position sensing \$\endgroup\$ – Intrastellar Explorer Jul 13 at 22:18
1
\$\begingroup\$

A voice coil actuator alone simply produces a force that is proportional to the current passing through the coil (with a bit of friction in the bearings). Ideally that force is independent of the position of the shaft, but in practice it will vary a bit throughout the stroke.

To move it to a given position you need to measure the position by some means and use a controller and output amplifier to provide an appropriate current, resulting in a force, to move and maintain the position against whatever external forces and friction that may be present.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.