Positional Servo Control [DNO / VTX]

Notes on how to create a positional servo control.

There are two types of servo system: speed and position. These are explained in our Answers to FAQs on Battery Motors & Controllers. There is also an explanation, in our sister site 4QD-TEC of Pulse Width Position Servo as it is used in radio control systems.

This application note explains how a simple reversing controller might be used as a building block in a positional servo system in, for instance, a power assist steering system.

This is not an application we ourselves have implemented, so these notes are theoretical only.

Theory

A servo system is a closed loop – one input (demand) tells the system what the output should do and another input (response) senses the systems output. By comparing ‘demand’ with ‘response’ and error signal is given which corrects the system until the response equals the demand.

The theory behind such feedback systems is extensive and if you wish to be baffled by complex mathematical analyses, there are many books available which will do just that!

In a position servo, the response signal must sense the mechanical position of the output. Most positional servos are sensing angle of rotation: this applies to radio control servos, aerial rotators (a common use of servo systems) and to power assisted steering systems. In a rotational system, a rotary potentiometer is the easiest sensor to use.

Circuit

The standard DNO / VTX series controllers are simple reversing controllers well suited to ‘building block’ applications. They have a speed input and a direction input. For a servo system such as this, a single input is required which will control both speed and direction. The Joystick interfaces do this. The circuit below shows how you may modify our standard single axis interface daughter board (JSD-001). The interface circuits are given and explained in the Joystick interface circuits on the 4QD-TEC site. servo control

R1 and R2 apply half supply voltage to pin 5 of the IC and at the same time work with R5 to define the input gain as about 0.5. P1 is the ‘demand’ pot. Position is sensed by P2 and fed into pin 6: R2 and R4 define the gain of the stage as unity for the input P2 and 2 for input from P1. So the .5 gain defined by R5 with R1 and R2 is multiplied by 2 to give an overall approximately unity gain.

If there is any error signal between P1 and P2, then it causes an output which is converted by the rest of the circuit (as described in the page on the joystick circuits) into a signal to work the controller.

This is fairly crude circuit. However – the reponse of any electro-mechanical system such as this is as much a matter of mechanical parameters (motor drive gear ratio, stiction etc.) as it is of electrical performance and specifying mechanisms is well beyond the scope of this page. It should however illustrate the principle.

The circuit above is that of our Single-axis Joystick interface, Daughter version.


Other relevant pages

  • Servos in Answers to FAQs on Battery Motors & Controllers
  • Pulse Width Position Servo in 4QD-TEC circuits archive.
  • Joystick interfaces