modeling tracking servo

[garimella]

I was trying to simulate the simplified model of RDC (AD2S90). The issue is to design proper compensator for this model to track constant velocity with zero velocity lag. As suggested in the data sheet (page No.9) I tried to add a compensator (K/s * (s+a)/(s+b)) between demodulator and the VCO. But from simulation(step response), it appears to oscillate. Where am I going wrong? Almost all the data sheets that I came across recommend the compensator with the above mentioned transfer function, but in simulation, it does not work. How to proceed? When I provide velocity input, surprisingly the circuit works, gives no velocity lag forever. Reference signal is 50Hz

er!

 

[Pommie]

I know electronics, programming and general life, but have no idea what you're asking. Please rephrase.

Mike.

 

[alec_t]

The issue is to design proper compensator for this model to track constant velocity with zero velocity lag.

That's got me puzzled too.
Is 'servo' synonymous with 'resolver' (which the AD2S90 handles)?
If the velocity is constant, why do you need to track it?
According to the AD2S90 datasheet, the chip spits out a 12-bit serial number at a maximum bit-rate of 2MHz, i.e it takes at least 6us to give a reading. So how can you have 'zero velocity lag'?

 

[garimella]

RDC data sheet claims that for steady state and for constant velocity input the chip is capable of providing zero position/velocity lag. This is the characteristics of type2 servo loop. Stabilization is provided for the double integrator plant in the form of compensation. From my simulink model it appears to me that there is a positive feedback which is causing oscillations. Trying to figure out where it is and required help to fix it