darkfeffy
Electrical
- Jul 24, 2009
- 1
Hi all,
I am faced with an interesting transfer function as follows:
Gx(s) = C(s)*C_PID(s)*G(s)
Where:
(NB: "ex" represents 10^x)
C = 1/((5e-5)*s +1);
C_PID = Kp + Ki/s + Kd*s/(Tf*s+1) with Kp = 0.21205; Ki = 62.3191; Kd = 1.0773e-5; Tf = 2.9319e-5.
G = (-342.1 s^3 - 3.713e10 s^2 + 5.446e15 s + 6.906e17)/(s^4 + 1288 s^3 + 3.733e10 s^2 + 6.275e12 s + 2.352e16)
Using the bode function in MATLAB shows that the open loop transfer function has multiple zero crossings. So, how does MATLAB decide which of the zero crossings should correspond to the phase margin? I ask because by tweaking C and C_PID slightly, similar plots are obtained but with a different zero crossing chosen (by MATLAB) for the phase margin calculation. Any help here?
Thanks
I am faced with an interesting transfer function as follows:
Gx(s) = C(s)*C_PID(s)*G(s)
Where:
(NB: "ex" represents 10^x)
C = 1/((5e-5)*s +1);
C_PID = Kp + Ki/s + Kd*s/(Tf*s+1) with Kp = 0.21205; Ki = 62.3191; Kd = 1.0773e-5; Tf = 2.9319e-5.
G = (-342.1 s^3 - 3.713e10 s^2 + 5.446e15 s + 6.906e17)/(s^4 + 1288 s^3 + 3.733e10 s^2 + 6.275e12 s + 2.352e16)
Using the bode function in MATLAB shows that the open loop transfer function has multiple zero crossings. So, how does MATLAB decide which of the zero crossings should correspond to the phase margin? I ask because by tweaking C and C_PID slightly, similar plots are obtained but with a different zero crossing chosen (by MATLAB) for the phase margin calculation. Any help here?
Thanks