I met a problem on using the piezoelectric actuator to actively stabilize the homodyne interferometry to avoid the signal fading problem and to eliminate the experimental low frequency noises. The mirror is attached to the PZT actuator to reflect the reference beam. A Proportional-Integral circuit was designed , together with the PZT and homodyne interferometry, to constitute the feedback loop. I doubt the following problems: 1.Does the hysterisis or linearity and creep effects of the PZT influence the stabilization? Should these effects be seperatedly compensated or not, because the PZT is part of the closed-feedback-loop. 2. Does the response time of the PZT and the circuit influence the stabilization? 3. I cannot stabilize the homdyne interferometry well, what other factors may contribute to the unlocking of stabilization ? I do really appreciate your answers.
2. Does the response time of the PZT and the circuit influence the stabilization?
That's true! The matching circuit should be designed well.
Dear Professor Lu，
Thanks for your suggestions. Do you have any suggestions on how to compensate the response time of PZT or how to design the circuit? Relevent publications or details are appreciated. Thanks for your answers.
Best regards,
Guangzhen
Hello, there are two important factors for Stabilisation: amplification for your Feedback loop and Phase responce (or delay of your Feedback). Your Feedback Loop should act in anti-Phase to the incoming mechanical distortion Signal, otherwise instead of damping you will get amplification. Amplification factor of the Feedback circuit should not exceed Level of mechanical distortions. So therefore you Need to characterize mechanical distortions in your Setup and found Amplitude of the Feedback Signal which fit to These distortions. You will Need also to characterize delay of the your piezo-element and Feedback circuitry in order to understand Phase shifts in your Stabilisation System. Regards.
You might get an impedance analyzer, and follow the matching circuit theory to try the experiment. Of course, if the PZT work in the range of smaller voltage, you could measure the total circuit precisely.
1.Does the hysterisis or linearity and creep effects of the PZT influence the stabilization? Should these effects be seperatedly compensated or not, because the PZT is part of the closed-feedback-loop.
I think it depends on the working frequency of PZT.
Dear Professer Bychikhin,
Thank you for your answers. The experimental figures are attached. CH1 is the output of photodiode. CH2 is the feedback error signal applied to the PZT to stabilize the interferometry. To be noted, in the experiment, two PZTs were used. PZT1 is used for simulating the low frequency noise, PZT2 is used to stabilize the interferometry. The applied signal for PZT1 is 15 Hz, 100 mVpp and 500 mVpp respectively. Case 100 mVpp: after stabilization, the feedback has worked but the output of PD is still an sine, which is too large to be accepted. How can I reduce this signal further ? Then a 500mVpp signal is added to PZT1, the feedback does not work well then. I suppose the system bandwith limits the feedback, the Proportional-Integral factors should be well designed, and the response time of the system should be considered. Do you have any suggestions on how to design the system considering the response time of PZT and P-I circuits ? Thanks for your suggestions.
Dear Professor Lu，
The experimental detail has been added above. Do you have any suggestions considering the results? From above to the bottom are (a): 15 Hz， 100 mVpp applied to PZT1 before stabilization (b)15 Hz， 100 mVpp applied to PZT1 after stabilization (c): 15 Hz， 500 mVpp applied to PZT1 before stabilization (d) 15 Hz， 500 mVpp applied to PZT1 after stabilization. Thank you for your answers.
What is the bandwidth of Your Feedback? From your Pictures I have Feeling that your piezo has mechanical resonance around 500Hz (high frequency oscillations in Figs. c and d) and Feedback is not able to operate in this bandwidth. In case of 15Hz 100mVpp Signal I would advise You to vary proportional gain factor (+- 20-50% from actual value) to see an effect on PD Signal. Another advise: use PZT1 Signal with lower Amplitude, so that the movement of the PZT2 produce displacement lower than wavelength of your laser. I hope that you understand how stabilization is working: (your PD Signal is a periodical function of the linear displacement of the PZT2, so stabilization Feedback network can work only in case of monotonous part of the PD Signal, either increasing or decreasing PD Signal with displacement . In case when your mechanical dicplacements more than wavelength of the laser source in the Interferometer the Feedback will not be able stabilize the position of the piezo - which is probably the case in your 500mVpp experiment).
In order check resonances of your piezo you can record PD Signal while gently kicking the post or mechanical holder of your PZT2 with pen or small metal stick.
Which type of the Interferometer are you using? I suppose it is Michelson type...
Could you please send me an article or document where your experimental Setup was well described?
Regards.
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering
- Signal Processing