[with Zach and Dmass] We discussed about the stainless steel pmc design  and here are the list of what should be modified.

The drawing can be found, on svn.

pmc_endcap_v2,

•  the hole for the beam exit can be larger (x1.3) so that the exit beam can pass without clipping. It should not be larger than the pzt
•  [Important]the thickness of the PZT, thickness of the back mirror, end cap,  will determine the optical path of the PMC (with the beam centered at the input/output mirrors), see pmc_spacer_v2. Make sure to get the thickness right, so the beam is centered on all mirrors.

 pmc_clamp

• the clamp for the input / output window can have a little larger opening to avoid clipping the beam. The through hole diameter can be 0.9 in stead of 0.8".

pmc_base

• The height should be corrected for 3" beam height, measured from the table to the center of the mirror (the previous one was designed for 4" beam height. The size of the ball bearing has to be specified (Dmass said it was for 3 mm radius).

Materials

• find the ruby/ sapphire bearings for the 3-point mount sapphire ball.
• Other parts of the PMC (spacer, pmc clamp, end cap) will be made from stainless steel, the base will be made from brass.
• press fit slot and press fit cone will be made from hardened steel dowel pin from McMaster

assembly

• make sure the assembly has the right beam height (3" for CTN)
• make sure to glue (with epoxy) the mirror, pzt, end cap carefully so all the parts are parallel, no tilt, no yaw.
• order ruby or sapphire ball

PZT can be ordered from www.pi.ws.

The requirements for PZT from (LIGO-xx), are (A) pzt range = 2.7 FSR, for 0 - 375 V, (B) resonant frequency at 10kHz or above.

Zach is using model P-016.10H. The displacement is 15 um (with 1000V), OD = 16mm, ID = 8mm, L = 15mm, resonant frequency = 67kHz.  Assuming the pzt is linear, the displacement will be 5.6 um for 375V, this corresponds to 11 FSR for our cavity (FSR = 454 MHz). I don't know if this will cause some locking problem or not, or it might just give us an extra gain in the pmc loop.

If I follow the requirement, the displacment of 5um @ 1000V will be enough for us (model P-016.00H), but the length of the PZT will be 7 mm, and I think we have to fix the drawing accordingly.

Above, an excerpt from the pzt catalog, the full one can be found HERE.