Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.mrao.cam.ac.uk/projects/OAS/pmwiki/uploads/MROIFastTipTilt.TechMeet15/brainstorm.pdf Дата изменения: Thu May 19 19:09:54 2011 Дата индексирования: Sat Mar 1 04:01:53 2014 Кодировка: Поисковые слова: п п п п п п п п

Testing Mount Stability

Alex Rea adr34@cam.ac.uk

Cavendish Lab oratory, University of Cambridge

FTT Brainstorm Meeting Monday 16th May 2011

Alex Rea (Cavendish Labs, Cambridge)

Testing Mount Stability

16th May 2011

1 / 21


Goal

Alex Rea (Cavendish Labs, Cambridge)

Testing Mount Stability

16th May 2011

2 / 21


Goal
Want to discover if the mounts designed meet the stringent stability requirement that we have placed on them. To directly test the requirements, under a temperature change of 5 C, we must be able to measure Tilt of a Shear of Tilt of a Shear of Shear of flat mirror to better than 0.05 . a flat mirror (along optical axis) to better than 0.3µm. lens to better than 0.7 . a lens (perpendicular to optical axis) to better than 0.4µm. a lens (along optical axis) to better than 140µm.

Measuring a larger movement than these over a larger change in temperature is likely to be acceptable. Note that the displacements these angles correspond to are tens of nanometers. Ideally, such measurements will be (semi-)automated.
Alex Rea (Cavendish Labs, Cambridge) Testing Mount Stability 16th May 2011 3 / 21


Goal

In the first instance, we only want to see if these requirements are being met. Only if they are not do we want to investigate why (e.g. the mount is expanding in an unexpected way, or the optic is deforming). The tests for these two cases will almost certainly be different.

Alex Rea (Cavendish Labs, Cambridge)

Testing Mount Stability

16th May 2011

4 / 21


Measurement Methods
There are a great many of these! A selection includes: Resistive Potentiometers Inductive Sensor Capacitive Sensor Piezoelectric Sensor Strain Gauge Michelson interferometer Optical Beam Deflection Autocollimator Pencil-beam interferometry
Alex Rea (Cavendish Labs, Cambridge) Testing Mount Stability 16th May 2011 5 / 21


Measurement Methods
There are a great many of these! A selection includes:

Measure Linear Displacement
Resistive Potentiometers Inductive Sensor Capacitive Sensor Piezoelectric Sensor Strain Gauge Michelson interferometer

Measure Angular Displacement
Optical Beam Deflection Autocollimator Pencil-beam interferometry
Alex Rea (Cavendish Labs, Cambridge) Testing Mount Stability 16th May 2011 6 / 21


Measurement Methods
There are a great many of these! A selection includes:

Sensor near object
Resistive Potentiometers Inductive Sensor Capacitive Sensor Piezoelectric Sensor Strain Gauge

Sensor far from object
Michelson interferometer Optical Beam Deflection Autocollimator Pencil-beam interferometry
Alex Rea (Cavendish Labs, Cambridge) Testing Mount Stability 16th May 2011 7 / 21


Resistive Potentiometers

Precision Potentiometers

Pots are available in great variety, with specific surement requires a high-quality pot designed f rheostats, attenuators, volume controls, panel c

Fundamentally the moving object pushes a wiper on a resistive track. Resistive plastic or similar means resolution limited by electrical noise etc., not thickness of wound wire. Generously get down to 1 micron - so unlikely to be good enough.

FIGURE 6.1
Alex Rea (Cavendish Labs, Cambridge) Testing Mount Stability

Representative cutaways o
16th May 2011 8 / 21


Inductive Sensor

Various types; most sensitive are those based on Eddy Currents (down to 0.1µm). Non-contact. We have a conductive surface on our reflective optics already - but is it thick enough? Unlikely, 0.25mm required. Have to think hard about stray fields due to this sensitivity - including multiple probes.

Alex Rea (Cavendish Labs, Cambridge)

Testing Mount Stability

16th May 2011

9 / 21


Figure 2. Eddy-current sensor cons

Capacitive Sensor

One of the most sensitive options under ideal conditions, with resolution down to 1nm. Non-contact. Strongly affected by dielectrics appearing between surfaces, such as condensing water at low temperatures. Can use multiple probes on one surface, if synchronised. Would need to add such surfaces to lens.
130% of Sensing Area Diameter
Alex Rea (Cavendish Labs, Cambridge) Testing Mount Stability 16th May 2011 10 / 21


Piezoelectric

Inherently contacted, using pressure due to movement to create a voltage which can be measured. As a result, could easily disrupt behaviour. Particularly bad with curved surfaces of lens - uneven pressure over sensor, leading to unsure readings. Seem pretty difficult to track down, in any case.

Alex Rea (Cavendish Labs, Cambridge)

Testing Mount Stability

16th May 2011

11 / 21


Strain Gauge
A geometric change in the length of the conductor leads to a change in resistance, which can be measured. Only works properly when attached to a single solid body. As a result, unlikely to be useful for us when considering the gross movement of the optic. But perhaps useful for breaking down gross movement if needed. Only sensitive in one direction. Temperature dependence can be corrected for using manufacturer supplied data.
Alex Rea (Cavendish Labs, Cambridge) Testing Mount Stability 16th May 2011 12 / 21


Optical Beam Deflection
By reflecting a laser beam from the surface of the optical element, when the surface moves, so will the beam. Most common approaches involve a knife edge or bicell, and measuring the change in incident power seen.

FIGURE 6.114 Essential features of a position-sending detector (PSD), as represented by a photodetector shadowed by a semiinfinite knife edge.

where P is the Alex Rea (Cavendish Labs, Cambridge)total

incident beam esting = 2/r12, r1 isStability beam radius, and r ! 2 = x !2 + y ! 2. Tpower, a Mount the Gaussian 16th

May 2011

13 / 21


Optical Beam Deflection

Unfortunately for us, the displacements involved correspond to having a preposterously sized laser beam (on the order of micron diameter). In addition, need something further to determine whether tilt or shear is being seen.

Alex Rea (Cavendish Labs, Cambridge)

Testing Mount Stability

16th May 2011

14 / 21


Michelson Interferometer

Even without trying to be clever, can get down to

4

i.e. 0.15µm.

Need to count fringes though movement, so if multiple points wish to be monitored, need beams and detector for each. Good for shears, less so for tilts.

Alex Rea (Cavendish Labs, Cambridge)

Testing Mount Stability

16th May 2011

15 / 21


Autocollimator
Projects a crosshair onto the optic, which is reflected and then compared to a fixed scale in the eyepiece. Electronic autocollimators can get down to 0.01 , though are much more temperamental than their visual autocollimator counterparts. Can't directly measure lens shear, and certainly couldn't measure `defocus'.

Alex Rea (Cavendish Labs, Cambridge)

Testing Mount Stability

16th May 2011

16 / 21


Autocollimator

Mirror 1B Lens

Mirror 2B

Autocollimator A1
Mirror 1A

Autocollimator A2
Mirror 2A

Figure 3: Measurement set-up
Alex Rea (Cavendish Labs, Cambridge) Testing Mount Stability 16th May 2011 17 / 21


Precise Angle Monitor
Sends two beams to surface separated by a few mm, and then focusses their reflections through a lens and looks at interference pattern. Tilt moves the interference pattern on the detector.

Lens Detector
Alex Rea (Cavendish Labs, Cambridge) Testing Mount Stability 16th May 2011 18 / 21


Precise Angle Monitor

Resolution better than 0.01 . Much better than direct observation of single beam - due to interference fringe much sharper compared to intensity profile of single beam. Fitting multiple maxima and minima achieves this accuracy. Only requires small surface. Can generate two pairs of beams to detect rotation about orthogonal axes with one set of apparatus. Insensitive to shear.

Alex Rea (Cavendish Labs, Cambridge)

Testing Mount Stability

16th May 2011

19 / 21


Practical Implementation

Any sensor selected needs to be accommodated by the mounts and thermal enclosure. For example: For a capacitive sensor or similar, the sensor itself must be mounted sufficiently stably. Needs to physically fit - so unlikely to fit between lens and mount. For a method involving a light beam, assuming the beam source is outside the box, there must be a window in the thermal enclosure. The focal length of the lens is 1.25m, so similar windows could be needed for an exit beam.

Alex Rea (Cavendish Labs, Cambridge)

Testing Mount Stability

16th May 2011

20 / 21


Your thoughts?

Are there any sensor types you know of that I've missed? Do any of the sensor types I've mentioned have drawbacks I've omitted? Which of the methods I've talked about do you prefer?

Alex Rea (Cavendish Labs, Cambridge)

Testing Mount Stability

16th May 2011

21 / 21