The prior art bi-material MEM micro-cantilever devices bend, or deflect, when
irradiation is absorbed by an absorber element of the micro-cantilever and heats the bi-material section of the micro-cantilever, causing one of the bi-materials to expand at a greater rate than the other bi-material and resulting in deflection or bending of the micro-cantilever.
As a micro-cantilever device bends in response to incident radiation, it approaches a physical limitation to its degree of bending.
For example, if a micro-cantilever device is fabricated to bend freely downward in response to incident radiation, the physical limitation is reached when the micro-cantilever touches the substrate over which it is formed.
For a micro-cantilever device chosen to bend upward in response to incident radiation, this too will reach a physical limitation point past which it can no longer bend as a bi-material.
While a significantly higher
dose of radiation forces the micro-cantilever to bend slightly more towards its physical limitation, the degree of bending is not proportional and the device response is not linear in this region.
Since the intensity of an
optical image ultimately produced is based on the degree of bending, prior art devices have a poor
dynamic range and are limited
linearity, results in producing an image having the same shortcomings.
Moreover they possess a limited range of linear response to irradiative heating.
Aside from its cost and bulk, the
disadvantage is that power consumed by this thermoelectric refrigerator / heater is typically the single largest power drain in the imaging
system.
Another limit of prior art configurations has been the degree to which the bi-material bends for a given absorbed
irradiance.
However, this desirable bending has been limited in prior art by the choices of bi-material materials and bi-
material design such as the difference in
thermal expansion between the
expansive bi-materials, the high
thermal conductivity of the thermal isolating leg, and the stiffness of the cantilever support legs.
First, the prior art choice of bi-material materials has been limited to aluminum or gold as the
expansive material, and
silicon nitride as the relatively non-
expansive material.
Second, the bending has been limited by a necessity to achieve rapid thermal stabilization of the temperature of the absorber plate and bi-material within a fraction of a video frame period, so that the bi-material bending can be measured before the end of the frame.
Third, a limit of prior teaching has been the
mechanical resonance of the cantilever itself because the support leg, which includes the insulator leg plus bi-material, must be made stiff enough that the freely supported
mass of the absorber plate has a resonant frequency higher than that of ambient acoustic fields, typically higher than 20,000 Hz.
Yet another limit of prior art bolometers is their necessity to “blink” frequently with the display periodically going visibly black, using a mechanical
shutter to remove the target
irradiance from the array and zero out all the various readings of the
bolometer sensors as they slowly drift in base temperature.
This is occasioned by the limited irradiance range of the
bolometer: essentially all of the
dynamic range of the
bolometer is required for constructing the final image.
Therefore A / C-coupled systems employ a mechanical
chopper, with the attendant issues of weight, power,
life time and
delicacy as to mechanical shock.
A DC coupled thermal imaging
system measures a tiny
signal on top of a large DC background
signal, which is a primary cause for
noise limitation in the minimum discernable
signal.
The DC-coupled thermal imaging
system must
handle the relatively very large offset as well as the
signal of interest This complicates the system because the offsets differ from pixel to pixel, and the differences vary slowly with time, increasing
spatial noise in the system.
Both A / C and DC-coupled systems perform a similar comparative function by periodically shuttering the system at some time interval from a few seconds to a few minutes, usually resulting in an interruptive image freeze upon shuttering.
(a) Because the temperature of the sensor element must reach a stable level before measurement may be made, the thermal
time constant of the micro-cantilever must be made a fraction of the
frame time. As a result the thermal conductance of the thermal insulator leg must be large, the temperature rise for a given irradiance is less and the bend of the responsive bi-material is decreased.
(b) Because in the past the use of aluminum or gold for the expansive material and
silicon or
silicon nitride for the less-expansive material, the difference in
thermal expansion has been limited and the bend of the bi-material is decreased.
(c) Because the
mechanical resonance of the micro-cantilever must be higher than ambient acoustic frequencies, the bi-material leg must be made stiff, and the bend of the bi-material is in response to irradiance is decreased.
(d) In past micro-cantilever bolometers the measure of irradiance is an
analog signal, with the attendant 1 / f, shot, Johnson, and
amplifier noise, which together increase the irradiance required to exceed the
noise of the sensor.
(e) As a result of the noise and insensitivity, bolometers have required the longest
exposure possible to reach a stable temperature for measurement within the limited
frame time, and thus have a limited ability to operate at a higher
frame rate.
(f) Because of the need for stabilization of the substrate temperature, use of a costly and power-draining thermoelectric temperature
regulator has been common.