"; _cf_contextpath=""; _cf_ajaxscriptsrc="/cfthorscripts/ajax"; _cf_jsonprefix='//'; _cf_websocket_port=8578; _cf_flash_policy_port=1244; _cf_clientid='0AFB9B3F99CE8FD97A3F48265B53A9B5';/* ]]> */
Co-Fired Piezoelectric Actuators, 4.6 µm to 20.0 µm Travel
9.1 µm Stroke
4.6 µm Stroke,
20.0 µm Stroke
4.6 µm Stroke
Thorlabs' Co-Fired Piezoelectric Actuators are composed of stacked PZT layers separated by electrodes that span the entire surface area of the layer. The entire stack is sintered together into one monolithic structure. These piezoelectric actuators transform electrical energy into precisely controlled mechanical displacements and are ideal for applications requiring rapid, precise positional changes on the nanometer or micrometer scale.
Co-Fired Piezo Stacks differ from Discrete Piezo Stacks in the insulation and conductor design. In co-fired stacks, small amounts of glass filament are precisely located on each side of the stack so that every internal electrode is insulated from the external supply electrode of opposite polarity. This separation enables sequential layers to alternate connections to the positive and negative poles (or applied bias and ground). Compared to using an insulating gap as in discrete stacks, this design allows for a higher contact area and percentage of active PZT material than discrete stacks, which lends itself to lower internal stress. For more details and a diagram further comparing the two piezo stack insulation methods, please refer to the Operation tab.
These multilayer devices are ideal for nano- and micro-positioning. As the voltage applied to the actuator goes from 0 V to the maximum drive voltage, the piezo expands longitudinally. These open-loop piezoelectric actuators offer maximum displacements from 4.6 µm to 20.0 µm. Piezoelectric devices, such as these actuators, exhibit hysteresis, thus the displacement of the actuator is not solely based on applied voltage. When it is necessary to precisely track the displacement of the actuator, Thorlabs recommends our piezo stack actuators with attached strain gauges.
The green lead on the piezoelectric actuator must be connected to the high side of the voltage source used to drive the actuator. Do not drive the piezoelectric actuator with a reverse bias voltage, as this could destroy the device. Piezoelectric actuators should not be used in liquid, in the presence of combustible gases or liquids, or cleaned with organic solvents.
Thorlabs also offers an expanding line of 75 V, 100 V, and 150 V piezoelectric stacks and chips, including discrete piezo stacks with 5.2 µm - 100.0 µm travel and piezo chips with 0.7 - 3.6 µm travel, which are ideally suited for OEM applications. For applications requiring larger displacements, we offer amplified piezo actuators with 220 µm - 2500 µm travel.
Piezo Driver Bandwidth Tutorial
Knowing the rate at which a piezo is capable of changing lengths is essential in many high-speed applications. The bandwidth of a piezo controller and stack can be estimated if the following is known:
To drive the output capacitor, current is needed to charge it and to discharge it. The change in charge, dV/dt, is called the slew rate. The larger the capacitance, the more current needed:
For example, if a 100 µm stack with a capacitance of 20 µF is being driven by a BPC Series piezo controller with a maximum current of 0.5 A, the slew rate is given by
Hence, for an instantaneous voltage change from 0 V to 75 V, it would take 3 ms for the output voltage to reach 75 V.
Note: For these calculations, it is assumed that the absolute maximum bandwidth of the driver is much higher than the bandwidths calculated, and thus, driver bandwidth is not a limiting factor. Also please note that these calculations only apply for open-loop systems. In closed-loop mode, the slow response of the feedback loop puts another limit on the bandwidth.
The bandwidth of the system usually refers to the system's response to a sinusoidal signal of a given amplitude. For a piezo element driven by a sinusoidal signal of peak amplitude A, peak-to-peak voltage Vpp, and frequency f, we have:
A diagram of voltage as a function of time is shown to the right. The maximum slew rate, or voltage change, is reached at t = 2nπ, (n=0, 1, 2,...) at point a in the diagram to the right:
From the first equation, above:
For the example above, the maximum full-range (75 V) bandwidth would be
For a smaller piezo stack with 10 times lower capacitance, the results would be 10 times better, or about 1060 Hz. Or, if the peak-to-peak signal is reduced to 7.5 V (10% max amplitude) with the 100 µm stack, again, the result would be 10 times better at about 1060 Hz.
Triangle Wave Signal
For a piezo actuator driven by a triangle wave of max voltage Vpeak and minimum voltage of 0, the slew rate is equal to the slope:
Or, since f = 1/T:
Square Wave Signal
For a piezo actuator driven by a square wave of maximum voltage Vpeak and minimum voltage 0, the slew rate limits the minimum rise and fall times. In this case, the slew rate is equal to the slope while the signal is rising or falling. If tr is the minimum rise time, then
For additional information about piezo theory and operation, see the Piezoelectric Tutorials page.
Click to Enlarge
Figure 1: Diagram of Piezo Stack Insulation Methods:
(a) In-Chip Insulation Used in Standard Chips and Discrete Stacks,
(b) On-Stack Insulation Used in Co-Fired Stacks
Co-Fired Piezo Chips, Co-Fired Piezo Stacks, and Discrete Piezo Stacks
In the case of the chips (In-Chip Insulation), the internal electrodes of opposite polarities alternate. Each internal electrode layer is shorter than the full width of the piezo layer. All electrodes of one polarity have edges that are flush with one side of the chip, and all electrodes with the opposite polarity are flush with the opposite side of the chip. Because the electrode does not extend all the way to the opposite edge, the far end of the electrode is completely surrounded by PZT material. The PZT material enclosing the end of the electrode is insulating, which electrically isolates this electrode from the supply electrode of opposite polarity. This approach to electrically insulating the electrodes creates a region of stress at the edge of the electrode. The stress arises both due the abrupt change in thickness on either side of the electrode edge, as well as the tensile stress created when the PZT material sandwiched between electrodes responds to an applied voltage drive signal, but the insulating PZT material beyond the edge of the electrodes does not. This stress limits the maximum height of chips manufactured using this approach. The height of chips are limited to ensure internal stresses are low and do not affect lifetime or performance. Chips are sealed in a ceramic layer that offers superior resistance to humidity and heat than epoxy resin coatings.
One way of increasing the height, and therefore the maximum stroke, of piezo actuators based on these chips is to fabricate discrete piezo stacks. These are manufactured by bonding multiple chips together in series using a glass-bead epoxy. Discrete stacks can be fabricated to substantially longer lengths than co-fired chips or stacks, and this allows them to achieve higher maximum displacements while maintaining sub-millisecond response times and a low drive voltage range. As the constituent chips are sealed within a ceramic barrier layer, discrete stacks have superior resistance to humidity and heat than co-fired stacks, which are sealed in an epoxy resin coating.
In the case of co-fired stacks (On-Stack Insulation), the electrodes extend across the full width of the PZT layers. The edges of the electrodes are flush with all four sides of the stack, including the side with the supply electrode of opposite polarity. The edge of the internal electrode is insulated from that supply electrode by a layer of glass filament applied to the side of the stack. Precision localized application of the glass filament ensures that the electrode edge is electrically isolated from the supply electrode, and that the filament is applied over minimal surface area; the ability of the supply electrode to make electrical connections to the desired internal electrodes is not affected, and the small amount of applied glass filament does not affect the operation of the actuator. With their full-width electrodes, piezo actuators made using this insulating approach are characterized by homogeneous internal stress. Co-fired stacks can therefore be fabricated with greater heights than chips fabricated using the in-chip insulation approach. Co-fired stacks also have a higher percentage of active PZT material than the discrete stacks, which include inactive bonding layers of glass-bead epoxy. They are coated in an epoxy resin.
Interfacing a Piezoelectric Stack with a Load
To attach a load to one of our AE series piezo stacks, we recommend using a room-temperature epoxy, such as Thorlabs' F120. For connecting loads to our PC4WL, PC4FL, and PC4QM piezo stacks, we recommend using an epoxy that cures at a temperature lower than 80 °C (176 °F), such as our 353NDPK or TS10 epoxies or Loctite® Hysol® 9340. Loads should be mounted only to the translating, uncoated faces of the piezoelectric stack; the coated sides of our co-fired piezo stacks do not translate, and mounting a load to a non-translating face may lead to the mechanical failure of the actuator. Our PC4 Series Piezo Stacks are compatible with hemispheres and end cups to minimize internal stress when mounting. Please refer to the info icons () for item-specific compatibility. Some correct and incorrect approaches to interfacing loads with piezoelectric stacks fitted with end plates are discussed in the following.
Click to Enlarge
Figure 2: Actuation of a lever arm using stack fitted with a flat plate (A, Incorrect), and a hemispherical plate (B, Correct).
Click to Enlarge
Figure 3: Loads properly and improperly mounted to PZT actuators using a variety of interfacing methods.
Figure 2 presents incorrect (A, far-left) and correct (B, near-left) methods for using a piezoelectric stack to actuate a lever arm. The correct method uses a hemispherical end plate so that, regardless of the angle of the lever arm, the force exerted is always directed along the translational axis of the actuator. The incorrect interfacing of the stack and the lever arm, shown at far-left, endangers the stack by applying the full force of the load to one edge of the stack. This uneven loading causes dangerous stresses in the actuator, including a bending moment around the base.
Figure 3 shows one incorrect (near-right, A) and three correct approaches for interfacing a flat-bottomed, off-axis load with a piezoelectric stack. Approaches A and B are similar to the incorrect and correct approaches, respectively, shown in the image at left. Correct approach C shows a conical end cup, such as the PKFCUP, acting as an interface. The flat surface is affixed to the mating surface of the load, and the concave surface fits over the hemispherical dome of the end plate. In the case of correct approach D, a flexure mount acts as an interface between the off-axis flat mounting surface of the load and the flat mounting plate of the actuator. The flexure mount ensures that the load is both uniformly distributed over the surface plate of the actuator and that the loading force is directed along the translational axis of the actuator.
Operating Under High-Frequency Dynamic Conditions
Estimating the Resonant Frequency for a Given Applied Load
Quick changes in the applied voltage result in fast dimensional changes to the piezoelectric stack. The magnitude of the applied voltage determines the nominal extension of the stack. Assuming the driving voltage signal resembles a step function, the minimum time, Tmin, required for the length of the actuator to transition between its initial and final values is approximately 1/3 the period of resonant frequency. If there is no load applied to the piezoelectric stack, its resonant frequency is ƒo and its minimum response time is:
After reaching this nominal extension, there will follow a damped oscillation in the length of the actuator around this position. Controls can be implemented to mitigate this oscillation, but doing so may slow the response of the actuator.
Applying a load to the actuator will reduce the resonant frequency of the piezoelectric stack. Given the unloaded resonant frequency of the actuator, the mass of the stack, m, and the mass of the load, M, the loaded resonant frequency (ƒo') may be estimated: