Electronic earplug for monitoring and reducing wideband noise at the tympanic membrane

[0001] The present invention relates generally to the field of electronic hearing protector devices. More specifically, the present invention relates to an electronic earplug and a method for designing an electronic earplug that accurately monitors sound pressure at the tympanic membrane for a wide range of frequencies and provides active noise reduction (ANR).

[0002] Active noise reduction (ANR) earplugs have been identified as viable means for suppressing the sound pressure levels inside users' ear canals. The primary advantage of the ANR earplug is its ability to suppress noise inside a smaller occluded space than the corresponding occluded space in prior ANR earmuff designs. This reduction in the volume of the occluded space causes the corresponding reduction in dimension of the acoustic dynamic system to be controlled, thereby allowing control of a broader frequency bandwidth. A significant disadvantage associated with prior ANR earplugs is that all earplug constructions offered thus far cannot guarantee that the electronic control attenuation at an earplug microphone will be the same as the control attenuation at the end of the ear canal, i.e. at the surface of the tympanic membrane.

[0003] At issue in ANR earplug design is the need to control sound pressure at the eardrum over a wide frequency band. Prior research and design has focused on controlling the pressure at the earplug (or error) microphone. Prior ANR earplug designs have failed to appreciate the critical differences in reducing sound pressure at the earplug microphone in the control circuit versus reducing sound pressure at the user's tympanic membrane.

[0004] A number of ANR earplug designs are described in “An Active Noise reduction Ear Plug with Digitally Driven Feedback Loop,” by K. Buck, V. Zimpter and P. Hamery, a paper presented at Inter-Noise 2002, the International Congress and Exposition on Noise control Engineering, Aug. 19-21, 2002 (herein, “Buck”). One such design used a walkman-type loud speaker. According to Buck, the closed-loop performance of that design is not impressive, largely due to the electroacoustic transfer function of the walkman-type loud speaker.

[0005] Buck also describes a piezoceramic actuator. A flat-plate type device exhibited an electroacoustic transfer function that was amenable to ANR applications. However, the pressure output of the flat-plate piezoceramic actuator was insufficient for ANR applications, particularly in a noisy environment. A tube-type piezoceramic actuator was also tested and claimed as possible for use inside the ear canal, without specification of any dimensions. Like the flat-plate design, the transfer function of the tube-type piezoceramic actuator was acceptable, but the prototype device was too inefficient in output sound power for commercial applications. Buck concludes that the main obstacle in designing ANR earplugs is related to the transducers [actuators]. Controlling the pressure at the eardrum location not addressed by Buck.

[0006] Another paper, “Electroacoustic Design of an Active Earplug,” by Phillipe Herzog, a paper presented at Inter-Noise 2002, the International Congress and Exposition on Noise control Engineering, Aug. 19-21, 2002 (herein, “Herzog”), also discusses the design of earplugs with ANR. Herzog also comments on the design constraints posed by current choices for actuators:

• • [0007]  The piezoelectric speaker would allow to use a simpler control filter, but still require expensive developments. An cheaper solution, requiring also a simple control filter, would be electret speaker, if a relatively low pressure is to be controlled. Conversely, the emergence efficient numerical control filters may allow us to use existing dynamic speakers. In any case, the maximum pressure inside the ear canal remains a critical criterion.

[0008] Buck and Herzog both focus on the actuator as a main obstacle in the design of an effective earplug with ANR system. Both papers fail to address how to design an earplug so that the electronic ANR performance can be compared to, and tailored to match the performance at the tympanic membrane.

[0009] A publication by Thomas R. Harley, et al., titled “Digital Active Noise reduction Ear Plugs,” Air Force Research Laboratory Report AFRL-HE-WP-TR-2001-0042, points out that ANR earplugs tested in the ear canal simulator of a KEMAR mannequin exhibited discrepancies that showed “. . . cancellation at the earplug microphone did not guarantee cancellation at the (mannequin) eardrum.” However, human test subjects were not tested to determine if this discrepancy was also observed for human tympanic membranes. In addition, the paper does not offer any explanations for this discrepancy nor does it provide any solutions to overcome critical differences in performance between the electronic ANR system and ANR performance at the eardrum of a KEMAR mannequin or the tympanic membrane of a human.

[0010] In U.S. Pat. No. 4,985,925 issued to Langberg et al. (Langberg) presents the idea of a small electroacoustic actuator used as part of a feedback system to close the loop on a second electroacoustic transducer (microphone) to suppress the sound pressure levels observed by the microphone over a relatively narrow band of frequencies (20 Hz to 1 kHz). The speaker and microphone locations for the earplug embodiments identified in Langberg are defined simply as “in proximity of an ear canal.” Figure 2 of Langberg indicates that the speaker is just outside of the ear canal and the microphone is located just inside the ear canal entrance. Langberg discusses the phase of the electroacoustic transfer function as a fundamental obstacle to electronic control performance. However, there is no discussion of the impact of transducer placement on noise reduction performance at the eardrum, or the human perceived performance of the resulting control.

[0011] U.S. Pat. No. 5,305,387 issued to Sapiejewski teaches the use of a first electroacoustic transducer (actuator) arranged inside the concha of a user, with a second electroacoustic transducer (microphone) situated inside a front cabinet volume and adjacent to the actuator, and intercoupling of first and second transducers with feedback electronic circuitry to actively reduce the noise intensity inside the concha cavity.

[0012] U.S. Pat. No. 5,631,965 issued to Chang discusses the use of a microphone to receive an external acoustic sound signal to be electronically processed in the feedforward sense for creation of a sound signal that travels down a tube extending through the ear piece to the ear canal.

[0013] U.S. Pat. No. 5,740,258 issued to Goodwin Johansson also discusses a feedforward active noise suppressor that includes an input transducer used to generate an electrical signal in response to sound pressure waves entering the ear canal, then is processed to generate an inverse noise signal applied to the output transducer. No specifications or embodiments that correlate the output transducer measurements to the tympanic membrane sound pressures are provided or discussed.

[0014] What would be useful is an electronic earplug that monitors and replicates sound pressures at the tympanic membrane thereby offering electronic enhanced noise reduction performance at the tympanic membrane over a specific frequency bandwidth. Such an electronic earplug would also provide accurate monitoring; and high-bandwidth noise reduction performance at the tympanic membrane for the required conditions of a hearing protection application. The accurate resemblance, or mapping, of complex sound pressures at the earplug microphone to the complex sound pressures at the tympanic membrane would also permit calculations of noise exposure limits for humans in hazardous noise fields.

[0015] An embodiment of the present invention is an electronic earplug that provides means to accurately monitor an electronic reduction of acoustic pressures at the tympanic membrane (TM) over a wide range of frequencies. The electronic earplug according to this embodiment, monitors the TM sound pressures, determines an individual's time-intensity noise exposure (dosimetry), and provides ANR noise reduction at the TM at frequencies above 1.5 kHz.

[0016] Therefore, an aspect of the present invention is an electronic earplug that provides a high degree of predictable passive noise attenuation at the TM.

[0017] It is another aspect of the present invention to incorporate into an electronic earplug transducers located to provide accurate dosimetry of either uncontrolled (open-loop) or electronically controlled (closed-loop) sound pressure levels at the TM surface as well as ensure that any electronic noise reduction delivered to the earplug microphone will correspondingly be delivered to the eardrum and be perceived as actual control over the design bandwidth.

[0018] A further aspect of the invention is an electronic earplug that provides accurate dosimetry of the TM noise exposure without simultaneous electronic control.

[0019] Yet another aspect of the present invention is to measure electronic noise control attenuation at the earplug microphone location that is the same as the control attenuation at the tympanic membrane over some specified range of frequencies.

[0020] Another aspect of the present invention is a method of designing an electronic earplug wherein the method supports the determination of the frequency range over which the ANR control attenuation at the earplug microphone is the same as the ANR control attenuation at the tympanic membrane.

[0021] Yet another aspect of the present invention is an electronic earplug that extends the active noise reduction bandwidth to frequencies greater than 1.5 KHz.

[0033] An embodiment of the present invention relates the design features of an electronic earplug and the connectivity of such features to the anatomical features of the occluded ear canal space to the dynamic acoustic impedance at any selected location inside the earplug or ear canal. Prior art electronic earplug designs have not sufficiently addressed those acoustic dynamics. The control attenuation of these prior art designs is not comparable to (and is always greater than) the actual perceived active noise reduction performance achieved at the tympanic membrane. This is especially true at frequencies higher than approximately 1.0 kHz for prior art designs of electronic earplugs. This observation is critical for any noise attenuation goals that relate to either human perception of the control attenuation or hearing safety metrics that involve measurement of exposure (e.g. Occupational Safety and Health Administration (OSHA) calculations of human noise exposure limits). This is because the sound pressure level at the TM is directly proportional to the perceived loudness and the inner ear fatigue associated with hazardous sound pressure levels.

[0034] FIG. 1 illustrates an electronic earplug designed according to an embodiment of the present invention to accurately monitor sound pressure at the TM and to provide noise reduction over a wide bandwidth at the TM. Electronic earplug 100 comprises an earplug housing 105, an earplug microphone 110, and an acoustic driver 115. The earplug housing 105 reaches sufficiently into the ear canal 120 to accommodate a position for the earplug microphone 110 to monitor noise exposure at the TM 125 and to provide noise reduction performance at the TM to some determined bandwidth, as discussed below. Acoustic driver 115 provides means for combing electronic control and/or communication with noise monitoring.

[0035] In an embodiment, the acoustic driver 115 is a balanced armature hearing aid type receiver and earplug microphone 110 is an electret type transducer. In another embodiment, acoustic driver 115 is a miniaturized voice coil speaker. In still another embodiment, acoustic driver 115 is a piezoelectric speaker. As will be apparent to those skilled in the art, other devices that receive and actuate acoustic energy may be employed to perform the tasks of acoustic driver 115 and earplug microphone 110 without departing from the scope of the present invention.

[0036] In yet another embodiment, earplug housing 105 is custom molded to fit the ear canal 120 of a user. In an alternate embodiment, the custom molded portion of the earplug housing 105 is limited to the portion of the housing that extends into the ear canal 120. By way of example and not as a limitation, an earplug housing 105 may be constructed from polyvinylchloride, acrylic, a combination of polyvinylchloride and foam, and a combination of acrylic and foam.

[0037] FIG. 2 is a block diagram illustrating the signal paths in an electronic earplug designed according to an embodiment of the present invention. An audio disturbance 205 impinges on a user of an electronic earplug (see FIG. 1) and the sound pressure associated with audio disturbance (d) is directed to earplug microphone 210 and TM 220. The sound pressure of the audio disturbance at earplug microphone 210 is represented as P_d,e. The sound pressure of the audio disturbance at TM 220 is represented as Pd_im. Earplug microphone 210 is connected to controller 230. Controller 230 connects to acoustic driver 215. Controller 230 is characterized by transfer function G_c and acoustic driver 215 is characterized by transfer function G_sp. A portion of the output of acoustic driver 215 is fed back to earplug microphone 210 and impinges on tympanic membrane 220. The acoustic dynamics that are excited by acoustic driver 215 are represented by two transfer functions. The acoustic dynamics between acoustic driver 215 and the tympanic membrane are given by transfer function G_p,tm, while the acoustic dynamics between the acoustic driver 215 and the earplug microphone are represented by transfer function G_p,e. The combination of the audio disturbance sound pressure P_d,e and the control sound pressure P_u,e at the earplug microphone 210 forms the total sound pressure P_t,e measured by the earplug microphone. The combination of the disturbance sound pressure P_d,tm and the control sound pressure P_u,tm at the tympanic membrane forms the total sound pressure perceived at the eardrum tympanic membrane P_t,tm.

[0038] Based on the input-output relationships of FIG. 2, the total sound pressure at the surface of the tympanic membrane is described by Equation 1 below: $\begin{array}{cc}{P}_{t,\mathrm{tm}}=P_{d,\mathrm{tm}}+P_{u,\mathrm{tm}}=P_{d,\mathrm{tm}}+\frac{G_c{G}_{\mathrm{sp}}{G}_{p,\mathrm{tm}}}{1-G_c{G}_{\mathrm{sp}}{G}_{p,e}}{P}_{d,e}& \left(\mathrm{Eq}1\right)\end{array}$

[0039] Thus, the total sound pressure level perceived at the surface of the tympanic membrane is related to two different noise generation mechanisms, one from the feedback acoustic signal emitted acoustic driver 215, causing sound pressure P_u,tm, and one from the external disturbance noise, causing sound pressure P_d,tm. By considering each term more carefully, it can be observed that P_d,tm is the sound pressure field at the tympanic membrane after the passive control of the electronic earplug 200 has been applied, and P_u,tm is the sound pressure at the tympanic membrane from the active noise controller design.

[0040] One critical observation is that this expression for the sound pressure at the TM is considerably different than the equations used for design of electronic earplugs in prior art. Prior art typically assumes that the noise field being controlled at the earplug's microphone (FIG. 1,110 and FIG. 2,210) is identical to that at the TM (FIG. 1,125 and FIG. 2, 220). This design approach incorrectly assumes that the total sound pressure at the TM 220 (due to disturbance and control) will be reduced by a controller having a transfer function G_c in accordance with Equation 2: $\begin{array}{cc}{P}_{t,e}=\frac{1}{1-G_c{G}_{\mathrm{sp}}{G}_{p,e}}{P}_{d,e}& \left(\mathrm{Eq}.2\right)\end{array}$

[0041] Prior art designs based on minimizing the sound pressure expression of Equation 2 only results in minimization of the total complex sound pressure P_t,e at the earplug microphone 210 and not at the TM 220 where the control is truly desired.

[0042] Although Equation 2, derived from FIG. 2 is true at the earplug microphone, satisfying the expression of Equation 2 does not assure that the total sound pressure at the tympanic membrane P_t,tm will be minimized. An embodiment of the present invention utilizes the information presented in Equation 1 to design and construct electronic earplugs using active noise control to achieve effective electronic noisy reduction performance at the TM location for known bandwidths and to provide earplug microphone sound pressure measurements that accurately monitor the TM sound pressure (accurate dosimetry) for both open loop and/or closed loop.

[0043] Equation 1 represents the closed loop expression for the sound pressure at the TM. This expression can be rewritten as the ratio between the total sound pressure at the tympanic membrane P_t,tm and the disturbance-generated sound pressure P_d,tm (since the disturbance generated sound pressure at the earplug microphone P_d,e is largely equal to the disturbance generated sound pressure at the TM P_d,tm): $\begin{array}{cc}\frac{P_{t,\mathrm{tm}}}{P_{d,\mathrm{tm}}}=1-\frac{G_{p,\mathrm{tm}}{G}_c{G}_{\mathrm{sp}}}{1-G_c{G}_{\mathrm{sp}}{G}_{p,e}}& \left(\mathrm{Eq}3\right)\end{array}$

[0044] Equation 2 illustrates the relationship of the closed-loop complex sound pressure at the earplug microphone to the disturbance-induced complex sound pressure at the earplug microphone while Equation 3 illustrates the relationship of the closed-loop complex sound pressure at the TM to the disturbance-induced complex sound pressure at the TM.

[0045] An embodiment of the present invention is a method of designing an electronic earplug that provides control attenuation at the eardrum. This embodiment achieves such control attenuation by minimizing the result of Equation 3 directly through design of a frequency response function associated with transfer function G_c of the feedback controller 230 so as to affect frequency response functions associated with transfer functions G_p,tm and G_p,e. In this embodiment, placement of the acoustic driver 215 and earplug microphone 210 define these frequency response functions. If there are no considerations given to the placement acoustic driver 215 and earplug microphone 210, then transfer functions G_p,tm and G_p,e can be considered to be arbitrary. For such a case, the frequency response function associated with transfer function G_c of controller 230 is designed to minimize Equation 3 using a feedback control design technique that targets minimizing broadband disturbance. By way of illustration and not as a limitation, controller 230 uses a loopshaping method or an optimal method of feedback control. One advantage of this embodiment is that acoustic driver 215 and earplug microphone 210 can be placed virtually anywhere in the earplug, and the controller design can ensure designed electronic noise reduction performance at the tympanic membrane. However, the measured electronic noise reduction performance at the earplug microphone 210 will not reflect the realized noise reduction performance at the TM 220. Even so, this embodiment represents an improvement over prior art electronic earplug designs that focused only on minimizing the sound pressure expression of Equation 2.

[0046] Another embodiment of the present invention is to place the acoustic driver 215 and earplug microphone 210 so as to define the frequency response of transfer functions G_p,tm and G_p,e to provide the same control bandwidth at the TM as at the earplug microphone 210 while designing transfer function G_c of controller 230 such that its frequency response minimizes the result of Equation 2. This embodiment focuses on design and constructing an electronic earplug in such a way that will ensure that perceived electronic noise reduction performance at the TM 220 is the same as the measured electronic noise reduction performance at the earplug microphone 210 by governing the specific placement of the acoustic driver 215 and earplug microphone 210 in the electronic earplug 200.

[0047] Using this embodiment, the earplug microphone 210 is positioned in the electronic earplug 100 such that it is mathematically similar to the TM 220 in its acoustic impedance over some bandwidth. In general, the complex ratio
$\frac{G_{p,\mathrm{tm}}}{G_{p,e}}$
depends on the specific location of the earplug microphone 210 inside the electronic earplug 200 that will in turn be dependent on the construction of the earplug itself. It is helpful but not sufficient that this ratio equals unity over as large a frequency bandwidth as possible. If the two transfer functions are exactly equal, Equation 3 becomes Equation 2. (This occurrence rarely happens in practice, and only over a very small bandwidth).

[0048] FIGS. 3 and 4 illustrate how the frequency response functions associated with transfer functions G_p,tm and G_p,e for a typical electronic earplug affect the design methods according to this invention.

[0049] FIG. 3 illustrates the complex-valued frequency response functions associated with transfer functions G_p,tm and G_p,e, in magnitude and phase units for an electronic earplug designed according to an embodiment of the present invention. In this embodiment, the open loop magnitude and phase are nearly identical over a certain bandwidth as a result of the placement of acoustic driver 115 and earplug microphone 110. If the open loop magnitude is much greater than unity, this embodiment will produce a closed-loop electronic noise reduction performance at the TM as illustrated in FIG. 4.

[0050] In most cases, it is not possible to design the loop gain to be much greater than unity (which implies very high closed loop attenuation) and a more general result follows from Equation 3. FIG. 5 compares the closed-loop electronic noise reduction performance at the TM to the electronic noise reduction performance at the earplug microphone, based on the calculation of Equation 3, for the same earplug design illustrated in FIGS. 3 and 4. The differences in TM closed-loop performance between FIG. 4 and FIG. 5 are due to the additional dependence on G_c which has a low loop gain. FIG. 5 illustrates that as result of the design process of this embodiment of the present invention, the TM electronic noise reduction performance is approximately identical to the electronic noise reduction at the earplug microphone up to a frequency of approximately 3 kHz.

[0051] As previously noted, an objective of the present invention is knowledge of the complex sound pressure at the TM. The total TM sound pressure P_t,tm may be identified in any number of ways known to those skilled in the art. Identification methods include direct measurement with a probe that extends to the immediate vicinity of the tympanic membrane, or estimation methods based on a one-dimensional acoustic model of the occluded acoustic spaces after insertion of an earplug, or higher-order dimensional models created using analytical or finite element modeling techniques that are well known to acousticians. By way of illustration and not as a limitation, FIG. 6 illustrates a one-dimensional acoustic model according to an embodiment of the present invention. In this embodiment, the transfer function Ĝ_p,tm=G_p,tm is used for the identification of P_t,tm versus P_t,e. FIG. 6 illustrates how the total closed loop sound pressure at the TM (dotted trace) can compare to the total closed loop sound pressure at the earplug microphone (solid trace) for one possible electronic earplug design, when using the one-dimensional acoustic information. FIG. 7 illustrates the experimentally measured comparison of the same quantities (tympanic membrane sound pressure dashed), illustrating the adequacy of this particular TM pressure identification technique for the invention. In each of these cases it is important to note that the design was carried out (in simulation and experiment) to ensure that the control bandwidth was the same for the tympanic membrane as it was for the earplug microphone. As predicted by Equation 3, if the design is not carefully considered for the whole bandwidth, the tympanic membrane sound pressure can significantly deviate from the earplug microphone sound pressure as occurs in both simulation and experiment in higher frequencies of FIGS. 6 and 7.

[0052] As previously described, it is important to consider earplug microphone placement if the electronic earplug design is to achieve the same electronic noise reduction performance at the earplug microphone and the tympanic membrane. As a general rule, the TM electronic noise reduction performance illustrated in FIG. 5 requires that the earplug microphone and the TM be as close to monophase and equal magnitude conditions for as large a frequency band as possible within the geometrical constraints of an earplug user's ear canal dimensions. This depends on the occluded space acoustic dynamics only (ear canal and any exposed volumes inside the earmold) and not the electromechanical transducers' dynamics.

[0053] An embodiment of the present invention locates an earplug microphone and acoustic driver in an earplug based on the frequency of the occluded space acoustic half-wave resonance that occurs nominally above 8 kHz. By way of example, FIGS. 8-11 illustrate the affects of varying the earplug microphone position. In particular, FIGS. 8 and 9 illustrate the electronic noise reduction performance of electronic earplugs designed according to the prior art. FIGS. 10 and 11 illustrate the electronic noise reduction performance of electronic earplugs designed according to embodiments of the present invention wherein the tympanic membrane electronic noise reduction performance is equal to the electronic microphone electronic noise reduction performance over bandwidths at least up to 2 kHz, based on the design formulation presented above through Equation 3.

[0054] FIG. 8 is a comparison of two frequency response functions determined from an electronic earplug typical of the prior art. A dashed trace reflects the frequency response function of a speaker in the concha area of a custom-molded electronic earplug and the complex sound pressure at the surface of the TM. A solid trace reflects the frequency response function between the same speaker in the same concha area and an adjacent microphone location in the concha area. This electronic earplug construction is typical of prior art electronic noise canceling earplug designs where there is no effort or consideration made to locating the earplug microphone in the earplug to ensure commensurate electronic noise reduction performance at the TM. Both the magnitudes and phases of these frequency response functions quantitatively diverge at frequencies above 1 kHz.

[0055] A unique controller design process was then employed to obtain attenuation performance to 2.5 kHz at the earplug microphone. (Prior an controller designs have not been able to achieve this bandwidth even at the earplug microphone). This improved controller design process was carried out using a frequency response function derived by minimizing Equation 2. FIG. 9 compares the corresponding closed loop total pressures at the TM (dashed trace) and the earplug microphone (solid trace, based on Equation 3) for a controller designed to minimize sound pressure at the earplug microphone location. This design process is entirely consistent with and described in prior art (although the bandwidth realized at the earplug microphone in this example exceeds any prior an discussions of control bandwidth).

[0056] As illustrated by FIG. 9, this prior art electronic earplug design does not meet the wideband closed loop performance at the tympanic membrane enabled by embodiments of the present invention. Although the electronic noise reduction performance was designed to minimize the sound pressure at the earplug microphone up to 2.5 kHz, the attenuation perceived at the tympanic membrane never exceeds a frequency of 800 Hz. This particular earplug design, consistent with typical prior electronic earplug art, employs a speaker and an earplug microphone in the concha of the earplug. Such a design, when coupled with a high passive control electronic earplug, cannot provide wideband attenuation at the TM.

[0057] FIG. 10 is a comparison of two open loop frequency response functions of an electronic earplug designed in accordance with an embodiment of the present invention. A dashed trace represents a first frequency response function between a speaker in the concha and the TM. A solid trace represents a second frequency response function between a speaker in the concha area of an electronic earplug and an earplug; microphone location further down the earplug, at a design location selected to match the first frequency response function between 0 and 4 kHz. The exact location of the earplug microphone is dependent on the location of the speaker, the acoustic dynamics in the electronic earplug itself, and the shape of the controller. It is not simply sufficient to place the earplug microphone in front of the speaker, nor is it sufficient to state that it must be within a certain distance of the speaker, or tip of the electronic earplug. Instead, in order to ensure that the electronic noise reduction performance reaches the TM, the design process described in reference to Equation 3 must be used. Equation 3 summarizes that design process either for the controller, or the electronic earplug as a system.

[0058] FIG. 11 illustrates the corresponding closed loop performance comparisons between the closed loop total sound pressure at the TM (dashed trace) and the earplug microphone (solid trace), illustrating wideband (up to 3 kHz) closed loop performance at the TM. FIG. 11 validates the design process of the present invention and demonstrates that placing the specific components of the electronic earplug in accordance with the embodiments of the present invention provides substantially the same electronic noise reduction performance at the TM that is achieved at the earplug microphone. This is contrasted with prior art earplug designs that generate the performance shown in FIGS. 8 and 9, where component placement prevents tympanic membrane control beyond 800 Hz.

[0059] As is known in the art, the open loop characteristics described herein are determined with G_c equal to zero. For dosimetry, or accurate monitoring of sound pressure at the eardrum, the goal is simply to determine accurately the magnitude of the sound pressure at the TM over the desired bandwidth of accuracy for either closed loop or open loop operation. Depending on the acceptable tolerance of the dosimetry measurement, the magnitude may deviate from an exact equality over certain frequency bands and still provide an accurate dosimetry calculation. The phase agreement between the open loop and closed loop transfer functions is not critical to achieve an accurate amplitude measurement used in dosimetry. Therefore, the open loop FRF comparisons presented in FIGS. 8 and 10 are illustrative of the reliability of predicting the TM sound pressures based on placement of the earplug microphone.

[0060] In addition, it has been illustrated that the methods used to identify the TM sound pressures (i.e. determination of G_p,tm) are sufficiently accurate such that either open loop or closed loop values of the TM sound pressure levels can be reliably predicted for any frequency range desired. Therefore, it is possible to utilize information from an earplug microphone placed at any location within an electronic earplug to calibrate the measured levels against the predicted levels (using the acoustic models described earlier for either open loop or closed loop) for accurate estimation, and subsequent monitoring, of the TM sound pressures. For example, the actual open loop exposure at the tympanic membrane can be computed using knowledge of the control to earplug microphone path frequency response magnitude and the control to TM frequency response magnitude. The difference in the magnitudes of these frequency response functions forms a frequency dependent calibration factor that is applied to noise measured at the earplug microphone to determine the estimated TM exposure. Equation 4 illustrates this concept. $\begin{array}{cc}\mathrm{EstimatedOpenLoopTM}\mathrm{pressure}=\left(G_{p,\mathrm{tm}}-G_{p,e}\right)+P_{t,e}& \left(\mathrm{Eq}4\right)\end{array}$

[0061] Here the magnitude of the sound pressure measured at the earplug microphone is modified by the difference in magnitudes of the transfer functions between the frequency responses to arrive at an estimate of the actual sound pressure at the tympanic membrane, based on the measured earplug microphone signal.

[0062] Similarly, this calibration for dosimetry can be performed when an electronic earplug is operating in closed loop for additional noise attenuation. The calibration factor itself is modified by the closed loop nature of the system and can be represented by Equation 5. $\begin{array}{cc}\mathrm{EstimatedClosedLoopTM}\mathrm{pressure}=\left(\frac{1-G_{p,e}{G}_{\mathrm{sp}}{G}_c+G_{p,\mathrm{tm}}{G}_{\mathrm{sp}}{G}_c}{1-G_{p,e}{G}_{\mathrm{sp}}{G}_c}-\frac{1}{1-G_{p,e}{G}_{\mathrm{sp}}{G}_c}\right)+P_{t,e}& \left(\mathrm{Eq}5\right)\end{array}$

[0063] Therefore, embodiments of the invention have been shown that enable the ability to conduct electronic control of sound pressure at the TM for a wide range of frequency, while simultaneously monitoring the sound pressures at the TM for all frequencies, based on the relationships and methodologies presented above. It is clear that the overall invention may include many types of earplug materials, acoustic driver designs, microphone designs and feedback control algorithms without deviating from the scope discussed herein.