本資料說(shuō)明了反饋取消在最新數(shù)字助聽(tīng)器中提供的功能對(duì)節(jié)奏的?R3920安森美半導(dǎo)體。反饋取消不同于其他反饋減少策略，因?yàn)樗粶p少前進(jìn)正常工作條件下助聽(tīng)器的增益。相反，該算法減少反饋通過(guò)制定

　　真助聽(tīng)器反饋信號(hào)的內(nèi)部估計(jì)然后減去這個(gè)估計(jì)從麥克風(fēng)信號(hào)。如果內(nèi)部估計(jì)是準(zhǔn)確的，那么實(shí)際反饋被取消，助聽(tīng)器不會(huì)尖叫。正確估計(jì)真正的助聽(tīng)器反饋信號(hào)是必不可少的正確操作算法。由于反饋信號(hào)隨時(shí)間而變化算法必須對(duì)內(nèi)部進(jìn)行連續(xù)的調(diào)整模型，以確保外部反饋準(zhǔn)確匹配。在半導(dǎo)體上的反饋消除器中，這些調(diào)整自動(dòng)發(fā)生，不需要探針信號(hào)或其他用戶(hù)干預(yù)可能干擾助聽(tīng)器的操作。最新的反饋消除算法安森美半導(dǎo)體新機(jī)制增加保持音頻質(zhì)量的適應(yīng)速度。這機(jī)制自動(dòng)決定是否當(dāng)前反饋條件需要大或小的調(diào)整內(nèi)反饋?路徑模型。大調(diào)整用于快速消除聲反饋。當(dāng)大不需要調(diào)整，選擇較小的調(diào)整最大化音頻質(zhì)量。下面的部分描述反饋消除操作更詳細(xì)地強(qiáng)調(diào)影響反饋的因素檢測(cè)器的性能。文章的結(jié)論與可用于優(yōu)化反饋的方法的討論檢測(cè)器的性能。這些方法允許用戶(hù)關(guān)于半導(dǎo)體數(shù)字放大器的增加產(chǎn)品性能。

　　PRINCIPLE OF OPERATION Adaptive Feedback Cancellation （AFC）， from ON Semiconductor， is a true feedback cancellation （also known as phase cancellation） algorithm. The algorithm operates by subtracting an internal estimate of the hearing-aid feedback signal from the microphone signal. The internal feedback signal is obtained by passing the hearing-aid output signal through an internal， feedback-path model. The feedback-path model in AFC is intended to match the impulse response， or time response， of the external feedback path. The feedback-path impulse response is the time signal that would be observed at the microphone output if an impulsive signal were produced by the hearing aid. As an example， the measured impulse response of a typical BTE hearing aid is shown in Figure 1?a. This plot shows an initial period of approximately 0.5 ? 0.75 ms with little or no energy. This time window corresponds to the analog-to-digital and digital-to-analog converter delays as well as any acoustic delays due to transducers， tubing， sound propagation through the air， etc. Following the initial delay is the main part of the feedback-path response which， in this example， persists for more than 2 ms. The net effect of the feedback canceller algorithm is to remove a certain portion of the feedback path impulse response. Since computational resources inside a hearing-aid signal processor are limited， only a portion of the total feedback path response can be cancelled. To achieve maximum effectiveness of the canceller， the cancelled part of the response should be that portion of the feedback path with the largest energy. This ensures that the net response has the lowest energy for the given length of canceller. For example， if the cancellation window is restricted to 1ms， then for the feedback path of Figure 1?a， the internal feedback model should look like the plot of Figure 1?b. When this is subtracted from the feedback?path response， the net feedback path is as shown in Figure 1?c. The corresponding frequency response is shown in Figure 2. This figure illustrates that the net effect of the feedback canceller in this example is to reduce the peak response of the feedback path by approximately 12 dB. Consequently， the feedback canceller in this example should provide approximately 12 dB of added stable gain （ASG） that is achieved without any impact on the forward gain of the hearing aid.