Resource | 16.06.2023 | By Imran Mulla

6.6. Test box measures

Introduction:

There are UK quality standards and practical guidance on electroacoustic measures for HAs, BCHDs, and auditory implant sound processors – most commonly CIs.

‘Quality Standards for the use of personal FM systems’ by the UK Children’s FM Working Group was published in July 2008 by the NDCS. This was updated in February 2017 as ‘Quality Standards for the use of personal radio aids: Promoting easier listening for deaf children’. The multi-agency group is now known as the UK Assistive Listening Technology Working Group (ALTWG).

The latest version of the guidance on QS8 Electroacoustic checks with auditory implants was published in March 2022

Of particular relevance is Section 5 of the Quality Standards on the Management and use of personal radio aids, which includes QS7 and QS8.

QS7: Subjective checks of radio aids must take place regularly.

Perform listening checks of the radio aid system daily, with and without the hearing instrument, using appropriate devices such as a stetoclip for hearing aids, monitor earphones for cochlear implants, listeners for bone conduction hearing implants, or a dedicated headphone set for the radio system.

QS8: Electroacoustic checks must be performed regularly and whenever a part of the system is changed.

Complete regular (test box) checks in order to compare the frequency response curves with baseline settings provided at the time of set-up.

A good practice minimum is every half term – this may need to be more often for young children and will depend on the user.

Explanation

Terms used

Hearing device is a catch-all term for hearing instruments and includes auditory implant sound processors, HAs, and BCHDs. The most common auditory implant devices are CIs.

Remote microphone (RM) system is a catch-all term for proprietary microphones associated with specific hearing devices and general radio aids. Systems can be directly connected to the hearing device with dedicated cables or by making use of the telecoil in the hearing device with induction-loop systems.

Health and safety

It is important that health and safety is considered. For example, neck loops are not recommended for young children. Also, induction-loop systems are vulnerable to electromagnetic interference and general performance may be affected (see also the comment below on test box responses). For young children, vulnerable young people, and adults, a proprietary battery lock should be fitted (see ‘Further reading’).

As part of routine checks, such as listening to hearing devices prescribed to others or evaluating hearing in a soundfield, it is essential to protect the tester’s ears with appropriate attenuation, for example, a stetoclip with a variable attenuator or ear plugs and ear defenders. This type of personal and protective equipment (PPE) needs to be considered alongside hand hygiene, and infection prevention and control products.

Test boxes are heavy, sotake care when lifting and handling them. It is useful to consider wheeled cases for transport and appropriate care of components when transferring between venues. You do not want things to move about and get damaged.

Subjective checks

It is important to first listen to the whole system together and then to its separate components when troubleshooting. The whole system comprises the hearing device and its connection to a receiver and an associated transmitter:

hearing device receiver transmitter

All parts of the chain of communication should be checked with a stetoclip or monitor earphones (device dependent) (see ‘Further reading’).

Electroacoustic checks

A visual and listening check should always be carried out before test box measures are made. Regular checks compare the frequency response curves with baseline settings provided at the time of set-up.

The most common measures to record for hearing devices are responses to sounds at soft, conversational, and loud levels; and to check for distortion. The aim is to make regular comparisons against a baseline recording to see if performance is still appropriate. For tests 1 and 2, use the international speech test signal (ISTS) or an equivalent speech-like signal. On the Fonix FP35, DIG SPCH is required.

Frequency responses in SPL (sound pressure level) should be recorded as 50, 65, and 80 dB.
The test box should be set to GAIN and curves repeated at 50, 65, and 80 dB.
Total harmonic distortion (THD).

The most valuable test is GAIN; given time pressures in the field, it may be best to focus only on Test 2. Gain curves are also used for ‘balancing’ remote microphones with hearing devices (see section below).

For the set of curves in test 1, the response to 80 dB will be the highest curve and the 50 dB curve the lowest. The 65 dB curve will be in between. Check that these curves match baseline records.

For the set of curves in test 2, the response to 50 dB will be the highest curve and the 80 dB curve the lowest. The 65 dB curve will be in between. Check that these curves match baseline records.

What if the curves are different?

Replaceable microphone filters should be changed regularly. Your visual and listening checks before using the test box would have revealed actions to be taken before testing. For example, moisture or wax (or worse) may be occluding the tubing connected to the HA, or the filters on the CI processor can be blocked.

Does the user notice things sound differently?

Despite remedial action, if performance is degraded then refer to the user’s local audiology or auditory implant service. If there are contraindications to testing in the first instance, for example, noticeable infection or acute odour, then parents/carers should be asked to seek medical advice.

Test 3: As speech noise is made up of multiple frequencies, it cannot be used as an input to measure distortion. Total harmonic distortion THD is a calculation between the fundamental frequency and the intensity of subsequent harmonics. THD can only be measured with a tonal stimulus as each tone is presented with equal energy. Ideally, distortion will be below 10%. However, for digital signal processing (DSP) devices, THD may appear artificially high (15% or more as harmonics are given greater energy than the fundamental in channels with DSP). When measuring distortion, take a note of the reference level (even if it is high) as the level should be repeatable over time (the same DSP). Where there are large changes between periods of measurement, refer back to the service that fitted the device.

Different test box options

There are three common test boxes in use today in the UK (others are also available): Frye FONIX FP35, Affinity systems, and Aurical Hearing Instrument Test (HIT).

The Affinity and Aurical systems meet modern requirements for testing hearing devices (BSI 2021; 2022) and have the ISTS. The FONIX FP35 Analyzer needs to have the latest software to be able to select ‘Aid Groups’ as ADAPTIVE and to select the test signal DIG SPCH (digital speech). DIG SPCH is an interrupted composite signal and is not equivalent to ISTS. In addition, measurement tolerance is higher with the FP35 (see tolerance section below). The latest FP35 software is also needed to test modern HA fittings with an Open Fit Coupler.

Telecoil (t-coil) responses are known to be reduced in the lower frequencies to avoid interference in this area (Putterman & Valente, 2012). Therefore, in all test boxes, induction-loop RM system curves will be reduced in the lower frequencies and the response curves are more likely to match above 1 kHz.

Other equipment needed

Dedicated monitoring/listening components are necessary for some hearing devices. The Connevans DCTEST4 lead and earpiece (button receiver) allows baseline measurements to be made in a HA test box. It can also be used with Advanced Bionics Naida CI, Naida CI M, or Sky CI M processors with their dedicated listening checks; and Cochlear N7 and N8 processors. It is also compatible with the MED-EL SONNET and RONDO processors, but additional accessories are required (see ‘Further reading’).

Balancing (acoustic transparency): verification and validation with RM systems

We verify to ensure that the level and quality of sound a child hears through the RM matches what they hear through the microphone of the sound processor. If the gain of the system is too low, it will make little difference to the user and their access to spoken information is disadvantaged. They can soon become disenchanted and reject the resource. If the gain is too high, compression in the whole system could lead to distortion.

People with sensorineural deafness have a reduced dynamic range between the softest sounds they can hear and the loudest sounds that are comfortable to hear. Suitable levels are crucial, even more so for CI users because the dynamic range of cochlear stimulation is further reduced. Dynamic range has implications for both speech perception and sound quality. The normal-hearing dynamic range is between 0 dB and 120 dB. The hearing device reduces sound detected by its microphones to an input dynamic range (IDR), ie very soft sounds are omitted and loud sounds are compressed. IDR is manufacturer dependent. In a CI sound processor, this is compressed to about 40 current units, and further signal processing produces a narrow electrical dynamic range for each electrode channel (Wolfe & Schafer, 2015).

Balancing RM systems with HAs and BCHDs

Actions for proprietary microphones (generally manual gain or non-dynamic systems)

The signals presented to the hearing device and to the RM are equal and at 65 dB SPL.

Device	dB SPL level to device	dB SPL level to RM
HA or BCHD	65	65

Record the frequency response of the hearing device to a speech signal of 65 dB.
Connect the receiver (if necessary) to the hearing device and activate/pair the RM.
Record the frequency response of the whole system to a speech signal of 65 dB to the RM.
Check that responses 1 and 2 are within 2 dB and within the tolerance value of the test box (see ‘Tolerance’ section below).

For BCHDs, specialist equipment is needed. A clinical skull simulator is the most accurate device, but proxy measures can be recorded with a kit provided by Connevans.

Although responses may be elevated in the test box using the Connevans kit in comparison to clinical results, they are comparative with the same hearing devices across test boxes. For example, frequency response of a Baha 5 sound processor in an Affinity system and a Fonix FP35 were found to be within 1 dB with the Connevans kit.

Skull sim:

Connevans kit:

Actions for dynamic radio aid aids, eg, Phonak Roger systems

The signals presented to the hearing device and to the RM are equal and at 65 dB SPL.

Device	dB SPL level to device	dB SPL level to RM
HA or BCHD	65	65

Record the frequency response of the hearing device to a speech signal of 65 dB.
Connect the receiver (if necessary) to the hearing device and activate/pair the RM.
Activate Verification Mode* on the transmitter (RM).
Record the frequency response of the whole system to a speech signal of 65 dB to the RM.
Check that responses 1 and 2 are within 2 dB and within the tolerance value of the test box (see ‘Tolerance’ section below).

* Verification Mode is required as the transmitter is not able to act dynamically to current test signals. In order to fix this variable, Verification Mode sets the system gain at 10 dB. In standard mode, the system can provide up to 20 dB of gain.

Balancing RM systems with CI Sound Processors in non-clinical settings

Device	dB SPL level to processor	dB SPL level to (RM)
CI sound processors	65	65

An acoustic transparency method with signals of equal intensity can be used with all hearing devices, even with non-linear signal processing (Husstedt et al, 2022).

Electroacoustic test signals in quiet clinical settings

Some CI processors have first-stage compression that can be observed. Although signals of 65 dB SPL can be used, it may be useful to present a softer signal to those processors where possible (Whyte, 2019).

CI processor	dB SPL level to processor	dB SPL level to (RM)
All Advanced Bionics processors	65	65
Cochlear Nucleus 6 and above	55	55
MED-EL SONNET and RONDO series	55	55

It is thought that signals of equal intensity will be similarly compressed by implant sound processors, and a signal below the compression level may not be necessary (Husstedt op cit). Further research is being conducted to establish the suitability of signal levels through qualitative methods and behavioural testing.

Tolerance

Adjust the volume/gain of the RM, or ‘FM advantage’ or ‘EasyGain’ level of the radio aid receiver so that the combined system response curve matches the sound processor response curve and achieves ‘transparency’ or ‘balance’. If the difference is 2 dB larger than the measurement tolerance, then two output signals are different and transparency is not achieved; therefore, continue adjusting. For example, in the Aurical HIT, if the two responses are within 3 dB, then transparency is achieved.

Adjusting the receiver gain should preferably be done by starting at a low gain and then increasing.

Tolerance or signal accuracy up to 5,000 Hz	Transparency difference overall
Affinity < ± 1.5 dB	± 3.5 dB
Audioscan ± 1.0 dB	± 3.0 dB
Aurical ± 1.0 dB	± 3.0 dB
Fonix FP35 ± 2.5 dB	± 4.5 dB

Transparency or balance can be judged by eye, by offset calculation, or by root mean square (RMS) values or calculations (see GPGQS8 – ‘Auditory implants’ for details of calculations. Worked examples are available in the ‘Example resources’ section).

Validation

Following fitting and verifying, the balance (electroacoustic transparency) of the hearing device and RM system must be validated. It is vital to assess speech recognition in quiet then in noise, with and without the RM system.

Ideally, use an adaptive test to evaluate speech recognition performance in noise, which will avoid ceiling (100%) and floor (0%) effects that may occur with fixed stimuli.

Alternatively, use an SNR test using signal levels that represent the most common environment where the wireless remote microphone (WRM) is to be used.

“Measures of benefit must not only include objective measures, but also patient perception of benefit” Clark et al (2017:617).

Consider using:

observation
behaviour
functional listening
questionnaires, eg
- LIFE (Nelson et al, 2020)
- COSI (Dillon et al, 1997)
  - COSI – Child

Clark, J. L., Pustejovsky, C., & Vanneste, S. (2017). Objective and perceptual comparisons of two Bluetooth hearing aid assistive devices. Disability Rehabilitation Assistive Technology, 12(6), 614–617. https://doi.org/10.1080/17483107.2016.1201153

Useful resources

Testing Advanced Bionics (AB) processors in the Aurical HIT Box: AB_AURICAL

Testing Cochlear processors in the Aurical HIT Box: Cochlear_AURICAL

Testing SONNET processors in the Aurical HIT Box: SONNET_AURICAL

Updated – Which test lead do I need? – May 2023: Updated – Which test lead do I need – May 2023

Video resources and other websites:

Test box video demonstrations (Ewing Foundation)

Testing a hearing aid with the Aurical HIT (Connevans)

Setting up a Phonak Roger Touchscreen Mic using the Aurical HIT (Connevans)

Checking a Phonak Roger Pen using the Aurical HIT (Connevans)

Setting-up an fmGenie using the Aurical HIT (Connevans)

FP35 Hearing Aid Analyzer PowerPoint Presentation

Guidelines for fitting hearing aids to young infants

Considerations for QToDs

A Phonak Receiver Audio Checker and adapters are very useful for doing listening checks with Phonak Roger X receivers (additional adapters are required for the range of other Roger receivers).

AB_AURICAL Cochlear_AURICAL SONNET_AURICAL Updated – Which test lead do I need – May 2023

It is important to maintain up-to-date software and firmware

Hearing device settings

Has the hearing device been set up to use the RM? What settings need to be present or activated?

For children and young people, mixing ratios of the processor microphone and aid radio aid should be 1:1 or 50:50.

If a loop system is to be used, does the telecoil need to be enabled?

Are there features active in the hearing device that may affect performance in the test box, eg directionality or phone detection? Remember to do like-for-like comparisons against baseline recordings.

6.6. Test box measures

Introduction: