“It is innate and natural that audio signals are part of the Human Machine Interface (HMI) of medical instruments. Whether it is for ordinary home sphygmomanometers, or for various complex instruments in a formal hospital setting (such as pulse monitors, cardiac waveform (EKG) devices, infusion pumps and respirators, as well as ventilators, oximeters, etc.) , audio signals are important to inform users of patient conditions, trends, critical/risk situations, and equipment health.
By Bill Schweber
It is innate and natural that audio signals are part of the Human Machine Interface (HMI) of medical instruments. Whether it is for ordinary home sphygmomanometers, or for various complex instruments in a formal hospital setting (such as pulse monitors, cardiac waveform (EKG) devices, infusion pumps and respirators, as well as ventilators, oximeters, etc.) , audio signals are important to inform users of patient conditions, trends, critical/risk situations, and equipment health.
However, various types of devices emit their own sounds, and these sounds combined can lead to misunderstandings, confusion, or missed alarms, or even life-threatening errors in emergency situations. To unravel these jumbles of audio, ISO/IEC 60601-1-8 builds a framework that defines what sounds (in pitch and sequence) should sound from medical electrical equipment, and under what conditions. Situations range from routine functionality and continuous monitoring, to critical alert situations and more. These sounds can range from basic sounds from a buzzer to more complex audio sequences such as melodies or compositions, or even voice messages.
As device manufacturers continue to integrate more functions into each medical electrical device, they also need to make provisions to incorporate more types of warning sounds. The task of the circuit design engineer is to ensure that the appropriate hardware (such as speakers or sirens, drivers/amplifiers and their physical mounting) is provided to create the desired specified sound patterns to provide consistency and avoid ambiguity in these common stressful environments .
This article will not delve into the complexities of IEC 60601-1-8; like all IEC standards, this standard is complex and requires careful study of its many requirements and exceptions. This article discusses the use of basic sirens and speakers. The article will take the products of Mallory Sonalert Products and PUI Audio as examples to explain how to use such products to meet the hardware requirements of the standard.
ISO/IEC 60601-1-8 Basic knowledge of alarm systems
The document ISO/IEC 60601-1-8 “Medical Electrical Equipment C Parts 1-8: General Requirements for Basic Safety and Essential Performance” is a detailed 71-page standard detailing various performance requirements as well as alarm systems related tests. Note that although much of the standard deals with audible alarms, these can be either visual or audible. The standard does specify specific melodic patterns, mnemonic lyrics, and in many cases the rationale for mapping melodies to alarms (Table 1). Various industry experts have even published audio files with representative examples (see the references provided by the University of Sydney, Australia for related examples).
Table 1: The IEC 60601-1-8 standard contains melody patterns, some sources add mnemonic lyrics and the rationale for associating melody with alarm lyrics. (Image credit: Penn State University)
Given the complexity of modern medical electronics, it is impossible to develop a “best” solution for one or several sounds in any case, as shown by the acoustic and cognitive problems summarized by the researchers (Table 1). 2).
Table 2: Each sound and sound mode has acoustic and cognitive problems, and these problems can vary between individuals and environments. (Image credit: U.S. National Library of Medicine, National Institutes of Health)
Alerts can be divided into two broad categories: physiological alerts related to patient conditions, and technical alerts related to equipment status. The latter covers situations such as low battery, disconnected wires, or kinked lines.
Although it is important to get the attention of medical personnel, appropriate alert and emergency levels must be set. Obviously, when the battery has 20 or 30 minutes of charge remaining, they’re not as critical as if they had only a minute or two left. For this and other reasons, IEC 60601-1-8 defines three different hazard classes:
・ DANGER: Indicates a hazard with a high risk level which, if not avoided, will result in death or serious injury
・ Warning: Indicates a hazard with medium risk which, if not avoided, could result in death or serious injury
・CAUTION: Indicates a hazard with a low risk level which, if not avoided, could result in minor or moderate injury
One of the many goals of the standard is to match the audio produced to the hazard level, so as not to cause unnecessary hazard indications in case of caution, but also not to downplay or mislead the real hazard.
As a comprehensive specification, the standard also covers what types of medical conditions should trigger an audible warning sound. Which not only defines specific frequencies, rise/fall times, waveforms, sound levels (in decibels (dB)), pulse widths, repetition rates, and harmonics for each sound, but also leaves room for equipment manufacturers A certain amount of flexible space.
For example, the standard requires that the fundamental frequency (pitch) of a single sound pulse must be between 150 and 1000 Hz, there must be at least four harmonics (overtones), and that the amplitudes of these harmonics must be within 15 dB of the amplitude of the fundamental frequency inside (Figure 1).
Figure 1: The IEC 60601-1-8 standard requires that the fundamental frequency of the tones lie in the range 150 to 1000 Hz, have at least four harmonics, and have an amplitude that is within 15 dB of the amplitude of the fundamental frequency. (Image credit: Mallory Sonalert Products)
The standard also focuses on the harsh reality of the healthcare environment: the incidence of false positives. According to records from outside researchers, the incidence of false positives varies from 10% to 90%, and in some cases as high as 90%. For too many false alarms, the normal response of the relevant personnel is to disable the alarm function so that it no longer sounds, which is allowed by the IEC standard.
There is also the issue of “acoustic masking” given the many possible audio sources, alarms and sounds. In this case, due to the limitations of human senses, concurrent alarms can cause one or more of the alarms to be inaudible. It’s the same reason that the observer can’t see something because of visual overload and confusion. One solution would be to use real spoken text in addition to tones or sound patterns for exceptionally critical high-level alerts, which are likely to stand out in the mix of voices.
This risk of cover-up is one of the reasons why many cockpits (alarm environments similar to operating theatres or intensive care units (ICUs)) in modern aircraft use short, jerky voice messages to warn of dangerous situations. Aircraft warnings include: “Pull up! Pull up!”, “Caution, terrain!”, “Stall Imminent!”, “Windshear! Windshear!”, “Traffic!Traffic!”, and “Descend!Descend!” (see Aircraft Owners and Pilots member associations and Wikipedia references).
Start with a basic beep
For simple single-function medical devices, such as home blood pressure monitors for non-technical casual users, there is little or no need for complex audio outputs. In this case, the audio indications are simple beeps to indicate conditions such as “device incorrectly placed”, “device problem” (may include low battery warning), and “read complete” .
These common needs can be met with a basic internally driven magnetic buzzer like Mallory Sonalert Products’ ASI09N27M-05Q (Figure 2). This Surface Mount Technology (SMT) device measures 8 × 9 mm, is 5 mm high, and operates from a single 3.0 to 7.0 V supply (5 V nominal). When operating at nominal supply voltage, the product emits an audio tone of 2700 ±300 Hz, has a sound pressure level of 80 dB at 10 cm, and consumes 30 mA of current.
Figure 2: Magnetic single-tone buzzers with internal drive circuitry, such as the ASI09N27M-05Q, are easy to use and suitable for some simple medical device applications. (Image credit: Mallory Sonalert Products)
With internal drivers, no external audio sources or waveforms are required. Only a gated DC voltage is required to operate, and even low-side discrete transistors can be used to switch the voltage source and current. Although the unit operates at a fixed fundamental frequency, if placed in an appropriately sized resonant enclosure, it can also generate harmonics (up to the fourth overtone) acceptable to applicable standards.
Play Songs, Melodies and Voices with Speakers
Many types of medical Electronic devices require more complex tonal sequences and melodies than can be generated by a basic single-tone buzzer. This also applies to non-mandatory voice alerts. In these cases, speakers (or just “horns”) can produce sound with frequency components spanning part or most of the audio band (often considered 20 Hz to 20 kHz) with “reasonable” to “very good” fidelity and low distortion.
The sound pressure level (SPL) delivered by these speakers depends on frequency, speaker efficiency, and drive signal level. There are various styles of speakers on the market, with different sizes, frequency response curves, packaging, connections and durability levels. Almost all speakers have a nominal impedance of 4 Ω or 8 Ω.
For example, PUI Audio’s general purpose AS02008MR-5-R loudspeaker is an 8 Ω loudspeaker rated at 500 mW (up to 800 mW), delivering up to 86 dB SPL at rated power levels (Figure 3). Its 3 dB bandwidth ranges from 500 Hz to 4 kHz (at 5% total harmonic distortion (THD)), covering the portion of the spoken-voice frequency band required for clear speech. This ultra-thin small speaker measures 20 mm in diameter, 3.80 mm in height and weighs 2.4 grams. The speaker cone is constructed from polyethylene terephthalate (PET) material and uses powerful NdFeB magnets to achieve this performance in a small and lightweight package.
Figure 3: The general purpose AS02008MR-5-R loudspeaker is a small, low profile loudspeaker that provides the loudness and bandwidth needed to understand voice alerts and messages. (Image credit: PUI Audio)
For applications requiring higher fidelity and stronger frequency response, choose PUI Audio’s AS03208MS-3-R. This is an 8 Ω general purpose loudspeaker that can handle up to 3 W of power in the frequency range from 200 Hz to 20 kHz (90% of the audio band) and deliver up to 85 dB SPL (Figure 4).
Figure 4: For higher fidelity, the AS03208MS-3-R 8 Ω speaker provides response as low as 200 Hz and as high as 20 kHz. (Image credit: PUI Audio)
This speaker features a rubber paper cone ring and a square non-resonant cone frame. The speaker enclosure is IP65 rated, so it’s dust-proof and protected from water intrusion from the nozzles (but not completely waterproof). This level of protection is required in some medical environments (Figure 5).
Figure 5: The AS03208MS-3 loudspeaker includes a square cone frame and a rubber-paper cone ring, so it is dust and splash proof and IP65 compliant. (Image credit: PUI Audio)
The AS03208MS-3-R loudspeaker measures 32 × 32 × 16.5 mm, requires only 1.5 mm of free cone space, and includes contacts for connecting discrete leads.
Good Audio: More Than Just the Speakers
Selecting the most suitable speakers is only one aspect of the design challenges that need to be addressed to meet the desired audio performance. Speaker mounts and enclosures are also important factors. Install the speaker so that it forms a seal along the outer edge of the basin frame. Back-to-front pressure wave cancellation occurs when sound waves from the front of the speaker diaphragm/cone interact with sound waves from the back of the speaker diaphragm/cone, and this mounting reduces this cancellation. In the critical region below 1 kHz, this cancellation is more likely to be a problem.
Another key specification for speakers or other audio output sources is the self-resonant frequency. This means, among other things, that loudspeakers are the most efficient at converting electrical input power into actual sound pressure levels. Mounting the loudspeaker in an enclosure improves the performance of the loudspeaker at or below the loudspeaker’s resonant frequency. For AS02008MR-5-R, this resonance frequency is 500 Hz ±20%, so it has good low frequency performance. The self-resonant frequency can also guide the design of the enclosure to avoid unwanted hum and click from the speaker due to mechanical resonance with the speaker itself.
In addition, attention needs to be paid to the electrical input power level and waveform. Speakers have two power ratings: average (continuous) power rating and maximum power rating. When the signal sent to the speaker is not a sine wave, the power may exceed the maximum power rating specification. The power rating can be determined by a simple formula:
Power = (peak voltage)2/impedance value
Exceeding this power rating (which is different from the maximum instantaneous power used by voice or music power) can cause damage to the speaker over time, including broken voice coil brocade leads (voice coil burnt, resulting in an open load resistor), Or the voice coil bobbin (the rigid cylinder around which the voice coil wire is wrapped) deforms, locking the voice coil in the magnetic motor.
Accelerate Speaker Evaluation with Kits
Buying multiple speakers, connecting them to an amplifier, and evaluating their audio performance and mechanical fit can be confusing and can be time-consuming, especially for voice alarms. To simplify this task, several evaluation kits are available, such as PUI Audio’s 668-1692-KIT (Figure 6). The kit contains eight conventional speakers with different power ratings and impedances (4 Ω and 8 Ω).
Figure 6: The 668-1692-KIT Audio Amplifier and Speaker Kit contains several different speaker types (with different sizes, power ratings, and impedance values) to speed up the evaluation of speakers in the final application. (Image credit: PUI Audio)
The kit also includes PUI Audio’s ASX02104-R exciter, a 4 Ω audio generator with a diameter of 21 mm and a height of 8.5 mm. The exciter has a rated input power of 250 mW and an average sound pressure level of 72 dB, covering the frequency range from 640 Hz to 10.5 kHz (Figure 7).
Figure 7: The ASX02104-R exciter in the 668-1692-KIT evaluation kit is more than a loudspeaker as it contains the resonant chamber and sounding enclosure. (Image credit: PUI Audio)
The exciter is a self-contained audio source that avoids some of the challenges of using speakers. This is because there is no need for a resonant chamber, there is no need to worry about environmental damage to the speaker, and there is no need to change the appearance of the product to meet the need for speaker holes. It is driven in the same way as a speaker and is waterproof and dustproof, further enhancing its suitability for certain medical devices.
To enable active electronics to drive speakers or exciters, the kit also includes PUI Audio’s AMP2X15 audio amplifier board (Figure 8). This Class-D audio amplifier provides single-channel (mono) or dual-channel (stereo) audio, delivering 15 W per channel into an 8 Ω load. The evaluation board measures 76.2 × 50.8 × 20 mm and operates from a single 9.5 V to 20 V supply.
Figure 8: The 668-1692-KIT evaluation kit includes the AMP2X15, a complete off-the-shelf dual-channel Class-D audio amplifier capable of delivering up to 15 W per channel. (Image credit: PUI Audio)
At the heart of the AMP2X15 is Texas Instruments’ TPA3110D2 Class D amplifier IC for maximum signal fidelity (Figure 9). This 28-lead HTSSOP IC uses a unipolar 16 V DC power supply and is capable of delivering 30 W into a mono 4 Ω load and 15 W per channel into an 8 Ω load, although it can operate between 8 and 8 Ω. Operates on a 26 V DC power supply.
Figure 9: The amplification and other functions of the AMP2X15 audio amplifier board are provided by the TPA3110D2 Class-D audio amplifier IC, which supplies power to the speaker load for low distortion and high energy efficiency. (Image credit: Texas Instruments)
Without a doubt, understanding and meeting the complex and trivial requirements of the IEC 60601-1-8 standard for audio alarms for medical electrical equipment and systems can be a daunting challenge. Even experts have different opinions on how to implement the standard guidelines, and which sound patterns and types (such as buzzer, melody, or voice) are best for each scenario and user situation.
Fortunately, the hardware side is relatively straightforward when it comes to implementing audible alerts. There are many high-performance, easy-to-use small buzzers and speakers on the market, giving design engineers a variety of options. These products not only have clear performance attributes, but also minimize design-in challenges. Therefore, when adding audio capabilities to medical devices with these products, the hardware side becomes fairly straightforward as long as basic audio and mechanical guidelines are considered and followed.