Excessive Noise Level, Alerting Device, Noise Alarm

Excessive Noise Level Alerting Device (ENLAD): US Patent 7315244

The ELC Audio Engineering Excessive Noise Level Alerting Device (ENLAD) is a patented (US Patent 7315244) alarm / monitoring system designed for use in hospitals, offices, or any environment where excessive noise levels are of concern. The ENLAD is not a sound level meter (SLM) or noise dosimeter that simply triggers once a sound pressure level (SPL) or equivalent noise level (Leq) is exceeded. Instead, the ENLAD uses several conditions / variables to trigger the alarm so that false alarms are kept to a minimum. To illustrate, here is an example: The sound resulting from the accidental dropping of a tray within the alarm's perimeter will most certainly be louder than the ENLAD's pre-set threshold, but because the duration of the event is short and occurs only once, the alarm is not triggered. Similar, single-occurrence events are generally not considered “preventable noises” and, therefore, do not trigger the ENLAD. In comparison, a dosimeter will trigger an associated alarm if the energy level of a single, transient event (such as dropping a tray) is equivalent to or greater than the energy level of quieter, longer-duration events. Other non-preventable noises include air flow noise emanating from AC or cooling vents. This type of noise is typically low-frequency, steady-state sound: Selective filtering prevents alarm triggering from such noises. When it comes to sound level meters, a choice of weighting (A or C) and sound level meter ballistics (FAST or SLOW) may help prevent some false alarms, but a sound level meter's SLOW response will miss repetitive transient events such as hand clapping or a ball bouncing (events deemed avoidable) and the FAST response will result in too many false alarms. This is explained in more detail below.

Notes regarding studies that are designed to measure a correlation between physiological parameters and adverse or objectionable sounds

A number of researchers have contacted me about the detrimental effects of noise (apart from hearing loss) on patient well-being and how "objectionable" noise can be monitored or curtailed. One of many possible deleterious effects is increased morbidity in highly stressed, fragile infants who are susceptible to noise. Noise is most certainly a cause of stress, and stress can be indirectly quantified using physiological measures (heart rate, blood pressure, and respiration being examples). Another parameter regarding stress and infants would be number of days in an NICU (this alone might summarize the physiological parameters?). I imagine researchers would use an independent samples t-test, or compare their group (sample) means to averages obtained from other NICUs or other patient areas. It also seems feasible to compare treatment to non-treatment, and then look at results of no treatment (no alarm) where a treatment (alarm) has already been implemented but then removed. I do not claim to be an expert on the statistical methods needed for such studies. I am, however, an expert when it comes to sound and its measurement...

When discussing the purchase or installation of a sound level alarm with an acoustic/sound engineer or technician, here are some important points to consider. Most people who are involved with noise surveys are concerned with noise in the workplace and whether workers are at risk for hearing loss or should wear personal hearing protection. Sound level meters and dosimeters are, then, appropriate because they are usually designed to measure the average noise level, peak levels (in some instances), and noise dosage or dose. HOWEVER, some studies involve physiological responses (such as heart rate) and psychological stress, and the researcher's concern is NOT one of hearing loss, but patient well-being. The use of an alarm based on conventional sound level meters that measure peak level or average sound pressure level, as well as noise dose (or Leq which means “equivalent level”) can result in too many false alarms. When there are a lot of false alarms, people start to ignore the alarm altogether. It is for these very reasons that I designed and patented the ENLAD alarm circuit.

Here are examples to provide to your sound engineer or technician: One of the most objectionable noises in the hospital environment is talking, especially ongoing conversation. The talking may not be particularly loud, but a patient’s ability to sleep is compromised or his/her heart rate is affected by constant chatter. Setting a sound alarm at typical voice levels will result in false alarms because there are many transient events that exceed the level of human-generated speech. Transient events are often events that cannot be avoided and, therefore, should not trigger an alarm. Most persons involved with noise surveys or OSHA compliance in the workplace would suggest using meter ballistics that give an average reading in lieu of a peak reading (the SLOW response mode on a sound level meter is generally specified in noise survey guidelines). The SLOW, or average reading/response, helps some with false positives, but it misses objectionable, repetitive, short-duration events such as someone bouncing a pencil on a desk or clapping his/her hands. Speech, too, can have loud and soft peaks that are “flattened out” by a sound level meter (SLM) set for SLOW response. A FAST response catches transient events, but then single, short duration events (often unavoidable) would be triggered by an alarm based only on sound pressure level (SPL, expressed in decibels).

Most sounds that are preventable are human-generated and have a pattern (talking, bouncing a ball, clapping hands, etc.), whereas "accidental" sounds (such as dropping a tray) occur once or at highly irregular intervals. A dosimeter-type of alarm will trigger with a single noise if it's loud enough because "dosage" is a measure of energy. A single loud sound (albeit an accident) has as much or more energy as a moderately loud (intense), longer-duration sound. This is why dosimeters tend to give "false alarms" in the NICU. Unless the concern is hearing loss, or it doesn’t matter whether the sound is accidental (tray dropping) or preventable (such as talking), alarm triggering based on dosage is not be the best solution. Just remember: Too many false triggers/alarms will result in personnel ignoring the alarm.

My patented sound alarm doesn’t rely solely on average SPL, dosage, or a SLOW or FAST response. Instead, it momentarily holds a warning "flag" (held in memory) for any sound event that is louder (technically, more intense) than a user-defined, predetermined SPL. The sound "event" could be the loud part of speech (not terribly loud) or a tray dropping (very loud). The warning flag circuitry initiates a timer, but it does NOT trigger the alarm. If the peak occurs only once or twice, the timer resets itself. However, if the warning circuit stays HIGH (flag up) for an appreciable amount of time, this means that the sound is ongoing (such as speech) or that the transient event is occurring at regular intervals. In such circumstances, the alarm’s timer does not reset itself before timing-out (as it would with a single event) and the alarm is triggered after a user-specified time (e.g., 5 seconds of talking or 5 seconds of clapping). This is NOT the same as dose or average response; it is simply a way of looking at sounds above a certain threshold, steady-state or transient, and seeing whether these sounds persist for a period of time (5 seconds being an example). The advantage to all of this is that single, loud "accidental" events do not trigger an alarm while objectionable, ongoing noises will trigger the alarm. Examples of noises that would be considered objectionable and will trigger the ENLAD include: Human voice levels above a preset threshold and for a time greater than, for example, 5 seconds; repetitive transient events (e.g., hammering or clapping one's hands repeatedly); or steady-state machinery noise above a preset threshold. Because some types of noise in hospital rooms are unavoidable (e.g., ventilators), the ENLAD uses a remote sensor (wired remote, not RF--see Footnote) that can be mounted on walls or ceilings away from such noise sources. The primary intent of the ENLAD is to provide a visual warning via a flashing light or sign in response to noise and noise levels that are deemed unnecessary or preventable.

A brief technical description of my original Excessive Noise Level Alerting Device (ENLAD) is given below. Note: Changes to the circuitry do not preclude infringement on patent rights or privileges. The latest version is implemented via programmable ICs, but the fundamental operating principles remain unchanged.

The ENLAD is comprised of a remote (wired) sensor unit, a power supply/timer unit, and an alerting sign. The sensor unit includes an omnidirectional electret microphone, a 2-stage preamplifier, a triggering device (Schmitt trigger IC), and a switching transistor. More than one sensor can be connected to a single timer/power supply module. An optional LED on the sensor illuminates in response to transient signals and is independent of the alerting sign. The timer unit consists of electronic circuitry that is housed with the power supply. AC or DC power is easily supplied to any number of alerting signs that are connected to the power supply. Unlike the optional LED, the primary alerting sign does not illuminate in response to transient events unless they're repetitive in nature; instead, the alarm flashes (or logs the event) when the noise level exceeds 75 dBA (75 dBA being typical; this level is user adjustable) for a period of five seconds or longer. Please note: Both the threshold SPL and time period are user-adjustable. This feature, in addition to selective filtering, helps prevent false triggers that would otherwise occur as a result of accidental transient noises or low-frequency background noise.

The power supply/timer unit consists of a regulated DC power supply, a timer unit, and an optically-isolated solid-state switch. The timer circuit uses a transistor and an astable multivibrator IC to set the delay time and the ON/OFF cycle that follows. When the transistor is forward biased for a period greater than td (the delay time), the alerting sign begins flashing at a rate t1 + t2 (the sign is ON for a period t1, then OFF for a period t2, and so on). The delay time td and duty cycle t1 + t2 can be adjusted, as can be the sensor's trigger level.

The output of the timer circuit is connected to the input of an optical triac driver. The triac driver, along with a high-voltage triac, serves as an electrically-isolated, solid-state switch that optically isolates the higher-voltage alerting sign from sensors located in patient rooms. Any lamp or alerting sign, AC or DC powered, can be used for a visual alerting device. Data logging is an option when a visible alerting sign would be objectionable.

The sensor units are initially calibrated using speech-weighted noise (USASI) as the noise source. Each sensor's trigger level can be adjusted via a multi-turn potentiometer. A precision (Type 1) sound level meter is used to pre-set the trigger level at 75 dB SPL. It could be observed that the LED on the sensor unit's LED flickers in response to noise levels slightly below 75 dB SPL, but 75 dB SPL is needed to keep Q3 (re schematic) "ON" for a period equal to or greater than td.

Two additional features will be available as options. One option is a digital counter that counts the number of times the noise level has exceeded 75 dB SPL (nominal) for a period greater than five seconds. The second option records "excessive" noise using a ChipCorder™ IC or other types of solid-state memory. Either option may be useful for research purposes or identifying the sources of objectionable noise.

The physical dimensions of the power supply unit vary depending on the number of signs it will power and the number of sensors that can be attached to it. The power supply unit plugs into a standard 120 volt AC outlet via a 3-conductor plug. At present, the sensor modules are housed in a lightweight plastic chassis measuring 4 x 3 x 1.5 inches. Their light weight makes the units ideal for ceiling mounting. They are connected to the power supply unit via a durable, but thin and flexible shielded cable. The cable is about 3/16” in diameter (inclusive of shield and insulating jacket). Each sensor units connects to a power supply unit via a 4-conductor plug/socket arrangement.

Footnote: While it would be easy to implement a wireless remote sensor, RF or IR radiation is not desirable in many settings (e.g., hospitals). The shielded cable leading to a sensor carries very low current and 5 VDC.