Willemstan I-SHAT System

Infrastructure for Severe Harm Alert Technology

This article was last updated on: Semptember 26th 2022

Charlie Nolan's dog just chilling as the SHART-1 goes off.

"I don't know I just like building big powerful things that are borderline illegal."

- Charlie Nolan in the Willemstan documentary on the SHART series.

Lore

The Willemstan I-SHAT system (Infrastructure for Severe Harm Alert Technology) is a concept that has been used in many countries- a centralized way to send important messages to civilians. This includes things like civil defense sirens, radio broadcasts, and phone notifications. This can notify citizens about emergencies or other important news. The I-SHAT system currently includes Willemstan Government Alerts and the SHART sirens. It used to be integrated with the Live Camera Feed livestreams. The I-SHAT system is now partially automated, and is integrated with the SuS Bot and a local weather station. When it detects severe weather (such as high winds) a notification is sent to the operator of the SHART siren who can decide if the weather is worthy of sounding a signal indicating danger.

The SHART Series

The SHART series is a series of electronic civil defense sirens. Development on the SHART sirens began in early June, with building and tests of an early model, the SHART-1. The SHART (Severe Harm Alert Repeater Tower) series are electronic civil defense sirens, capable of battery backup, solar power, and their own tone generation, and are able to drive up to 160 watts of speaker output. The current version uses an 80 watt output. The SHART series generate their tones and provide an interface to the operator with an Arduino Uno, and drive speaker output through digital-logic H-bridge motor controllers. They get power from a 12 volt car battery. The SHART sirens are able to effectively double their output power by outputting an AC square wave signal to the speaker. Recent iterations of the SHART are weatherproof. Since early August, the unit has not been running daily tests due to fear of noise complaint, but will still be activated for monthly tests, around the same time as the county (1 PM on the first Saturday of each month). The most recent long-distance test confirmed that the siren is clearly audible at over 1,100 feet. The current SHART siren is capable of 17 signals in total, with customizable tone pitches and it is the only digital siren controller that mimics analog/electromechanical siren characteristics. If it were produced commercially, it would technically be the most versatile outdoor warning siren tone generator on the market, as no current siren controllers have 17 signals, support dual-tone, or analog-type windup sounds. The controller is also extremely efficient, by using digital-logic amplifiers, with no ambient power leakage, and very low standby power usage for the whole unit. The design and function of the SHART series was inspired by the analog Whelen sirens of the 1980s and Federal Signal's EOWS series, and the sounds were inspired by analog and digital Whelen siren controllers (such as the ESC-864), and electro-mechanical Federal Signal sirens of the 1960s to 1990s (such as the 3T22). The naming scheme is decided by the number of the major version, followed by the letter of the minor revision, followed by the maximum wattage delivered to the speakers.

SHART-1 Series

Prototypes, not meant for prolonged outdoor use. In videos, the output speaker is mounted on a few garden stakes. The SHART-1 used a 60 watt car speaker for its output, outputting about 45 watts.

SHART-1: It was programmed on June 13th, 2020. It was capable of 11 functional tones, and a 30 second test function. It was not weatherproof and battery life was not very efficient.
SHART-1A: The SHART-1A was given a weather-resistant housing, and an additional solar panel in order to make it capable of a very long life without any power line connections.

The SHART-1, in the flesh.

SHART-2 Series

The SHART-2 series used a new naming scheme. The 2 means it is a member of the SHART-2 series and the letter means the revision. The following numbers are maximum output wattage. The 2 series was more than an early prototype. The 2 series stayed outside exposed to the weather, as a test of reliability. The notable issues include dysfunctional daily testing until revision E, and the PA function failing to work throughout the entire lineup of prototypes.

SHART 2A56: The SHART 2 series increased output volume while not using any more power than before. The housing was also tweaked to make it more water resistant. The SHART-2 also used redesigned programming for more realistic tone generation.
SHART 2B56: The 2B56 had better water resistance, and some better tone generation. It used a lower frequency for better sound range. The new programming also addressed various audio glitches and the functionality of the test button. The 2B56 also had water-absorbent material added into the control box to prevent buildup of moisture. There was also a mount added for a solar panel to improve efficiency.
SHART 2C56: The 2C56 had some button functionality and tone generation changed. Tones were changed to be more distinguishable from one another. This model was in use during a day of heavy rain and survived with no damage. This model fixed the Airhorn tone which did not previously sound correct.
SHART 2D56: The 2D56 introduced many changes. The naming scheme was changed, to omit SHART from most naming as that just confuses people. The 2D56 was also upgraded to have tighter waterproofing so that less water will leak in. A circuit was added which supports control of a rotator motor, so that if desired, newer models can rotate their speakers for a wider area of output. Also, tone generation was changed to be more realistic.
SHART 2E56: The 2E56 made automatic daily testing functional and reliable, a feat which no previous versions found possible. It also fixed the growl tone which was previously not working.

The SHART-2D56.

SHART 3 Series

The 3 series used new speaker drivers, with an increased theoretical maximum output of 110.8 dB at 100 feet, and a calculated typical output of 109.4 dB at 100 feet, similar to that of some real civil defense sirens. Although this is output assumed in a best acse scenario weather-wise and scene-wise. It has been tested to be above ambient noise level at over 700 feet (originally incorrectly reported as 1000 feet).

SHART 3A56: The 3A56 was similar to the 2E56, except the output was fine-tuned so that it would be more effective with the new speakers. The 3A56 also got a functional PA mode, but with heavy distortion and a very quiet output. Siren signals were tested at 600 feet and it is calculated that the siren has an effective range of at least 1,500 feet from where the speakers are pointing.
SHART 3B56: The 3B56 made a number of improvements. It introduced dual-tone signals, lowered the pitch of single-tone signals (to improve coverage) and significantly improved the clarity and volume of the PA function. It also introduced the “Doomsday Chimes” tone and fixed the silent test mode. Dual tones signals offer better distinguishability but single tone signals offer better coverage. The 3B56 has been tested to be louder than ambient noise at 700 feet.

The SHART 3A56. The speakers are visible but the control housing is not.

SHART 4 Series

The 3B56 died an unfortunate death after filling with water some time between September 7th and September 9th of 2020. The audio amplifier and battery regulator were ruined and it was clear the SHART 3B56 was not waterproof. It was also horrifyingly filthy inside, and a large number of dead bugs and even a dead earthworm were found in the control housing. It can be determined that this was quite the bacteria brewery. Some electronics also had terrible corrosion. These and previous versions of the SHART also had problems such as low serviceability and very limited controls, which the 4th major increment of the SHART hoped to solve. It was also decided that a mount, increased and regulated speaker output, and smarter power usage would be some goals for the project. So it was decided that major overhauls had to be made for the SHART 4 Series.

SHART 4 early prototypes: The early prototypes used a system of buttons such as previous models, but were ultimately never created as real sirens, as the programming and control system were completely reworked after better options were explored.
SHART 4A80: The SHART 4A80 increased the maximum theoretical power output by 24 watts, by stepping up DC voltage before it enters the audio amplifier. This increased the output from 56 to 80 watts (40 watts per speaker driver), the maximum power that the speaker drivers can accept. Also unlike previous revisions, the speakers get 80 watts no matter how charged the battery is, so maximum output can be used unless the battery is entirely dead. This method draws up to 120 watts from the battery when the speakers are running, but vastly improves the output. The amplification system was also upgraded to use dual amplifiers (one per tone) which support a maximum of 10 amps of output continuous (179 watts) meaning that up to 8 speakers could be connected in parallel to the output. The 4A80 also added additional circuitry that would cut-off power from the DC voltage converter and amplifier if no tones are being played, reducing power consumption when idling. The 4A80 added a 4 foot tall wooden mount which could be rotated, improving coverage. The housing was upgraded to be truly waterproof and open from the side, and only have a hole in the bottom, to avoid the problem that killed the 3B56. The control scheme was improved with the use of a labeled keypad, and was located in the housing to prevent tampering and water infiltration. The internal circuitry was also laid out better to improve longevity and ease of repairs. Audio was also changed so that some tones no longer use a linear wind-up sound, and instead mimic mechanical or analog electronic sirens, for better recognizability. The 2-tone tones were also changed to use a 5 to 6 ratio primarily instead of 6 to 7, to sound like a perfect minor third. With the press of a button, tone ratios could be changed. The battery was kept in a different compartment than the control circuitry. There is a video of the control board running from a power supply with only one speaker driver attached.
SHART 4B80: The 4B80 revised programming for rotational support, and slightly changed the control scheme. Tone generation was also slightly modified. Solar charging was fixed by replacing a faulty component. Programming support was added for the addition of a plug-and-play wireless control module.

The SHART 4B80. Notice the separated battery housing (below) and circuit housing (on wooden beam) containers. It is also apparent that the wood used in construction of the unit became extremely warped.

S5 Series

Planning of the SHART 5 began in December of 2020, after the SHART 4 proved unreliable in cold weather, and the output could not even travel half a mile.

S5 prototypes: The SHART name was further shortened to just 'S', due to mockery of the name among nerdy 'siren enthusiasts' who did not understand the irony of the name. The control board was upgraded, from an Arduino Uno to an Arduino MEGA for improved tone generation, and more functionality. It was decided that some components, such as a DC step-down power converter and the solar charge controller would need to be stored in a battery cabinet, because they would not fit in the control panel. This would leave room for the main control box to house the Arduino, the control interface, the new SD card reader, the amplifier, and rotation motor control circuitry. The entire siren would be turned into a 24 volt system instead of a 12 volt system, as this would require less drain on the batteries when the siren was running, and thinner wires. The possibility of an upgraded output was explored, from 80 watts to 400 watts: a 9x increase in output with only a 1.5x increase in current consumption. The amplification system was also changed from 2 seperate amps to 1 amp, with one input being able to switch between the high tone and low tone to enable single tone or dual tone mode. A solar panel with a maximum output of 100 watts would be used to charge the batteries, as standby current increased from 40mA to an estimated 100mA, and the current solar panels seemed inadequate. A new mounting solution for the speakers would also be designed. The speakers would rotate back and forth, for greater coverage.
S5A: Many of the above changes were not implemented for various reasons. The siren got rid of solar power backup, because there was inadequate sunlight in the immediate area. The batteries now constantly charge from wall power. A second battery was added to improve power capacity, but the it remained a 12 volt system. Speaker rotation was not implemented. Circuitry was heavily simplified, and the charge controller was removed. The control scheme was also made easier to use. An SD card reader was added, so that the siren can now follow commands from pre-written scripts, and play audio recordings. Programming was also enhancewd so that siren tones would restart in the event of a power failure. The 2 chime tones were also modified so that they would run more quietly and sound more natural. The silent test mode was enhanced. Sound characteristics of multiple tones were changed. No full-length tests of this iteration have yet been performed.

The Death of SHART

In the summer of 2022, after several months of storage, the siren fell over and the speaker mounts broke. A few days later, it was officially decomissioned and deconstructed. However, plans are in the works for a whole new generation, one which would be feasible to sell to municipalities...

Severe Harm Alert Repeater Tower Testing

Here is a list of all of the videotaped tests for each SHART model, as well as the result. You can find all of the tests in a playlist on the Silly Goose YouTube channel. The testing procedure is not very constant, but usually the siren will run on the first Saturday of the month, between the hours of 12 and 3 PM, for 1 minute. This procedure may change, as FEMA guidelines suggest testing all outdoor warning sirens that exceed 123 dB at least once per month in a 1 minute steady signal, and a 1 minute wail signal.

SHART-1

Test occurred on June 13th of 2020. Outcome -> Success, Alternate Wail tone works.
Test occurred on June 13th of 2020. Outcome -> Success, Fast Wail tone works.
Test occurred on June 13th of 2020. Outcome -> Success, Wail tone works.
Test occurred on June 13th of 2020. Outcome -> Success, Alert tone works.
Test occurred on June 13th of 2020. Outcome -> Success, Whoop tone works.
Test occurred on June 13th of 2020. Outcome -> Success, Horn tone works.
Test occurred on June 13th of 2020. Outcome -> Success, Westminster Chimes works.
Test occurred on June 13th of 2020. Outcome -> Success, High-Low works.
Test occurred on June 13th of 2020. Outcome -> Success, Growl works.
Test occurred on June 13th of 2020. Outcome -> Success, Pulsed Alert tone works.
Test occurred on June 13th of 2020. Outcome -> Success, Pulsed Wail tone works.

SHART-2A56

Test occurred on June 15th of 2020. Outcome -> Success, Wail can be heard from a long (impressive considering the limitations) distance.
Test occurred on June 16th of 2020. Outcome -> Success, Horn tone works.
Test occurred on July 4th of 2020. Outcome -> Success, ran Westminster Chimes successfully.
Test occurred on July 4th of 2020. Outcome -> Failure, did not turn off due to control system failure.

SHART-2B56

Test occurred on July 6th of 2020. Outcome -> Success, Mechanical Hi-lo tone worked.

SHART-2C56

Test occurred on July 8th of 2020. Outcome -> Success, horn tone works, and the SHART-2C56 works during the rain.

SHART-2D56

Test occurred on July 10th of 2020. Outcome -> Success, Alert tone works.
Test occurred on July 10th of 2020. Outcome -> Success, Mechanical Hi-low (alternate steady) works.
Test occurred on July 10th of 2020. Outcome -> Success, Fast Wail tone works.
Test occurred on July 10th of 2020. Outcome -> Success, Pulsed Alert tone works.
Test occurred on July 10th of 2020. Outcome -> Success, Pulsed Wail tone works.
Test occurred on July 10th of 2020. Outcome -> Success, Alternate Wail tone works.

SHART-2E56

Test occurred on July 14th of 2020. Outcome -> Success, Whoop tone works.

SHART 3A56

Test occurred on July 28th or 29th of 2020. Outcome -> Success, growl and Horn tone work.
Test occurred on July 30th of 2020. Outcome -> Success, daily automatic test of Westminster Chimes works.
Test occurred on July 31st of 2020. Outcome -> Success, siren can be heard at over 150 feet when running Westminster Chimes.
Test occurred on August 1st of 2020. Outcome -> Success. Siren was tested for a short time in wail mode, along with the official sirens of the county. It was easily above ambient noise level at over 500 feet.

SHART 3B56

Test occurred in early August of 2020. Outcome -> Success. The PA function is easily audible and understandable at multiple hundreds of feet.
Test occurred on August 5th of 2020. Outcome -> Success. The siren had been running daily automatic tests for a few days in its unique Cheerless Chimes tone, and was recorded on August 5th at around 400 feet.
Test occurred on August 7th of 2020. Outcome -> Success. The daily automatic test of Cheerless Chimes was above ambient noise level at 700 feet, even with various obstacles blocking a direct sound path and very high humidity.

SHART 4 prototypes

Test occurred on September 10th of 2020. Outcome -> Success. The nonlinear windup and additional button controls work.
Test occurred on September 14th of 2020. Outcome -> Success. The new control panel and wail tone work.
Test occurred on September 15th of 2020. Outcome -> Success. The fast wail tone works.
Test occurred on September 19th of 2020. Outcome -> Success. The new control board and amplifiers appear to work flawlessly.

SHART 4A80

Test occurred on September 20th of 2020. Outcome -> Success. The new amplifier, control scheme, boost converter, and all other components function. The only things left to add are the solar panels.
Tests occurred on September 21st of 2020, which were not uploaded to the YouTube channel. Outcome -> Success. Everything works as it should in the permanent location.
A monthly test was attempted on October 3rd of 2020, but the siren would not turn on and was in low power mode. The problem was diagnosed as a dead battery- the battery was down to 6 volts, when it was supposed to be at over 11 volts. The battery was recharged externally. It appeared that the solar charging had not been functioning for the 3 weeks or so that the unit was running.
Monthly minute-long test occurred on October 4th of 2020, The siren ran a wail tone for 1 minute in its new location. It was also determined that charge controller was faulty, so a replacement was ordered. Multiple cameras were set up and the signal was clearly audible for over 1000 feet.

SHART 4B80

Monthly minute-long test occurred on November 7th of 2020. The siren ran the High-Low tone and a camera was positioned 1/2 mile (~2800 feet) away from the siren. The siren did not loudly broadcast to that range, test was considered a failiure.
Monthly minute-long test occurred on December 5th of 2020. The siren was initially activated in Alternate Wail mode, but it immediately shut off due to a low battery. Tests were retried with other signals but the battery was too low. The battery was low because it only charged with solar power, and the past many days had been cloudy.
Monthly minute-long test occurred on January 2nd of 2021. The siren had been sitting for a few days without a battery in it, because the battery was dead and was being heated and charged indoors. For the test, the battery was reconnected, and the siren was run in Alternate Steady mode for 1 minute, and recorded at a distance of 1,100 feet.

S5 Prototype

A monthly test was completed on March 6th of 2021. The siren used the new preprogrammed procedure system, and the new built-in SD card reader. The test included a prerecorded voice announcement, a minute of a steady tone, and a minute of the wail tone. Only 1 speaker was used in the test.
A monthly test was completed on April 3rd of 2021. The whoop and yeow tones were used in this test. For an unknown reason, the siren restarted the test partway through it. The test had to be cancelled early when the batteries were disconnected.

A monthly test was completed on September 4th of 2021. The siren ran at full power with both speakers, for 1 minute of alternate wail. There were no issues, an what appeared to be a malfunction was user error.
A monthly test was completed on October 2nd of 2021. The siren ran the "Cheerless Chimes," and 1 minute of single-tone wail. The tests were activated by Ian Hampton.

Willemstan Radio