Lynnette Reese

AI-assisted screening for COVID-19 using CT images

April 14, 2020 By Lynnette Reese Leave a Comment

Although COVID-19 tests are one means to determine infection, the rapid spread of the disease has created a shortage of test kits and a tight supply of reagents that are used to create test kits. Symptoms are another indication, but some of the infected show very mild or no symptoms. Another method leading to a swift diagnosis is to take an X-ray or CT scan of the patient’s lungs. CT scans provide a higher resolution image than X-rays. CT scans have also shown a higher sensitivity than real-time RT-PCR (reverse transcription-polymerase chain reaction). RT-PCR is known as the most sensitive technique for detecting mRNA, as a test in diagnosing COVID-19.

Part of the crushing burden of COVID-19 on medical personnel is also visited upon radiologists who determine whether X-ray or CT scans of a patient’s lungs indicate pneumonia, COVID-19, or something else. Other diagnostic tools include pulse oximeters, which indicate the amount of oxygen in the bloodstream. However, radiologists are vital in interpreting the severity of COVID-19 based on lung images. COVID-19 shows a “ground glass” appearance of opacity in the lungs, among other things.

Technology lends a…brain

The Novel Coronavirus Pneumonia Emergency Response Epidemiology Team (NCPERT), operating in China, built and deployed Artificial Intelligence (AI) as a preliminary analysis of COVID-19 in lung scans. In a pre-print, unreviewed paper submitted to medRxIV the authors state, “Working in an interdisciplinary team of over 30 people with medical and/or AI background, geographically distributed in Beijing and Wuhan, we were able to overcome a series of challenges in this particular situation and deploy the system in four weeks. Using 1,136 training cases (723 positives for COVID-19) from five hospitals, we were able to achieve a sensitivity of 0.974 and specificity of 0.922 on the test dataset, which included a variety of pulmonary diseases.” The AI system also highlighted all areas with legions, which allowed for a faster examination. The deep learning (DL) system was deployed in 16 hospitals and has performed over 1,300 screenings per day.

NCPERT took advantage of end-to-end deep neural network models, adapting the models they had previously developed for other diagnoses, and constructed a training-interference pipeline. After quickly evaluating several promising models, they compiled training data from positive cases, including a negative set from existing images of other lung diseases so the model could effectively learn to distinguish a COVID-19 lung image from others.

AI-assisted COVID-19 screening — Figure 1: The framework of the system included a) data from 5 hospitals, b) three-stage annotation and quality control, c) a deep learning model training pipeline, and d) deployment to hospitals. (Image: medRxIV.org)

Using sample images from 5 hospitals and 11 different models of CT scanners to generalize the model, NCPERT “developed a three-stage annotation and quality control pipeline, allowing inexperienced data annotators to work with senior radiologists to create accurate annotations, with minimal time investment from the radiologists.” NCPERT continuously retrained the model to improve accuracy as more data came in. The tool was delivered in an easy-to-use, low-cost platform for hospital IT staff to set up, with a remote upgrade function. Using the tool, physicians were able to rapidly locate slides and regions for a detailed examination that assisted in improving their diagnosis, regardless of their personal experience with COVID-19.

Figure 2: Some examples of the typical predictions of the segmentation model, highlighting lesion regions, as seen from the AI tool’s user interface. The model identified typical lesions found in COVID-19 pneumonia, including ground-glass opacity, intralobular septal thickening, air bronchogram sign, vessel thickening, crazy-waving pattern, fiber stripes, and “honeycomb lung” syndrome. (Image: medRxIV.org)

The model missed some positives for ground-glass opacity of less than 1 cm in diameter and did not perform well with multiple types of lesions or images where subjects introduced motion or metal objects. However, the AI tool was able to help overwhelmed hospitals by providing preliminary results that sped up the triage process of filtering suspected COVID-19 patients. Also, less experienced radiologists were able to learn to better detect COVID-19 indicative features. After receiving the first set of data, NCPERT produced the first useable model in just one week.

Much more about the NCPERT team of Chinese engineers, scientists, and physicians that built this AI model for screening CT lung scans for the presence of COVID-19 can be found at medRxIV in the pre-review pre-print paper titled “AI-assisted CT imaging analysis for COVID-19 screening: Building and deploying a medical AI system in four weeks.”

References
https://www.medrxiv.org/content/10.1101/2020.03.24.20042317v1
https://www.thermofisher.com/us/en/home/references/ambion-tech-support/rtpcr-analysis/general-articles/rt–pcr-the-basics.html
https://www.medrxiv.org/content/10.1101/2020.03.19.20039354v1

The working principle, applications and limitations of ultrasonic sensors

August 6, 2019 By Lynnette Reese 4 Comments

Figure 1: HC SR04 ultrasonic sensor. (Source: Digikey

The ultrasonic sensor (or transducer) works on the same principles as a radar system. An ultrasonic sensor can convert electrical energy into acoustic waves and vice versa. The acoustic wave signal is an ultrasonic wave traveling at a frequency above 18kHz. The famous HC SR04 ultrasonic sensor generates ultrasonic waves at 40kHz frequency.

Typically, a microcontroller is used for communication with an ultrasonic sensor. To begin measuring the distance, the microcontroller sends a trigger signal to the ultrasonic sensor. The duty cycle of this trigger signal is 10µS for the HC-SR04 ultrasonic sensor. When triggered, the ultrasonic sensor generates eight acoustic (ultrasonic) wave bursts and initiates a time counter. As soon as the reflected (echo) signal is received, the timer stops. The output of the ultrasonic sensor is a high pulse with the same duration as the time difference between transmitted ultrasonic bursts and the received echo signal.

Figure 2: Representation of trigger signal, acoustic bursts, reflected signal and output of echo pin. (Source: HC-SR04 User Guide)

The microcontroller interprets the time signal into distance using the following functions:

Theoretically, the distance can be calculated using the TRD (time/rate/distance) measurement formula. Since the calculated distance is the distance traveled from the ultrasonic transducer to the object—and back to the transducer—it is a two-way trip. By dividing this distance by 2, you can determine the actual distance from the transducer to the object. Ultrasonic waves travel at the speed of sound (343 m/s at 20°C). The distance between the object and the sensor is half of the distance traveled by the sound wave.[iv] The following equation calculates the distance to an object placed in front of an ultrasonic sensor:

Applications

Multiple areas of engineering use ultrasonic sensors. “No-contact” distance measuring is very useful in automation, robotics, and instrumentation. Below, we investigate the applications of ultrasonic sensors:

Ultrasonic Anemometers

Figure 3: A 2D Ultrasonic anemometer detects the horizontal component of wind speed and direction (Source: Biral)

Weather stations commonly used anemometers since they detect wind speed and direction efficiently. The 2D anemometers can measure only the horizontal component of wind speed and direction, whereas 3D anemometers can measure the vertical component of wind, as well.

Apart from measuring wind speed and direction, ultrasonic anemometers can also measure temperature because the speed of ultrasonic sound waves is affected by variations in temperature while maintaining independence from changes in pressure. Temperature is calculated by measuring speed variations in ultrasonic sound.

The ultrasonic anemometer is more durable as compared to the cup anemometer and vane anemometer since it has no moving parts and it operates using ultrasonic sound waves. [vi]

Figure 4: A 3D Ultrasonic anemometer measures both horizontal and vertical components of the wind speed and direction. (Source: Biral)

Tide gauge

A tide gauge is used to monitor sea level. It also detects tides, storm surges, tsunamis, swells, and other coastal processes. [vii] A tide gauge can use an ultrasonic sensor to detect real-time water level. The gauge is often linked to an online database where a record is maintained, and in case of a risky situation, the system can trigger an alarm.

Tank level

Measuring fluid level in a tank is similar to a tide gauge. However, in this case, the fluid can be clear water, a corrosive chemical, or a flammable fluid. Unlike optical sensors and float switches, ultrasonic sensors are less likely to corrode as they do not make contact with the fluid.

Functional in sunlight

The sunlight at Earth’s surface is composed of around 52-55% of infrared light.[ix] If an infrared sensor detects an object using infrared light, the process is disturbed due to the interference of infrared light present in sunlight. However, ultrasonic sensors are not affected by the infrared spectrum present in sunlight.

Web-guiding systems

Web-guiding systems position flat materials (e.g., newspaper, plastic film) and widely use ultrasonic sensors. According to Maxcess, “In 1939, Irwin Fife invented the first web guide in his Garage in Oklahoma City, Oklahoma, solving a newspaper owner’s challenge of keeping paper aligned in his high-speed newspaper press.“ [x] A web-guiding system uses a non-contact sensor for detecting and tracking objects at multiple stages. The purpose is to ensure that the material is positioned correctly. If the material is moving out of alignment, the system mechanically places it back on the machine’s processing path. Ultrasonic sensors are suitable for web-guiding systems as the process requires non-contact, high-speed, and efficient functionality.

UAV navigation

Figure 5: Ultrasonic sensor measuring height during drone’s flight. (Source: RadioLink)

Unmanned aerial vehicles (UAVs)—or drones—commonly use ultrasonic sensors for monitoring any objects in the UAV’s path and distance from the ground.

The autonomous feature of detecting safe distances enables the aircraft to avoid crashing. And as the flight of path changes instantaneously, the ultrasonic detection of distances can prenvent a drone from crashing.

Figure 6: Ultrasonic sensor measuring distance from object during drone’s flight. (Source: RadioLink)

Limitations of ultrasonic sensors

Ultrasonic sensors such as the HC-SR04 can efficiently measure distances up to 400 cm with a slight tolerance of 3 mm. [xiii] However, if a target object is positioned such that the ultrasonic signal is deflected away rather than reflected back to the ultrasonic sensor, the calculated distance can be incorrect. In some cases, the target object is so small that the reflected ultrasonic signal is insufficient for detection, and the distance cannot be measured correctly.

Furthermore, objects like fabric and carpet can absorb acoustic signals. If the signal is absorbed in the target object’s end, it cannot reflect back to the sensor, and hence, the distance cannot be measured.

Figure 7: Representation of ultrasonic signal deflected due to target object’s position, resulting in error. (Source: Macduino)

The intense sensitivity of ultrasonic sensors makes them efficient, but that sensitivity can also cause problems. Ultrasonic sensors can detect false signals coming from the airwaves disturbed by an air conditioning system and a pulse coming from a ceiling fan, for instance.

Ultrasonic sensors can detect objects placed within their range, but they cannot distinguish between different shapes and sizes. However, one can overcome this limitation can by using two sensors instead of just one sensor. One can install both sensors a distance away from each other, or they can be adjacent. By observing the overlapped shaded region, one can get a better idea of the shape and size of the target object.

Figure 8: Representation of overlapping region when one places two ultrasonic sensors at a distance or adjacent to each other. (Source: msu.edu)

https://www.mouser.com/ds/2/813/HCSR04-1022824.pdf[ii] https://www.mpja.com/download/hc-sr04_ultrasonic_module_user_guidejohn.pdf
https://www.mpja.com/download/hc-sr04_ultrasonic_module_user_guidejohn.pdf
https://www.teachengineering.org/activities/view/nyu_soundwaves_activity1
https://www.arrow.com/en/research-and-events/articles/ultrasonic-sensors-how-they-work-and-how-to-use-them-with-arduino
https://en.wikipedia.org/wiki/Anemometer

ToughSonic® Ultrasonic Sensors Help Provide Tsunami Warning

Principles of Thermal Ecology: Temperature, Energy and Life; Clarke, Andrew. 2017
http://www.maxcessintl.com/fife
https://en.wikipedia.org/wiki/Web-guiding_systems
https://www.maxbotix.com/uav-ultrasonic-sensors.htm
https://www.mouser.com/ds/2/813/HCSR04-1022824.pdf
http://cmra.rec.ri.cmu.edu/content/electronics/boe/ultrasonic_sensor/1.html
https://www.egr.msu.edu/classes/ece480/capstone/fall09/group05/docs/ece480_dt5_application_note_nkelly.pdf
https://www.egr.msu.edu/classes/ece480/capstone/fall09/group05/docs/ece480_dt5_application_note_nkelly.pdf

Embedded security: Do you know what you don’t know?

August 5, 2019 By Lynnette Reese Leave a Comment

by Lynnette Reese, contributing editor

Most of us assume that security comes from the operating system or application software. However, firmware is also at risk, including anything that barely hovers over the physical layer (silicon) in the chip. Embedded security is especially important in blocking seemingly innocuous pathways for hacking, such as in an automotive infotainment center.

Most of us assume that security comes from the operating system or application software. However, the firmware is also at risk, including anything that barely hovers over the physical layer (silicon) in the chip. Booting up a chip means that some hardcoding in the silicon itself is permanent (unless a portion of the chip is FPGA-based). Anything above that permanent coding, which acts like a tiny boot loader, is vulnerable to hacking.

After a chip is powered up, it looks for firmware in memory, usually ROM. That firmware provides boot management services. However, for decades firmware so low on the stack hasn’t been armored with security measures like requiring it to be cryptographically authenticated by the tiny silicon-based boot loader. Firmware further up the stack for system management can be vulnerable without any checks for authenticity, too. If the firmware is designed without any consideration for security, then hackers can subvert the fundamental, root-functioning operations to do their bidding without detection by changing the firmware. Any subsequent security checks from anything higher in the stack, such as the operating system or application programs, could be orchestrated to do what the hacker wants. Therefore, it’s imperative that security starts at the physical layer as a function provided by the chip manufacturer.

Embedded security — Figure 1: The stack on the left is far less secure than the one on the right. The secure hardware starts at the physical layer using fuses in the silicon to create identification. The Hardware Root of Trust (HWRoT) is a secure area, including memory, on the physical layer that cannot be accessed by unauthorized sources. (Source: Sequitur Labs)

The depth of the process for securing everything from silicon to product design, development, volume product manufacturing, and throughout the product’s lifecycle (including firmware and software updates) is a tremendous learning curve. Companies like NXP offer a means to secure embedded products starting in the silicon. However, seamlessly incorporating and coordinating the embedded security throughout the supply chain is a daunting project with a steep learning curve.

Figure 2: The main steps in securing from ROM to the operating system. Setting fuses on and locking the system on chip, creating a unique device ID, cryptographic key pair and certificate; creating domain isolation, instantiating trusted execution, installing secure functions, encrypting each device uniquely (in volume production), loading and storing keys and certificates for each embedded product, closing the programming tool (e.g., JTAG), and handing off supervisory control to the operating system (e.g. Linux). (Source: Sequitur Labs)

The importance of a secure supply chain

Even securing an automotive infotainment system is vital, since it’s possible to hack any of a hundred or more control units positioned in different systems throughout modern vehicles today. NXP provides the security at the silicon layer in NXP’s i.MX 8 and i.MX 8M application processors. NXP’s i.MX 8M high-performance chips feature up to four Arm^® Cortex®-M4 and Cortex^®-A53 cores designed to meet the requirements of an enhanced multimedia user experience for smart applications. According to an NXP fact sheet, “Built with advanced media processing, secure domain partitioning and innovative vision processing, the i.MX 8 applications processor family will revolutionize multiple display automotive applications, industrial systems, vision, HMI, and single-board computers.”

It’s critical that integrated chips, passives, power supplies, connectors, and all electronic devices be authentic. Counterfeit products can cause anything from catastrophic failure to a shorter product life that results in poor product reviews. Counterfeit products can be extremely difficult to spot. Besides purchasing individual components directly from the manufacturer, purchasing components from an authorized distributor with anti-counterfeit parts risk mitigation measures in place, such as Mouser Electronics, is imperative.

Starting with secure silicon is important, but how do you ensure secure boot managers, firmware, application delivery during manufacturing, and subsequent firmware updates for vehicles already in service? One answer is to do it yourself, but how do you know what you don’t know? There are some companies that have been working for years on embedded security with a wholistic, viewpoint. For instance, Sequitur Labs has a vetted cradle-to-grave security solution for NXP’s i.MX 8 and 8M families of application processors. Sequitur Labs’ EmSPARK Security Suite™ can ensure an easy security process because they’ve abstracted the tedious process from end-to-end. Using NXP’s Trust Architecture, development tools can provide an extended secure boot process bound to a hardware root of trust.

Establishing a hardware isolated secure domain (enclave) can also extend through protecting firmware in volume manufacturing, implementing the necessary “plumbing” to ensure that any updates to processors can only be accomplished by authorized entities. If the production of an embedded device is farmed out to external assemblers or manufacturers, the external company can only load encrypted software, which locks them out from accessing any intellectual property. Not all embedded security tools go this far. For example, with tools like the EmSPARK Security Suite working with NXP’s i.MX8, only software with the proper security key that you create can get unlocked by the chip just after powering up. Any updates required during the lifetime of the product – for example, a car infotainment center – would be enabled; but only if the car’s NXP infotainment processor authenticates the update as genuine by using the keys that were originally generated by the Sequitur Labs’ tools suite. Even the update process is flexible, however, and one can delegate authority. For example, if an OEM creates a product for a customer, but the customer wants to manage the devices themselves, the customer can own the secure update process.

Figure 3: Full value can be extracted from the NXP trust architecture by the nearly turnkey EmSPARK Security Suite of tools. (Source: Sequitur Labs)

It’s unfortunate that the digital gifts that technology has bestowed upon mankind are vulnerable to the criminal intentions of others. Technology has grown tremendously by building on the shoulders of giants. Engineers, scientists, developers, hobbyists, and others who use technology to create marvelous things that enhance human productivity and create joy in everyday living must now consider securing intellectual property, life, and limb. It’s good to know that NXP and their ecosystem partners like Sequitur Labs work hard in safekeeping the creative process for others.

What could happen if there’s an EMP or solar superstorm? Part 2

May 7, 2019 By Lynnette Reese 1 Comment

Read Part 1 here.

In 2006, the creators of the TV series Jericho imagined how nuclear detonation would affect a small Kansas town. Large solar-based EMPs have been documented. In 1859, a solar storm dubbed the Carrington Event hit the Earth’s magnetosphere. One of the largest solar storms on record, the CME in 1859 caused telegraph equipment to spark and an aurora borealis bright enough for people to read a newspaper outside at night. According to Robert M. Schoch, in Forgotten Civilization: The Role of Solar Outbursts in Our Past and Future, “the all-important transformers along gridlines could heat up and melt, causing massive blackouts. This famously happened in Canada’s Québec province, when on March 13, 1989, a solar storm disrupted Hydro – Québec transmission lines and in less than two minutes a blackout occurred, affecting 6 million people for 9 to 12 hours (Berge 2011). In this particular case, a billion-ton cloud of gas and plasma was released from the sun, with the energy of thousands of nuclear bombs exploding simultaneously, and headed toward Earth at a speed of approximately 1.6 million kilometers an hour (Odenwald 2009). Besides the Québec blackout, due to the solar flare accompanying the March 1989 solar storm (which was nowhere near the size of the Carrington Event), shortwave radio signals were disrupted…. And there were still more direct consequences, as reported by Sten Odenwald: ‘in space, some satellites actually tumbled out of control for several hours. NASA’s TDRS-1 communications satellite recorded over 250 anomalies as high-energy particles invaded the satellites sensitive electronics.’”

A similar note was sounded by William G. Duff, PhD., whose 53-year career centered on surmounting electromagnetic compatibility issues for computer and communications systems. During an interview for the IEEE History Center in 2012, Duff, a past president of the EMC Society, explained, “If you have an explosion, [a] high altitude nuclear explosion over the center of the US, it will wipe out all the electronic equipment that’s not protected over the whole US. There’s a lot of interest in that problem now. We’re so dependent on grids. We’ve got communication grids. We’ve got power grids. And all of them are potentially susceptible to EMI problems.”[i]

solar superstorms — Figure 1: The U.S. power grid has a network of regions that share power, buying and selling it to each other to meet fluctuating demand. (Image: solarstorms.org)

In December 2013 the Defense Threat Reduction Agency (DTRA), asked solutions providers to come forward if they were interested in the opportunity to further the knowledge of “high altitude weapons electromagnetic pulse effects modeling.”[ii] However, in noting the DTRA’s action, Robert Heckhusen, in an article published first on medium.com and then in Foreign Policy, calls “the risk that a rogue state could emit a blast of electromagnetic energy” an “overblown theory,”[iii] at the same time acknowledging that EMP from a powerful solar storm could mean “lights out.” Despite the skepticism by some, DTRA has continued to address the issue. For example, by 2016, the agency had sent out a Request for Proposal (RFP) for participation in a “commercial R&D program to create EMP protected microgrids for critical infrastructure needed both on and off military bases and other defense critical infrastructure.”[iv]

By June of 2017 the Commission to Assess the Threat to the United States from Electromagnetic Pulse (EMP) Attack had completed gathering information and in April of 2018 made the report available to the public (PDF, 2MB). The commission warned, “Within the last decade, newly-armed adversaries, including North Korea, have been developing the ability and threatening to carry out an EMP attack against the United States.”[v] The same report also noted that “Solar superstorms can generate natural EMP over remarkably wide areas.”

Options

You can shield electronics from EM radiation with a Faraday bag, which is a Faraday cage made of flexible fabric made of woven metallic thread. The military has ruggedized laptops for the battlefield. You can cage equipment to protect it (see Figure 2). The EMP Commission report makes “preppers” look wise, with the grim observation that ”a long-term outage owing to EMP could disable most critical supply chains, leaving the US population living in conditions similar to centuries past, before the advent of electric power. In the 1800s, the US population was less than 60 million, and those people had many skills and assets necessary for survival without today’s infrastructure. An extended blackout today could result in the death of a large fraction of the American people through the effects of societal collapse, disease, and starvation. While national planning and preparation for such events could help mitigate the damage, few such actions are currently underway or even being contemplated.”[vi]

It’s not clear if such a powerful electromagnetic pulse will take place naturally or otherwise. A mild hit would destroy some, but not all exposed electronics and electrical systems. The cost to harden all power grids against such an event is so high that none are willing to do more than keep a few replacement parts on hand. However, looking at the research commissioned by the government, it’s clear that a large enough hit (natural or otherwise) could wipe out many power grids at once, cripple cars, and destroy exposed (and operating) electronics. Some present-day equipment is life-sustaining. For example, without refrigeration, insulin for diabetics would deteriorate after 28 days, as stored insulin must be refrigerated.

Figure 1: A Faraday cage shielding equipment at a power plant in Heimbach, Germany. (Image: By Frank Vincentz – Own work, CC BY-SA 3.0)

Being without power for a long time exposes the potential frailty of a modern, technically intricate society. A few years back, a friend of mine had no power for an entire week due to a severe snow storm. He couldn’t drive because the car was under several feet of snow. The contents of the refrigerator and freezer were put outside, showers were ice cold, and they heated the house with the fireplace while scavenging every blanket they owned. Cell phones were impossible to charge, and the usual modes of entertainment were dead. He admitted that the situation was annoying at first, then became frustrating. The grocery store within walking distance would only take cash, as all the credit card machines were down and they had none of those little carbon copy slips left over from the ‘80s. Without power, ATMs were inoperable and traffic signals dead. His family survived the technologically primitive week and truly appreciated the recovery of their accustomed standard of living. However, compared to a bad storm that left a city without power for a week, an EMP attack could destroy equipment that takes weeks or months to repair.

Living with electricity has become a fundamental part of life as agriculture, banking, and transportation are entwined with layers of embedded technology, including the internet. The internet is near the top of the technology stack, above computers and electricity. If a large region were bespeckled with electrical damage from a solar storm, neighboring regions with less damage would also be affected (because we are interconnected) and respond. Extensive electrical damage would mean longer term hardship than a storm for many, however, and change the standard of living for a multitude, not just the ones who are directly hit. For centuries, humanity has demonstrated that no man is an island. Modern technology has interconnected us all, and continues to connect us to an increasingly greater extent as technology continues to build upon itself.

References

[i] https://ethw.org/Oral-History:William_G._Duff
[ii] https://www.fbo.gov/index?s=opportunity&mode=form&id=b91c8673d13b567f3f5da4c41628db4f&tab=core&_cview=0
[iii] https://medium.com/war-is-boring/bzzt-military-wants-to-protect-satellites-from-emp-weapons-fde1f86d14bc ; https://foreignpolicy.com/2014/01/03/military-looks-to-shield-its-satellites-from-electromagnetic-attacks/
[iv] http://www.stop-emp.com/Final-DTRA-IAN-Press-Release%20-w-logos-centered.pdf
[v] www.firstempcommission.org/uploads/1/1/9/5/119571849/executive_report_on_assessing_the_threat_from_emp_-_final_april2018.pdf
[vi] Ibid.

What happens to an MCU if there’s an EMP or Solar Superstorm? Part 1

May 3, 2019 By Lynnette Reese Leave a Comment

An Electromagnetic Pulse (EMP) is an impulse or blast wave that’s generated by geomagnetic activity due to solar flares, solar storms, solar superstorms, or the detonation of nuclear weapons. The horrible truth is that an EMP can wipe out electronics by inducing inductive coupling on traces, wires, or any conductive media within hundreds to thousands of miles from the blast, depending on the strength and altitude of the EMP source.

Power lines subjected to EMP can induce considerable current into every connected device. Modern cars traveling within range would stop working as expected. Drive-by-wire (electronic throttle) is now common, having replaced the old-fashioned mechanically linked steering. Drive-by-wire has no physical linkage, weighs less, reduces the number of moving parts, and eliminates the same need for maintenance like the old mechanical system. A 2002 report from the EMP Commission assessed the nature and magnitude of an EMP attack made on the United States, back when the automobiles with the most electronics were typically luxury vehicles. According to Pico Technology, “Today’s systems have a specific ECM that incorporates the necessary programming to facilitate the input signals from the throttle pedal potentiometers and signal outputs to the electronic throttle body.”[i]

Solar flares are also capable of wreaking havoc on modern life. Coronal mass ejection, or CME, is a solar superstorm. One such storm was the 1859 Carrington Event. Among much else, that solar superstorm sent sparks flying from telegraph equipment. Power lines are susceptible to low intensity, long duration EMP. EMP of the long duration type induces “quasi-direct current”[ii] in the power transmission lines. There’s “a shift in the magnetic field operating point in the core of the transformers.”[iii] Rather than its usual helpful role, exposed transformers would generate abnormal harmonic waveforms, overwhelming itself and the grid. The upcoming Lagrange solar warning mission “will allow monitoring of the potentially hazardous sunspots and high-speed solar wind streams before they come into view from Earth, and detect solar events and their propagation toward the Earth with higher accuracy than is possible today.”[iv] Solar flares emit radiation, including X-rays and ultraviolet radiation, which travels at the speed of light. In March of 1989, large flares disrupted long-range communications, including satellite links and shortwave transmissions, as reported from Holloman Air Force Base in New Mexico, arriving in just 8 minutes after the solar eruption.[v]

Figure 2: The European Space Agency (ESA) has begun assessing two future space weather missions to monitor and anticipate potentially hazardous solar events. (Image: www.esa.int)

Understanding advances

A key step in developing solutions to protect electronics and electrical systems has been figuring out how to simulate an EMP event, which requires storing and discharging massive amounts of electrical energy in brief pulses. In the 1970s, Physics International began research into simulating the type of EMP that’s caused by detonating a nuclear weapon in space. In a project inspired by the initial efforts of Physics International, a 2015 R&D 100 award went to the Nuclear Technologies Department of the U.S. Defense Threat Reduction Agency (DTRA). The R&D result was an ion band implanted switch (IBIS) photoconductive semiconductor switch, or PCSS, which can be used for electromagnetic pulse simulation. Along with DTRA, on the team which developed the switch, was the science and research firm UES Inc. and L3 Applied Technologies, Inc. (Physics International was a forerunner of L3 Applied Technologies.)

Satellites are not protected by Earth’s atmosphere and can experience high-energy particle radiation that can cause them to malfunction. For instance, Geostationary satellites use the Earth’s magnetic field to determine orientation. If the Earth’s magnetic field is disturbed by a solar storm, it can cause satellites to flip and tumble if they aren’t manually overridden. Heavy shielding and radiation-hardened devices are common options to avoid the need for repairs in any device that’s exposed to EMP radiation. EMI/EMP filters can provide some circuit protection. Smiths Interconnect has products able to protect “against high energy transients from direct lightning strikes, standard EMP, high-altitude EMP (HEMP), non-nuclear EMP (NNEMP) and electromagnetic interference (EMI).”

What if…?

EMPs have captured the imagination of fiction authors considering “future history” and the attention of the scientific community, think tanks, and government agencies. One example includes a 2009 issue of Military &Aerospace Electronics magazine, where a senior program manager for a maker of electronic surge protection products said, “Explode a nuclear weapon 200 miles above Kansas City, and you can kiss off the power grids in all the United States. We’re back to being hunter/gatherers.”[vi] William Forstchen, in his 2009 novel One Second After, fleshed out the social norms that fall apart following a HEMP launched over the U.S. by a hostile power. Newt Gingrich, former U.S. Speaker of the House, in a foreword to the book states, “Few in our government and in the public sector have openly confronted the threat offered by the use of but one nuclear weapon, in the hands of a determined enemy, who calibrates it to trigger a massive EMP burst. Such an event would destroy our complex, delicate high-tech society in an instant and throw all of our lives back to an existence equal to that of the Middle Ages. Millions would die in the first week alone, perhaps even you who are reading this if you require certain medications, let alone the most basic needs of our lives such as food and clean water.”[vii]

Unprotected power grids are vulnerable. The effect of a solar storm or EMP can ruin a transformer with a large dose of current. According to SolarStorms.org, “When space weather disturbances cause ‘Geomagnetically Induced Currents,’ these GICs can enter a transformer through its Earth ground connection. The added DC current to the transformer causes the relationship between the AC voltage and current to change at the source of the electricity… Because of the way that GIC currents affect the transformer, it only takes a hundred amperes of GIC current or less to cause a transformer to overload during one-half of its 60-cycle operation. As the transformer switches 60 times a second between being saturated and unsaturated, the normal hum of a transformer becomes a raucous, crackling whine. Regions of opposed magnetism as big as your fist in the core steel plates crash about and vibrate the 100-ton transformer nearly as big as a house in a process that physicists call magnetostriction.”[viii]

Mistakes to avoid when submitting a design for PCB assembly

April 8, 2019 By Lynnette Reese 1 Comment

Simple prototype Printed Circuit Boards (PCBs) require, on average, 11.6 physical prototypes, whereas more complex PCBs require as many as 16.[i] And according to Advanced Assembly, a prototype PCB assembly house, “94% of assembly data is accidentally submitted with errors.”[ii] PCB assembly houses will extract design data from PCB design files and create a PDF image.

A good assembly house has engineers who review and verify your design for component fit and rotation, identifying mistakes for customers. Catching another’s mistakes is decent, but can cause delays, especially if the PCB assembler can’t quickly find a correction for a simple error by looking at other project files.

Some common mistakes that can cause delays or create unusable PCBs include:

Missing or corrupt design files such as Gerber files (including the stencil or paste mask layer), any XYRS data files, and the Bill of Materials (BOM) cause delays. All the files are used in fabricating a PCB. Gerber files (.GBR) describe the layers of the PCB, which can include layers made of copper (e.g., traces and ground planes), solder masks, data for drilling holes on the PCB, silkscreens, the board outline (which may be included in another file), information on inner layers, and so forth. XYRS files have a .XYRS extension and are used for placing parts in automated pick-and-place systems. XYRS files indicate where the part is placed on an X-Y axis, component rotation, which side of the PCB the component is located, the type of component (surface mount or through-hole), part footprint, and a few other details. A BOM is a list that includes a short reference name, a part number, the manufacturer, and the quantity of every component on the PCB, but does not typically note location. Review the BOM carefully. Many mistakes are made in a BOM simply because you’ve looked at it so many times that your brain automatically assumes that it’s correct just as it was three design revisions ago! Carefully number your reference designators so that it’s unambiguous as to which component goes where. Make certain that all design files are provided, not corrupted, and in the correct format for the PCB assembler.
Incorrect markings on a PCB can cause components to be stuffed Make sure that the polarity markings on the PCB are correct and present. E.g., if you have a diode, make sure the anode and cathode are marked correctly. A polarized capacitor should be indicated with a plus sign in the correct place on the PCB. For an Integrated Chip (IC), identify pin one by marking it in the correct location on the PCB. Circuit simulators can add virtual parts, so make sure your final PCB files include or exclude them as needed.
Sub-optimal sizes for land patterns can be a subtle problem. A land pattern has a slightly different meaning than a footprint, although they are often used However, for PCB assemblers, there’s a huge difference between them that affects manufacturing quality. A footprint is the imprint of a component (that is sometimes loosely termed “real estate”) for the PCB surface area that’s consumed. Think of the footprint as you would if you pressed the component into soft dirt. A land pattern refers to both the outline and the size of the pads required for a component on a PCB. The land pattern will always be larger than the footprint since the land pattern designates the PCB pads where component leads will be landed and soldered. Land patterns are normally identified in component datasheets and should match up accurately with the component footprint. (See Figure 1). Proper land patterns will improve manufacturability.

PCB assembly — Figure 1: The land pattern for the Texas Instruments MSP430FR247x family of mixed-signal microcontrollers. (Source: TI Datasheet SLASE07.PDF)

Missing notes on height and length requirements on drawings can make it so your board will not fit into the enclosure that may already be fixed. Don’t forget the total height and thickness of your PCB may affect your ability to fit it into a fixed space after components have been stuffed. If your project is size-restricted, then start the design by placing taller components where height clearance is less of a concern. Another tip – note that standard sized PCBs can make a difference in cost versus custom sizes.
Missing additional information can lead to avoidable issues. The PCB assemblers cannot read your mind. Use a readme.txt file bundled with all the other files to add any special instructions or otherwise missing information. Other files can assist PCB assemblers in interpreting a mistake-free PCB for you, therefore, provide notes, schematics, 3D drawings, and identify any specific features (even how you plan to use the PCB) that will help them form a complete picture.

Meticulously reviewing your project, the PCB layout, and all files and information can make a world of difference in debugging your prototype later. Simple mistakes can burn hours or days in troubleshooting. For instance, choosing the wrong component or putting a component in the incorrect place will complicate the whole process. At best, such mistakes will create delays as the best PCB assembler finds a component marked for a spot that does not match the location’s land pattern. The worst case will be a finished PCB that doesn’t work.

Bundle all your files together by creating a .ZIP or .RAR file. Sending files as multiple separate attachments in different emails can cause delays. Remember that the PCB maker will likely have dozens, hundreds, or even thousands of customers that have submitted their files in the same day or week, with multiple employees handling the incoming requests. A request that’s fractured across two or more emails or uploads may not reach the same person handling your first submission. Bay Area Circuits has a free InstantDFM™ tool that can identify potential production issues in your files before you actually submit them, whether or not you actually use their services.

PCB assembly is placed on hold if it’s missing information, files, special or rare parts for population, or if it includes incomplete information or corrupted files. It’s a universal rule of thumb that any project (across all engineering disciplines) will cost more the longer it takes. Delays almost always add to the total cost, and prior planning prevents poor performance.

[i] https://www.aapcb.com/blog/3-things-affect-your-pcb-assembly-quote/

[ii] https://aapcb.com/knowledge-center/5-common-design-mistakes/

Understanding wireless across the spectrum

April 2, 2019 By Lynnette Reese Leave a Comment

Waves are part of our world as both waves in the ocean and radio waves coming in from the Sun’s corona share fundamental properties.[i] Wireless communication was born from the science of waves and how they travel. After more than a hundred years of study, we’ve organized waves into a large electromagnetic spectrum (EM) that is classified by frequency.

Figure 1: Hertz’s idea for creating and verifying the presence of radio waves in an experiment that included a spark gap transmitter (at left). The spark gap was two conducting electrodes separated by a space and powered by a Ruhmkorff coil (T). The receiver loop (at right) was a simple loop antenna with a spark gap. (Source: Wikipedia)

Wavelength and frequency are related:

λ f =c

where λ is wavelength, f is frequency, and c is the speed of light.^[ii] Frequency tells us how many wave peaks pass through a fixed position across an interval of time. Many frequencies are channels used to carry signals, creating an entire spectrum that varies from signals with very long wavelengths to the very short.

How did the spectrum come to be?

What Hertz discovered

Wireless technology began over a hundred years ago. Heinrich Rudolf Hertz (1857-1894) was a brilliant German experimental physicist who proved that electromagnetic waves exist, as predicted by James Clerk Maxwell’s (1831-1879) electromagnetic field theory. In 1888, Hertz ferreted out what would someday be called “radio waves.” Hertz employed an apparatus (Figure 1) using a zinc sheet to demonstrate “evidence of a standing wave that can only have been formed by a traveling wave combining with its reflection from the zinc screen.”[iii]

Over his short life, Hertz carried out a series of extraordinary experiments demonstrating noteworthy properties of the waves that he detected. Like light waves, Hertz’s waves could be polarized, reflected, and refracted. Hertz identified waves of a wavelength that falls within the radio spectrum, which is a part of the much larger Electromagnetic (EM) Spectrum (organized as pictured in Figure 2).

We define wavelength as the “distance between adjacent points of the electromagnetic wave that are in equal phase (e.g., wave crests).”[iv] The wavelength of visible light is between 3.8×10^{– 7} and 7.6×10^-7 (see Table 1). However, Hertz had discovered wavelengths much larger than visible light waves, on the order of > 0.3 meters. Hertz passed away in his late thirties with a condition that’s treatable with steroids today. The American Association for the Advancement of Science (AAAS) christened the unit as “one cycle per second,” or a “Hertz” in Hertz’s honor.[v]

Table 1: A major portion of the EM Spectrum, arranged in order of frequency and wavelength. (Source: Science.org)

There are nine bands that the National Telecommunications and Information Administration has organized within the radio spectrum (Figure 2). Waves of different frequencies exhibit different capabilities with varying advantages. For example, waves propagating at frequencies higher than 10 GHz have to worry more about attenuation; that is, being weakened by rain, fog, sleet, snow, and gaseous absorption (caused by oxygen and water vapor.)[vi] The majority of cellular systems keep to a “safe zone” in the spectrum by using frequencies below 6 GHz. Frequencies below 6 GHz translate to good coverage despite atmospheric variations. Although adjusting for weather conditions is usually “not part of initial system design,”[vii] carriers such as Verizon note, “Severe weather and seasonal conditions like heavy snow, storms or even trees in bloom can also affect [coverage].”[viii]

wireless spectrum — Figure 2: The orange section is the radio spectrum where most wireless communications are implemented. The radio spectrum is part of the electromagnetic spectrum that covers all known wavelengths. (Source: Transportation.gov)

The atmospheric effects at frequencies greater than 6 GHz factor into wireless system design. Design at greater than 6 GHz must take into account how backhaul reacts. (Backhaul is defined as the “intermediate links between the core network, or backbone network, and the small subnetworks at the edge of the network.”[i]) A key enabler of 5G network backhaul links can be tens of miles long.

When gaseous absorption (noted earlier) influences backhaul and other wireless system components, it’s the water vapor or oxygen molecules, which have specific frequencies of resonance, that cause peaks of absorption around affected frequencies. The absorption peak for water vapor is around 24 GHz (and multiples thereof), with 60 GHz being the main peak for oxygen.[ii]

One situation causing ample concern is heavy rain comprised of large-sized raindrops. Should rainfall be in the “cats’ and dogs’” realm of two- to four-inches an hour, a 10-20dB/km attenuation could take place, causing radio outages for both Local Multipoint Distribution Systems (i.e., LMDS at 28 GHz) or point-to-point microwave links (at 6, 11, 18, and 23 GHz).[iii]

Rules and differences

The EM spectrum, although intangible and invisible, is a resource. Rules and regulations governing the spectrum set conditions for coordination and guide each of the nine bands,[iv] primarily to avoid interference for competing services that are using the same spectrum in the same location.

While the need for rules applies to all the bands, the properties and applications for each band are not homogeneous. Waves in the lower part of the spectrum are used for TV and radio broadcasting, getting their start with scientists like Guglielmo Marconi (1874-1937) making “the leap from Hertz’s lab experiments to practical wireless telegraphy using electromagnetic waves as the medium of communication.”[v] Waves propagating at lower frequencies are also useful for mobile applications. Applications requiring higher throughput can capitalize on the generous amount of higher spectrum that’, but waves with a higher frequency are more influenced by atmospheric effects and need “near line-of-sight conditions.”[vi] Avoiding the need for line-of-sight or nearly line-of-sight made a success of Marconi’s invention of mobile communications using ships traversing the English Channel. Marconi sent the first successful radio transmission thousands of miles across the Atlantic Ocean, as radio signals bounced off the ionosphere to reach the next ship in spite of the curvature of the earth.

In recent history, the spectrum, although intangible and invisible, has been sold with intense competition as a resource. Excessive bidding on carrier frequencies in the spectrum, combined with costly satellite services and a few premature broadband wireless objectives triggered both failures and bankruptcies around 2000. Nevertheless, wireless communication has grown continuously as technology has persistently improved. Continued growth will come as a result of a hearty appetite for data use and the proliferation of new applications, including the Internet of Things (IoT).

Hints of our wireless future

Innovative wireless solutions cover the spectrum, aided by microcontrollers and other devices. Many new applications could someday result from activity encouraged by the Federal Communications Commission’s (FCC) recent proposal to “make a massive 21.2 gigahertz of spectrum above 95 GHz available for unlicensed use across four frequency bands.”[vii] In February 2019, Ajit Pai, FCC Chairman, states that he is “…also proposing to add a new experimental license type that would that would permit experimental use on any frequency from 95 GHz to 3 THz, with no limits on geography or technology.” Pai goes on to say, “We currently don’t know precisely what types of applications and wireless services the laws of physics will permit in these bands…..The history of wireless innovation is one of government creating space for broad thinking and entrepreneurs using that space to take us in unexpected directions. With this order, we’ll set up a big sandbox for engineers and technologists to work with — and we’ll then see what American ingenuity delivers.”

[i] https://en.wikipedia.org/wiki/Backhaul_(telecommunications)

[ii] http://morse.colorado.edu/~tlen5510/text/classwebch1.html#x5-40001

[iii] Ibid.

[iv] Ibid.

[v] Raboy, Marc. Marconi: The Man Who Networked the World. New York: Oxford University Press, 2016, 54.

[vi] http://morse.colorado.edu/~tlen5510/text/classwebch1.html#x5-40001

[vii] https://www.fcc.gov/news-events/blog/2019/02/21/behold-ides-march

Considerations for battery-powered designs

March 26, 2019 By Lynnette Reese 2 Comments

How can you make your resume really stand out? This has been the topic of books, articles, class assignments, and blogs for decades. It came up in conversation one day, and a friend of mine had a unique story. I’ve included it below, with some edits.

“With so much written on the topic, there is one resume that still sticks with me thirty-five years later. In the mid-1980s I was working as an electronics technician for a component manufacturer. The department had an opening to expand and hire a second technician. Before the internet, jobs were typically placed in a newspaper “want ads” with candidates mailing in the resumes and cover letters via U.S. mail. Among all the stacks of letters, my supervisor received one package about the size of a shoe box. Inside the box was a device that had an array of alphanumeric LED displays on its cover and written instructions on how to plug the device in and turn it on. After following the instructions, the red LED display lit up and started to scroll the applicant’s cover letter and resume in a continuous loop, which I still remember.

The applicant worked for Franklin Computer at the time. Franklin Computer was known for pocket dictionaries, spell checkers, and even an electronic Bible. In the mid-1980s they attempted to market Apple IIe clones.¹ Unlike the easy availability of similar devices today, through his employer the applicant had access to resources so he could program and burn his resume to non-volatile memory chips of the day and program a Microcontroller (MCU) to display it. Along with catching our attention, it also demonstrated his skills and ability to contrast against others with a custom device. He got the job.

Recently, a client wanted an Arduino design that would remain dormant for up to 20 days in a dark box that would be delivered by mail. When the recipient opened the box, a custom message would scroll across a 1.3” OLED display. I accepted the project. It was a young web designer that wanted his resume to stand out among his competition. He planned to purchase a low-cost Arduino and OLED displays, assemble them into a custom package, and mail them to hiring managers at companies where he was going to apply. The client knew that what he was asking for was technically possible but didn’t have the skills to execute it. I was happy to take the assignment and create my oven version of the electronic resume display I had seen 35 years ago.

Having an Arduino scroll a message across an OLED display is relatively easy with available opensource libraries. Some custom programming was necessary to get the text to display word by word, as requested. Wiring between the Arduino and chosen OLED display was simplified by using a two-wire I²C interface. Detecting when the box had been opened was just a matter of using a photoresistor (a Light-Dependent Resistor) in a voltage divider across the Arduino’s external interrupt. But what about the 20-day power requirement? With only seven days to complete the entire project, how could I guarantee that the battery selection of the finished design would last 20 days? With some careful design consideration, reading the specification, and actual measurements, the deliverable design far exceeded the 20-day requirement. The design was validated by actual testing later.”

battery-powered designs — Figure 1: The OLED display showing setup instructions. (Source: author)

Eliminate the USB in favor of FTDI chip

One of the biggest advantages in developing on an Arduino microcontroller board is the ease of writing code using Arduino’s Integrated Development Environment (IDE) and then downloading it the to the Arduino board with a common standard-A to micro-B USB cable. However, like everything else, convenience comes at a cost. Arduinos with onboard USB support will consume additional power, even when they are not powering the Arduino via USB. On an Arduino UNO board, the USB support costs an additional 15mA. Fifteen milliamps may not seem like much, but it could significantly impact the ability of an entire design like the above to last 20 days.

On the other hand, some Arduino boards can be programmed using an FTDI cable. FTDI is a company that’s been manufacturing USB-to-serial converters for several years. (It’s important to only buy genuine FTDI products since counterfeit chips may not work well.) A genuine FTDI cable for this purpose costs about twenty dollars. The FTDI cable itself has an embedded chip that converts the computer-side (host) USB signals to Transistor-Transistor Logic (TTL) serial signals on the Arduino. When the USB circuitry moves from the Arduino to onboard the FTDI cable, there is no additional power drain when the Arduino is being powered by a battery. One such Arduino that supports FTDI programming is the Arduino Pro Mini. The Arduino Pro Mini 328 and Arduino Pro 328 are each available in two versions; both offer a choice of 3.3V running at 8MHz or 5V at 16MHz. So how do you evaluate the best selection for this design?

Figure 2: One of the many FTDI Cables available, this USB-to-TTL serial cable is often used with Arduino boards and has an FTDI FT232RL USB-to-serial chip embedded in the translucent case on the USB side. (Source: Adafruit.com)

Choosing based on voltage level: 5Vdc or 3.3Vdc?

Most microcontrollers designed to run on Arduino boards typically come in either 5Vdc or 3.3Vdc logic level versions. The 5Vdc logic level comes from the legacy Vcc levels of the TTL serial technology dating back to the 1970s. The 5-volt logic level comes from the semiconductor processes and performance characteristics of the time. As such, many peripheral components like sensors and drivers were designed around these logic levels. 5Vdc logic level devices are still prevalent today as fully functional remnants of electronics in earlier times.

Advances in semiconductor processing have allowed lower voltage level highs at 3.3v, where 3.3v became the logical bit “1”, rather than 5v. Using a five-volt blip to register a logic level 1 consumes more energy than a 3.3-volt level for a logic level 1. (Near zero voltage level registers as a logic level 0 for both.) Thus, CMOS technology has created alternatives for TTL to perform the same functions at lower voltage levels. Lower voltage levels equal lower power consumption. Lower power consumption equates to longer battery life. Without any specific power calculations, choosing an Arduino that supports 3.3Vdc logic is the better choice. (Semiconductor process technology has brought voltage supply levels even lower, saving more power, but Arduinos don’t support voltage levels lower than 3.3v .)

Choose 8MHz or 16 MHz?

The driving goal of semiconductor development is faster and faster processor speeds. Other than cost, it seems rare that a designer would ever choose a slower processor speed over a faster one. However, remember that the faster the processor speed, the more heat that is dissipated by the processor. That’s why the latest computers will have a huge fan-cooled heat sink mounted directly on top of the processor chip. Taking a step back and looking at the OLED display design mentioned earlier, nothing in the design requires the top clock speed of 16MHz offered on the Arduino Pro or Mini Pro. The difference is just nanoseconds if you detect the opening of the box using an 8 MHz clock versus a 16 MHz clock, which is imperceivable to the user. The words displayed on the OLED display is in the range of hundreds of milliseconds at a time. By choosing the lower clock speed of 8MHz over the 16MHz, there’s no degradation in performance, and the MCU is consuming less power.

Power consumption evaluation

With our Arduino and OLED display chosen, what is our starting point for sizing the batteries to last 20 days? While we could continue to pour over the component documentation, there is no substitute for measuring the current draw of the circuit with a milliamp meter. Using the milliamp setting of a multimeter and attaching the probes as needed, we can see that we’re drawing about 15mA for the Arduino and OLED display. So, what size and how many batteries will we need to last 20 days? To calculate this, we need to look at the “milliamp-hours” each type of battery can supply.

Figure 3: A digital multimeter displays almost a 15 mA current draw. (Source: author)

Battery mAh capacity by type

While the ambient temperature and current draw influence battery life, it’s simpler to use the “milliamp hour” (mAh) rating for each alkaline battery size in Table 1 below. Note that since mAh refers to the amount of current the batteries can supply, connecting batteries in series to achieve the desired output voltage does not increase the mAh. One AA battery will supply 1.5Vdc for 2400 mA hours. Three AAs in series will supply 4.5Vdc for 2400 mA hours. You would have to connect the batteries in parallel to extend the capacity (in milliamp hours).

Calculate the number of days each battery type will last using the formula:

Days = (Capacity (mAh) / 15mA) x 1 Day/24 hours

You can get a maximum battery life out of the 6-volt lantern battery or a series of D batteries. Two D cells would provide 3 volts. However, that would put the supply at the low-end range of the selected 3.3V Arduino. While it would initially work, the batteries would only have to drain a few hundred millivolts before you’re out of range for a 3.3V Arduino. Therefore, it’s better to use three 1.5-volt batteries so that the batteries can drain by 33% before the supply starts to get into the problematic range.

Arduino deep sleep mode

The above is technically a solution, but a lantern and three D cell batteries are bulky. If bulky batteries are a problem, then you need to look at reducing the current draw of the system by implementing the “deep sleep mode” of the Arduino microcontroller. In deep sleep mode, the processor is halted until an external interrupt wired to the light sensor “wakes up” the Arduino and resumes the program. Since the processor is not running, only a small amount of current is used to monitor the external interrupt. After implementing the deep sleep mode, the supply current measures 1.9mA.

Figure 4: About 1.9 mA current draw in deep sleep mode

You now have more choices to meet the 20-day requirement, as shown in Table 3.

Table 3: Comparison of batteries by type and capacity against a 1.9mA “deep sleep mode” draw to run the MCU.

While AAA batteries meet the 20-day requirement, the AA batteries are a better choice, lasting more than twice the 20-day requirement and taking up not much more room than AAA batteries. AA batteries are significantly smaller than the C, D, or lantern battery choices.

OLED sleep mode

Although the OLED display has a sleep mode, it’s not worth implementing for this use case. While the Arduino is in its sleep mode, the OLED can be set to have all LEDs off. The display’s backlight LEDs in the off state reduce the current that’s drawn by the OLED, but the driver chip may be running at full power, albeit low. Implementing an OLED sleep mode might get the 1.9mA circuit current draw to under 1mA, allowing more confidence in selecting the AAA batteries, but the higher cost of AA batteries is worth the time and effort saved, considering that the scope of the project was a one-off shoe-box message delivery system. If you were building hundreds or thousands of these devices, it would be worth the time to design the lower current draw of the OLED driver, as well. However, for a dozen or so shoe box messaging systems, it isn’t worth it. Judgment calls regarding what is “good enough” to balance cost, schedule, and specification are all part of good engineering.

¹ Franklin Computer was sued by Apple in a landmark case that established that compiled software code and computer operating systems are copyright protected. This forced Franklin Computer out of the desktop computer market. This loss for Franklin Computer is probably why the technician was looking for another job in the mid-1980s. Ref: https://en.wikipedia.org/wiki/Apple_Computer,_Inc._v._Franklin_Computer_Corp

When asked the impossible, ask questions before proceeding

March 6, 2019 By Lynnette Reese Leave a Comment

Sometimes a client will ask for the impossible and expect it at low cost. Or worse yet, the sales engineer in your company will win a project on a low bid because he/she thought it could be accomplished in a “creative” (but impossible) way.

What if you were tasked with a wirelessly communicating embedded project whose electronics, circuit, components, PCB, and all, were required to fit inside a furnace? Typically, a thermocouple or RTD would be placed inside a furnace and wiring would take the sensor signal into a box that holds the electronics. But what if you were asked to create a project where the sensing devices, a 7-segment LED display, and wireless communication were all placed inside the furnace?

Figure 1: The planet Venus has been too hot to explore, because devices would burn up upon approach. (Source: nasa.gov)

Yeah, sometimes engineers get asked to do the atypical. In some cases, the challenge is met with creativity and innovation. And sometimes you must admit that the request can’t be done with today’s technology. With decades of experience, this would be my train of thought if asked to complete such a project.

Think at a higher level: what’s the real goal?

What is the objective here? Is to measure furnace temperature or build an embedded design that can withstand 125°C continuously? So many thoughts come to mind when I read this headline beyond just finding a technical solution for the chosen path. The first engineer I ever worked for gave me several pieces of advice I still practice today. Much of the advice can apply to the headline above.

Like me, you might have heard, “There must be a dozen ways to accomplish the same goal and more than a thousand ways to get it wrong.” Measuring and data logging temperatures well above the operating temperature of the electronic control devices is not a unique problem. Metal smelting furnaces and glass kilns operate between 1,000°C to 1,500°C. The material and walls of these ovens appear bright red at these temperatures! Why limit the suggested design to just 125°C? Why not make it 1,500°C? Obviously, because the design would melt in second; like Arnold Schwarzenegger in the finality of “Terminator 2”.

MIL-SPEC components are available at 125°C. These are considered high-reliability parts since they are designed, tested, and certified to operate continuously at 125°C. Component qualifications are submitted to life tests that subject sample lots to full power at a 125°C ambient temperature for 10,000 hours, perhaps cycling 1 ½ hours on and ½ hour off. The power cycling is supposed to further stress the expansion and contraction of the internal component materials. However, MIL-SPEC components are not actually used at 125°C continuously in practical designs. Rather, the fact that the components will operate at an acceptable failure rate at the extended temperatures, they are even more reliable at lower ambient temperatures.

Just because some electrical components are available at a 125°C rating doesn’t mean that placing them inside the oven they intended to measure is the only solution. Obviously, problems quickly arise trying to source all the components that can withstand the continuous temperature.

If someone asked me to do this project, I would take a step back and look at how others have accomplished this goal. What’s wrong with that? A more traditional approach would be to mount the thermocouples inside the furnace and wire them to the outside of the furnace. Depending on the design requirements, the thermocouple might be just insulted with high-temperature insulation or enclosed in a metal sheath. The thermocouple wiring would be long enough to reach the embedded control or monitoring electronics that are mounted in a more reasonable temperate environment.

The engineer must also consider that there is not always a direct commercial-to-military equivalent for all components. In fact, due to the time it takes to qualify high-reliability components for life testing, MIL-SPEC components often lag commercially available components by years. Many commercial components just don’t have an equivalent due to the demand and cost that it takes to qualify these components. In military, aerospace, and avionics applications, the reliability factor is more important than the latest technology.

What about the mechanics that are related to an embedded design?

Modifying the mechanics of the furnace for wiring access or mounting sheathed temperature probes might quickly get away from the skills and abilities of the electrical engineer who is designing the embedded system. People don’t want to be wrong, especially engineers, and tend to stick with the familiar. An electrical engineer will tend to provide an electrical solution. A mechanical engineer will tend to offer a mechanical solution. A software engineer will give you a software solution but blame the electrical engineer or mechanical engineer if their solution doesn’t work as expected (insert lopsided grin; no one is perfect).

However, the tongue-in-cheek statements above don’t mean that the electrical engineer shouldn’t seek the advice of someone qualified to change or modify the mechanics of the system. Another piece of advice from my first boss was, “You don’t have to know everything, you just need to know where to find the answer.” You got me…. these were in the days before the modern internet, so finding the answer meant more than just googling for answers. There is no substitute for talking to people and asking questions. (For which forums can be a good enough substitute, but not as good as whiteboarding with some experienced engineers.)

Although I’ve seen engineers in some companies hold tightly to their disciplines, the best designs are a balance of mechanics, electronics, and software. There are many advantages to seeking out a multidisciplined balance if you are ever presented with a tuck-it-inside-the-furnace problem. (This is not to say that I haven’t run across mechanical engineers who provide wiring access holes large enough for the wire bundle, but not the terminated connectors, too!) The approach needs to be a cooperative give-and-take of knowledge and ideas.

Who is asking for a narrow solution?

However, exchanging ideas with other engineers isn’t the only problem with the design objective. Is the client asking for the control and datalogger to be installed internal to the furnace without understanding the technical challenges and limitations? Maybe the client isn’t knowledgeable enough about what they are asking for, and the engineer isn’t confident enough to educate them and diplomatically suggest alternatives. There is a difference between doing what the client says and providing what the client needs. Again, if the realistic objective is to control and monitor furnace temperature, why is the engineer trying to mount the embedded controller inside the furnace?

If you look at the requested design, it’s asking for the entire embedded project, including the readout, to be placed inside the furnace. The question should be why you would try to source a seven segment LED display rated for 125°C? You will likely not be able to find one, because a human cannot survive reading an LED display while trapped in a 125°C oven. There is no such Commercial Off the Shelf (COTS) device; it would need to be custom made at a very high price. One application in need of such a rugged LED display might be on the planet Venus.

Better designs exist for the more fixable problems

If mounting several thermocouples externally on a furnace makes the wiring runs too long to multiplex (MUX) them at a single controller, then don’t take that approach. One possible solution could be that separate controllers are needed to read each thermocouple and then upload the data to a central controller. (Definitely cheaper than a custom 125°C tolerant display.) How you solve the problem all depends on the size and layout of the furnace that the engineer is trying to measure.

You don’t have to take the initial approach that’s requested and plow through with a suggested solution when so many others seem possible, which leads me to my final piece of advice from my old boss, “Why is there never enough time to do things right, but always enough to do things over?”Yes, the engineer who tries to put the whole project, electronics and all inside the furnace can finish a functional design, but it will never be reliable.

Taking a step back to seek more traditional solutions would not be a waste of time. Sometimes you have to take the right road and reject the unreasonable request….unless you have massive funds to commission custom devices or decades to do the R&D necessary to create a currently unavailable device.

This FAQ was inspired by a conversation on EDAboard.com. Check it out here.

The best (and nearly) free education that you can get…earning $67K to $103K a year in under 3 months

January 5, 2017 By Lynnette Reese Leave a Comment

If this sounds too good to be true, it is not: read on.

Image of Linux Foundation classes offered at Edx.org. (Source: Screen capture by Author)

Abe Lincoln was one of those guys that didn’t have a college education but made it to the top as a self-educated man. Back then, you didn’t have to have a 4-year college degree to get a decent paying job. Nearly everyone was an entrepreneur. What goes around must come around, then, because more people are selling goods and services as individuals (e.g., Uber, Airbnb, Etsy, eBay), and working remote is growing, as well. As of 2016, 37% of U.S. workers work remote.[i] The workforce is changing, mainly due to the fact that the internet has made it so easy to communicate, buy, and sell.

For bright and motivated “kids” that didn’t get a full ride to college, there’s an alternative to college that can land you a job without a college education: certification. One particular area that is in high demand (such that companies would hire trained monkeys if they could do it) and does not require a college degree is Linux Administration. Glassdoor.com states that the U.S. national average salary is $67,000 for this position. An experienced Linux Administrator, or Senior Linux Administrator, has a national average salary of $103,000 per year in the U.S.[ii] Linux Administrators are in very short supply at present. Those with certification are going to be hired as Junior Admins, but can be worth quite a bit after a couple of years of experience.

Granted, it’s going to cost at least a laptop and a $300 – $500 certification fee to get there, but certification can be achieved in less than a couple of months with The Linux Foundation. The first Linux certification course from The Linux Foundation is estimated to require about 40 – 60 hours of participation time (completed at one’s own pace). This first course is offered free on edX.org, a Massive Open Online Course (MOOC), which is an “affordable and flexible way to learn new skills, pursue [sic] lifelong interests and deliver quality educational experiences at scale,” according to edX, who apparently owns the MOOC.org domain. The extraordinary thing is that most of these courses are not only free, but most are bonafide distance-learning courses taught by prestigious colleges like Stanford and MIT. College students at these universities take them as “correspondence courses,” so to speak.

Introduction to Linux is not for those who don’t know how to use a computer and have no understanding of programming. That’s not a problem, however, as edX.org has classes like Introduction to Computer Science and Introduction to Cloud Computing. (For the uninitiated, “cloud computing” is another term for renting/using remote servers via the Internet.) The second course on the road to becoming a Linux Administrator, or “Admin,” is to take the Essentials of System Administration course at The Linux Foundation (www.linuxfoundation.org).

The work world is changing. We know that neither Steve Jobs or Bill Gates finished college, but both clearly demonstrated the smarts and the motivation to actively pursue their vision of what could be if they stuck with it. The Introduction to Linux course is not for the faint of heart. It is thick with terminology and one is expected to have at least have a laptop running Linux in order to participate in the exercises. But take a look at the job boards and decide for yourself if you have what it takes to learn what you need to pass the certification exam. Since college tuition and the price of textbooks are rapidly outpacing ordinary folks’ economic reach, take heart that most companies are so desperate for a second or third Linux Administrator that they would hire someone without certification as long as they could demonstrate competence. This information is not well-known; skeptics can read the job boards and decide for themselves.

The recent past has the typical professional with a piece of paper to prove they are smart enough and goal-oriented enough to make it through 4 years of higher learning. Companies may auto-reject applicants for lack of a college degree, but certifications represent an important interstitial position between blue and white-collar workers. In addition to intelligence and motivation, opportunity has always been a key factor for those who make it to the top. The internet and online learning have bootstrapped a new class of worker, one who at minimum has access to a computer at the local library, if only to cruise Craigslist for an affordable PC to convert to a Linux machine.

[i] Gallup, I. (2015, August 19). In U.S., Telecommuting for Work Climbs to 37%. Retrieved December 29, 2016, from http://www.gallup.com/poll/184649/telecommuting-work-climbs.aspx

[ii] “Linux Administrator Salaries.” Glass Door. N.p., Nov. 2016. Web. 29 Dec. 2016.

Lynnette Reese

AI-assisted screening for COVID-19 using CT images

The working principle, applications and limitations of ultrasonic sensors

Embedded security: Do you know what you don’t know?

What could happen if there’s an EMP or solar superstorm? Part 2

What happens to an MCU if there’s an EMP or Solar Superstorm? Part 1

Mistakes to avoid when submitting a design for PCB assembly

Understanding wireless across the spectrum

Considerations for battery-powered designs

When asked the impossible, ask questions before proceeding

The best (and nearly) free education that you can get…earning $67K to $103K a year in under 3 months

EE World Online Network

Microcontroller Tips

Lynnette Reese

Footer

EE World Online Network

Microcontroller Tips