With the advent of the GP200 and other AI chip platforms, power consumption in data centers has grown exponentially. The amount of heat generated by these chips can no longer be managed by fans alone. Microchips are now enveloped in cold plates and connected through hoses and liquid cooling connectors to an outer system—typically a coolant distribution unit (CDU)—to circulate the liquid before funneling it back through the server.

As these liquid cooling components take on greater design importance, manufacturers like Amphenol Industrial are advancing fluid connection technologies to meet the needs of this rapidly evolving space.

One of the main considerations for engineers designing liquid cooling systems is intermateability. This is where the Open Compute Project (OCP) comes in. OCP is a collaboration of thought leaders and subject matter experts within the data center, AI, server, and microchip industry. They meet weekly in different workstreams to create a universal spec that can be adopted globally.

For example, the Open Rack Version 3 (ORV3) is a specific server rack that’s designed with defined height, length, number of drawers, and the way liquid cooling flows through it. It is a standard that manufacturers follow so that data center infrastructure globally intersects.

Through OCP, the specs are constantly changing. The ORV3 is going through another iteration where it is getting 6 inches deeper and about 200 pounds heavier because a bus bar has been added, which is now being liquid cooled. These kinds of changes are driven by the next generation of AI chips, which require more power consumption and create more heat.

“Common challenges right now with quick disconnects (QDs) are figuring out how to create more flow with the same or smaller footprint, without having pressure drops,” says Albert Pinto, business development manager at Amphenol Industrial.

Amphenol Industrial’s Universal Quick Disconnect (UQD) follows OCP standards and specs. The first version used a slide latch, which could be hard to disconnect in tight spaces. The second version introduced a push-button latch, which follows the same diameter and spec but is easier to use. According to Pinto, the next version will have chamfers on the plug side of the socket to address friction. The series features a dry-break mechanism and a compact, low-profile latch design to minimize fluid loss and accidental activation.

A second revision of the UQD/UQDB is also in development, aligned with the latest OCP specifications. This update introduces a new interoperability mode in which a UQD plug can be used with a UQDB socket.

In blind-mate systems, where trays slide in and connect without access to the rear of the cabinet, tolerance becomes a determining factor. “What’s interesting about the UQDB—Universal Quick Disconnect Blind Mate—is that it has 1 mm of radial flow misalignment tolerance,” says Pinto. “Even with slight misalignment, the tray will still engage and stay retained by the front of the server tray, whether by bolts or some kind of clamp.”

The next evolution of that design is the Blind Mate Quick Connect (BMQC). “The major difference is that it has 5 mm of radial and 2.7° angular misalignment tolerance,” says Pinto. “It has a much longer funnel and pin.”

The Pivot Blind Mate Coupling (PBMC) is still in development. It maintains a radial tolerance on the pin side but adds tolerance on the socket side as well, splitting the compensation between both ends and enabling a return to a smaller form factor.

Amphenol also plans to launch a new Large Quick Connector (LQC), compliant with OCP standards, in Q4 of this year.

“Liquid cooling in the connector side is evolving at breakneck speed,” Pinto says. “In AI, the heat requirements and space constraints are so demanding that there’s constant evolution in how to keep all these servers functioning by cooling them while maintaining a small footprint.”

Where Amphenol differentiates is in termination options, offering many variations in thread type, barb fits, and right-angle configurations. “One competitor has around 18 part numbers around UQDs and we have about 130,” says Pinto. “Also, our lead time is very competitive: six weeks at volume.”

The company also offers PTFE and EPDM hose solutions, along with custom manifolds for distributing coolant. “We provide complete solutions — from connectors and hoses to fully customized manifold designs for every coolant distribution point in a data center, including CDUs, server racks, immersion cooling tanks, rear‑door heat exchangers and in‑row cooling systems — as well as our standard Blind Mate Manifold (ORV3), an OCP‑compliant design,” says Pinto.

Pinto encourages engineers designing liquid cooling components to immerse themselves in OCP, especially if they work in the AI space.

“Stay in the know, because this industry is moving faster than anything I’ve ever seen,” says Pinto. “Things are becoming obsolete within months because there’s such a higher demand for cooling.”

To learn more about Amphenol Industrial, at TTI.com.

The post Designing fluid connections for AI-driven data centers appeared first on Engineering.com.

From compact sensors embedded in wearable medical devices to rugged components installed in oil and gas platforms, pressure sensors must perform reliably across a broad range of conditions. Each application brings its own demands — whether it’s precision in a dialysis machine, resistance to corrosive media in a hydraulic pump or consistent performance across thousands of test cycles in a validation lab.

Honeywell offers a pressure sensing portfolio that spans the full spectrum of use cases. The company supports customers in healthcare, industrial automation, and test and measurement (T&M) by producing sensors that vary widely in form and function.

“Honeywell’s been designing and manufacturing board mount pressure sensors for more than 40 years,” says Simon Anderson, product manager at Honeywell. “We manufacture our own sense dies that are used in our board mount pressure products and our packaged pressure products. Honeywell advanced sensing technology provides industry-leading accuracy and stability.”

The portfolio includes compact board mount sensors used in medical systems and consumer electronics, packaged pressure sensors for industrial systems, and T&M-grade sensors that span ultra-low to extremely high-pressure ranges.

Board mount sensors in healthcare

In healthcare, where equipment often carries a 10-year service life, sensors must maintain accuracy over time. Honeywell TruStability board mount pressure sensors — composed of the RSC, HSC and SSC Series — are engineered to meet these expectations.

The RSC Series offers the highest level of accuracy and is used in calibration equipment and aerodynamic testing. HSC sensors are temperature-calibrated from 0 to 50°C, making them well suited to critical care systems such as ventilators and dialysis machines. The SSC Series extends the range from -20 to 85°C for broader use in medical and non-medical environments.

“TruStability sensors provide excellent long-term stability, which means our sensors exhibit very low levels of drift,” says Anderson. “This prevents the need for customers to recalibrate equipment and also prevents risky unplanned downtime and reduces service costs.”

For home-care devices such as blood pressure monitors, oxygen concentrators and CPAP systems, Honeywell offers the ABP, ABP2 and MPR Series. These sensors are designed for compact size and cost-effective production in high-volume applications. ABP2 Series also includes extensive media compatibility and optional protective gels for use in applications like emissions monitoring.

To further support medical OEMs, Honeywell provides customization options such as specialized ports and enhanced calibration. These modifications can improve overall system performance while helping manufacturers differentiate and protect their designs.

Packaged pressure sensors in industrial

Industrial systems often require pressure sensors that can operate at higher pressures and withstand exposure to fluids, temperature shifts and mechanical stress. The Honeywell MLH and MIPS product lines offer flexible, durable sensing across a wide range of use cases and environments.

The MLH Series is rated for pressures up to 8,000 PSI and is commonly used in railways, transportation and heavy-duty industrial applications. It supports multiple input voltages and provides both voltage and current outputs for analog systems.

The MIP Series is intended for mid-range pressures, approximately 15 PSI to 1,000 PSI, and is used in HVAC systems, chillers, hydraulic pumps and smart irrigation. It is available in both analog and digital formats, the latter offering reduced power consumption and built-in diagnostic features.

MIPs sensors include a laser-welded, stainless-steel diaphragm and are compatible with common and emerging refrigerants, including R290. This supports long-term sensor performance in systems that involve repeated pressure cycling or exposure to corrosive fluids.

Test & Measurement pressure sensors

Test and measurement environments require sensors that can cover a wide operating range while maintaining consistent performance. The Honeywell T&M portfolio supports measurements from as low as 10 inches of water up to 100,000 PSI, addressing applications across oil and gas, automotive testing and power generation.

All-metal, hermetically welded construction ensures durability and media isolation, while eliminating exposed adhesives or die-attached materials that could degrade over time. These transducers are used in hydraulic and pneumatic test stands, fluid systems and component validation machinery.

“We have a very competitive total error band that is as low as 0.05%,” says Derek Chung, global product application engineer at Honeywell. “With the laser-welded hermetic joints that are completely media-isolated, they’re very good for corrosion resistance.”

Honeywell T&M pressure transducers are certified for hazardous locations and include diagnostic features. The company also extends these products into OEM applications, including aerospace sectors where long-term measurement accuracy and environmental durability are essential.

Beyond its catalog offerings, Honeywell works closely with OEMs to develop tailored sensing solutions that meet application-specific requirements.

“If you visited the Honeywell website and looked at the sensors, you’ll only see half of what we actually build, because a lot of what we do for customers is create those custom solutions,” says Anderson.

To learn more, visit Honeywell at TTI Inc.

The post Pressure sensing across the spectrum of real-world applications—from micro to heavy-duty appeared first on Engineering.com.

DC-DC power supplies have come a long way. Earlier designs were built with large, inefficient components that focused primarily on delivering current and regulating voltage. Recent approaches reflect a different set of priorities. Advances in semiconductor technology are enabling smaller systems with more advanced control, especially for managing transient behavior and load variation.

“With added integration and reduction in size comes the need to run more efficiently, with the ability to support high power density and high current requirements,” says Nalin Boonyong, Senior Sales Engineer at Japan-based electronic components manufacturer TAIYO YUDEN. “This trend drives power supply design to be smaller, high power, faster response and a reliable power source.”

The evolving requirements have made in-house development increasingly difficult. “There’s now a need for engineers to buy off-the-shelf products from power supply companies rather than rely on their own designs, due to the complexity and many variables they need to deal with,” Boonyong says.

These changes are shaping how power circuits are developed in both consumer and automotive applications, with effects reaching every supporting component — including inductors.

Power inductors play a central role in DC-DC switching circuits. They store and release electrical energy in the form of a magnetic field — either stepping down voltage, as in a buck converter, or stepping it up, as in a boost converter, depending on the needs of downstream components. The magnetic field energy smooths ripple currents, stabilizes output voltage and improves circuit efficiency — ensuring consistent, uninterrupted power delivery to integrated circuits.

As electronics continue to shrink, engineers need inductors with smaller footprints to fit within limited board space. Trends in miniaturization include increasing the switching frequency of DC-to-DC converters and in turn requiring smaller, lower-inductance inductors.

Different applications bring different demands. In the case of smart phones, smart devices and small electronics where space is a premium, there is a demand for inductors that are smaller with lower profiles to fit in with limited board space. Servers and EVs need inductors that can support higher current while maintaining high efficiency. In automotive systems, the focus shifts to high reliability and high temperature requirements.

“Engineers designing power inductors face numerous challenges including core loss, saturation, current requirements and thermal performance,” adds Boonyong. “The magnetic material used in inductor development is key and plays a critical role to ensure low core loss, high saturation current, with high permeability and with good thermal performance. On top of this, engineers must balance performance, reliability, low cost and small case size.”

Addressing these trade-offs starts with the material itself. TAIYO YUDEN has developed MCOIL^TM, a proprietary magnetic material which contains iron-based powder mixed with various other metal elements. This material is then coated with thermally stable oxides to form a mixture that has stable temperature characteristics, the ability to handle high current and with low loss magnetic material.

The MCOIL lineup is offered in three form factors. The first is a multilayer SMD chip type, created by printing patterns on ceramic sheets containing MCOIL magnetic material. The sheets are laminated, fired, assembled, and finished with external electrodes.

“This is the world’s first metal multilayer inductor available in the EIA case sizes 0402 and 0806,” says Boonyong. “Its operating temperature goes up to 165°C.”

The second is a high-performance wire wound SMD chip type in which a flat wire wound is formed and then enclosed with MCOIL magnetic material by a soft molding process, then external electrodes are formed at the end. It comes in EIA sizes from 0805 to 1008 and is rated up to 150°C.

The third is a large square type of wire wound SMD inductor formed with a magnetic MCOIL material core through a sintering process. Then the core is trimmed down, and electrodes are formed. Wire is then wrapped around the core and coated with magnetic metal resin.  It is available in JIS sizes from 1.64 mm × 1.64 mm to 4.9 mm × 4.9 mm, with an operating temperature of up to 125°C.

Compared to ferrite inductors, MCOIL components can handle significantly higher current, which allows engineers to use smaller case sizes and reduce board space. Their magnetic saturation characteristics are more gradual, so inductance decreases slowly as current increases, rather than dropping off sharply once the limit is reached. MCOIL inductors also feature lower DC resistance, which helps lower heat generation and in turn improve overall efficiency. Thermal stability is another advantage; unlike ferrite, which can shift performance with temperature, MCOIL maintains consistent inductance under varying thermal conditions. The material’s shielding properties and lower flux leakage also help reduce electromagnetic interference (EMI).

MCOIL inductors are used in a broad range of systems, from consumer electronics and automotive applications to data centers, industrial equipment and energy systems.

“When engineers are working with power inductors, they should be mindful of how different characteristics of the inductor affect circuit performance,” says Boonyong.

Inductance should be selected to balance ripple current and transient response; for more demanding applications, tighter inductance tolerances may be necessary. It is also important not to exceed the inductor’s saturation and RMS current, as this can lead to overheating, circuit malfunction or premature failure. Engineers should determine the maximum temperature the product will face in operation and account for inductor self-heating under maximum current. Choosing core materials with low core loss can also improve power efficiency. EMI can be minimized by selecting shielded inductors with low magnetic flux leakage. All TAIYO YUDEN inductors are shielded.

For your power inductor needs, TAIYO YUDEN has an extensive line up of power inductors to choose from. In a wide range of case sizes from smaller chip type 0402 EIA to larger 8mm x 8mm size, with various inductance range and current handling ratings, in both metal MCOIL and ferrite materials.

To learn more, visit TAIYO YUDEN.

The post Choosing the right power inductor for today’s circuits appeared first on Engineering.com.

Wide band gap semiconductors like gallium nitride appear to be a case where you can have your cake and eat it too. With a high breakdown voltage, and a higher switching frequency compared to silicon, the technology would lend itself to multiple applications, but GaN devices also offer higher power density and high thermal conductivity, making them uniquely adaptable to both power and signal applications.

EEworldonline.com editor-in-chief Aimee Kalnoskas explains how and why it works in conversation with engineering.com’s Jim Anderton. 

For the audio only version:

***

Catch up on the latest engineering innovations with more Industry Insights & Trends videos and podcasts.

The post Why gallium nitride is the next big thing in semiconductors appeared first on Engineering.com.

This episode of Designing the Future is brought to you by New England Wire.

Imagine an electrical design engineering challenge where usual cost and time-to-market pressures co-exist with the need to build equipment on which lives depend on performance and reliability.

That’s the world that medical device designers work in, and like all electrical equipment, moving power and signals through cables is a fundamental part of the product. Safety is always a consideration, but in the medical device industry, designers must also work within a highly regulated environment. Cables must endure everything from sterilizing agents to stray RF radiation, to mechanical shock, yet the choice of cabling is often left late in the engineering design process. There are, however, good reasons to think about cable design in medical equipment early in the process.

Jim Anderton spoke with Joshua Spaulding, Design Engineering Manager with New England Wire Technologies about the challenge of cabling in medical devices, and how to spec the right product. 

* * * 
Learn more about New England Wire’s design and manufacturing of high-performance, custom cables in medical electronics.

The post Medical device design: Cables that deliver in a tough, critical environment  appeared first on Engineering.com.

AllSpice.io, a collaboration platform for electrical engineers, has raised $15 million in Series A funding. The company says that the funding will help it scale its enterprise features and bring the AllSpice AI Agent out of private beta.

Lately many engineering software developers are taking their cues from software in other domains. AllSpice was inspired by software development itself. It resembles the developer platform GitHub, offering a CAD-agnostic repository for design files with version control, branching and merging, comments and comparisons, and more.

And of course, no modern engineering software platform would be complete without some AI thrown in for good measure. AllSpice says its AI Agent is “like adding a superhuman to your team” that can analyze designs for errors, suggest improvements, and create documentation.

“By significantly building out our Gen AI capabilities, largely leveraged by the fact that we understand these design files so well, you can transform this data to present extremely impactful decisions for hardware engineering teams,” Kyle Dumont, AllSpice co-founder and CTO, said in the press release.

The Series A round, which was led by Rethink Impact, brings AllSpice’s total venture capital investments to $25 million.

Siemens introduces Design Copilot NX and other updates

Siemens has announced the latest updates to NX and NX X (now part of the Designcenter suite). Since no modern engineering software platform would be complete without some AI thrown in for good measure, Siemens has introduced the new Design Copilot NX.

“By leveraging natural language input and querying, the NX copilot capabilities enable users to find answers to technical queries, best practices and documentation quickly and efficiently,” according to the Siemens release.

In other words: another product support chatbot. Wake me up when it does something Google couldn’t ten years ago.

(To be fair, NX CAM has some interesting AI features that I learned about at Siemens Realize Live 2025 in Detroit. Stay tuned for upcoming coverage of that.)

Alongside Design Copilot NX, the latest release also includes NX Immersive Collaborator, the ability for multiple NX users to share a simultaneous session in virtual reality; NX Inspector, which extends NX’s model-based definition with characteristics for downstream quality and manufacturing processes; Design for Manufacture (DFM) Advisor, which analyzes part geometry to reveal manufacturing challenges and offer suggestions; an improved NX Mold Wizard with enhanced cooling channel simulation tools; and NX CFD Designer software, a new design simulation tool based on FLOEFD.

You can find more details on the update in Siemens’ announcement.

The engineering AI gap

It turns out there’s a pretty big gap between what engineers expect out of AI and what it can currently do, according to new research from SimScale. The cloud simulation developer commissioned a survey of 300 senior engineers about AI and published their findings in a report called The State of Engineering AI 2025.

The big takeaway is that, while engineers are reporting mild productivity gains today, they believe that the technology can offer much more. Here’s a comparison of their reality against their expectations:

The trick, of course, is how to get from here to there. I recently moderated a webinar that digs deeper into the survey results with SimScale CEO David Heiny and Nvidia distinguished CAE architect Neil Ashton. You can watch it on demand here: Mind the Engineering AI Gap: Why Engineering Teams are Struggling to Realize the AI Opportunity and How to Fix It.

Design and Simulation Week 2025

Engineering.com’s second annual Design and Simulation Week is not just any week, it’s next week.

Starting Monday, July 14, this series of expert webinars will explore the top trends in engineering software from some of the leading voices in the industry (and me). You’ll learn about AI, automation, multiphysics and how to make the most of modern tools.

Register for Design and Simulation Week now.

Quick hits

• Siemens has closed its $5.1 billion acquisition of Dotmatics. That was quick—they announced the deal back in April and didn’t expect to finalize it until at least October.
• Mastercam has released Mastercam 2026, the latest release of its CAM software. The update adds several productivity enhancements alongside a new—make sure you’re sitting down—AI product support chatbot thrown in for good measure, called Mastercam Copilot.
• Onshape just celebrated its 200^th release. Right on, Onshape.

One last link

Engineering.com editor Ian Wright writes on the history and subtypes of electric propulsion in The state of electric propulsion in aircraft.

Got news, tips, comments, or complaints? Send them my way: malba@wtwhmedia.com.

The post AllSpice.io raises $15M to scale GitHub-inspired EE platform appeared first on Engineering.com.

As modern vehicles grow more sophisticated, automakers are integrating an increasing number of electronic features inside the cabin. Infotainment systems, steering wheel controls, LED lighting arrays, heads-up displays, smart mirrors, and power-operated windows and seats all rely on compact, high-performance electronics embedded throughout the vehicle interior.

Unlike “outside-the-box” connectors—those used in safety-critical environments and governed by standards such as USCAR — “inside-the-box” connectors are installed within sealed electronic modules. These internal modules don’t face the same thermal extremes but must still operate under conditions of shock and vibration within a limited space.

To address these constraints, Molex offers a range of miniaturized connectors engineered specifically for use within automotive electronic modules. These products support flexible configurations and incorporate features that help guard against common points of failure—offering practical solutions for a wide variety of in-cabin systems.

“Although automotive standards like USCAR and LV214 aren’t required, we still design our products and test them to many of these standards just as an added layer of assurance to de-risk the connectors in these applications,” says Nathan Piette, Group Product Manager for the Power and Signal business unit at Molex.

Key Molex “Inside the Box” Products

The Micro-Fit 3.0 connector system is a longstanding option in Molex’s compact connector lineup. With a 3.0 mm pitch and current ratings up to 10.5 A per pin, it comes in a wide range of configurations, including wire-to-wire, wire-to-board, and board-to-board. Designers can choose from termination styles such as through-hole, surface-mount, and compliant pin. Most versions are rated to 105°C, with some extending to 125°C. The system supports both V-0 and V-2 resin types and offers either tin or gold terminal plating. While tin is the standard choice for cost reasons, gold offers a corrosion-resistant alternative for harsher environments.

For additional retention strength, an optional terminal position assurance (TPA) feature helps ensure terminals are fully seated during assembly, reducing the risk of intermittent connections caused by incomplete insertion. TPAs also prevent terminals from backing out if cables are tugged or bent after installation.

Micro-Fit+ builds on this platform with improved current handling—up to 13 A per pin, with a 14-gauge option in development that will raise it to 15 A. It also reduces mating force by around 40% compared to standard Micro-Fit and other comparable solutions. Added features include a connector position assurance (CPA) mechanism to reduce the risk of unmating by providing a secondary locking feature, as well as TPA for terminal retention. The entire system is rated to 125°C.

“That’s a T3 level in USCAR automotive parlance,” says Piette. “A lot of inside-the-box applications only require 85°C or 105°C temperature rating. This is a super robust system that far exceeds the performance requirements of the typical use case.”

“It’s a premium product,” adds John Crimmins, Worldwide Account Manager at Molex. “It’s foolproof. You can’t mismate it. It’s the highest power in the industry for something that small.”

Where space constraints are more pressing, the Micro-Lock Plus series offers pitches as small as 1.25 mm and 2 mm, with 1.5 mm on the way. The mated retention force — the force required to pull the connectors apart once they’re engaged — is 49 N, which is unusually high for this class of interconnects.

“When you think of small connector systems, you might think they’re flimsy or maybe delicate,” says Piette. “This is a reliable, robust micro-miniature wire-to-board system.”

The product family also supports potting, so it’s well-suited for customers who use epoxy, conformal coating, or other techniques to seal their boards against environmental ingress. The 1.25 mm version supports up to 3.6 amps per pin; the 2 mm version supports up to 4.7 amps. The series also includes TPA features.

“Molex has the broadest portfolio of micro-miniature wire-to-wire and wire-to-board products from 2 mm pitch and below on the market,” says Piette.

The Pico-Clasp family is Molex’s flagship signal connector in the micro-miniature wire-to-board category. With a 1 mm pitch, it offers one of the most compact footprints in the portfolio. The series includes a wide range of layout and termination styles, including vertical and right-angle orientations, single- and dual-row formats, and surface-mount versions. A variety of locking features are also available, including friction locks for basic retention, and both outer and inner positive locks that provide audible feedback during mating.

“Our over-80-year history of connector design and manufacturing know-how really sets us apart —especially in the power and signal space,” says Piette. “These products are core to our product portfolios overall, and so widely used and applied in the market. We have had a lot of experience and feedback in developing and optimizing these. Micro-Fit has been out for decades. The rest of the world has since copied and pasted that design because of its industry-leading quality and capabilities.”

“While some competitors have a lot of these attributes, rarely any of them have all,” adds Crimmins. “We have the most options, and they’re readily available through TTI with no lead time.”

To learn more, visit Molex at TTI.

The post Advancing automotive electronics by thinking inside the box appeared first on Engineering.com.

Electric vehicles (EVs) continue to grow in popularity and market share, and electric current is the fuel of the future. Current sensors are a critical component of today’s EVs, serving two primary applications according to Ajibola Fowowe, global offering manager at Honeywell.

“The battery management system (BMS) uses current sensors, in conjunction with other sensors such as the voltage and temperature sensors, to monitor the state of charge and overall health of the battery pack. The other use for current sensors is in motor control, where it is relied on to quickly detect and isolate a fault in the electric drive,” Fowowe told engineering.com.

Regardless of use case, there are several considerations EV engineers must understand when selecting among the many available current sensors. Here’s what you need to know.

Types of EV current sensors

There are different types of current sensors that each have advantages and disadvantages for EV applications.

Closed loop current sensors

Closed loop current sensors have a feedback system for improved measurement accuracy. A magnetic core concentrates the magnetic field generated by the flow of current and provides a proportional voltage to the amount of current detected in the core. This enables the sensor to generate a precise current measurement. Because of their high accuracy and stability, closed loop sensors are well suited for use in the BMS.

The Honeywell CSNV 500 is a closed loop current sensor rated for a primary current measurement range of ±500 amps of direct current. The CSNV 500 features a proprietary Honeywell temperature compensation algorithm with digital CAN output, to provide high accuracy readings within ±0.5% error over the temperature range of -40⁰ to 85⁰ C for robust system performance and reliability.

The Honeywell CSNV 500 closed loop current sensor. (Image: Honeywell.)

Open loop current sensors

Open loop current sensors operate on the principle of magnetic induction. They consist of a primary winding, through which the current travels, and a secondary winding that measures the induced voltage. Open loop sensors require less additional electronics and processing compared to closed loop sensors, resulting in faster response times. However, they require additional calibration because they are more prone to variations in heat and magnetic field. This means they are also less accurate — reaching approximately 2% error of the primary readings.

The fast response time of open loop current sensors makes them ideal for motor control functions. Motor control applications don’t require the same level of precision as the BMS, so the loss of accuracy compared to a closed loop or flux gate sensor isn’t critical.

The Honeywell CSHV line of open loop sensors have a range of 100 amps to 1,500 amps, and their response times are as fast as six microseconds. They are used in fault isolation and fault detection, as well as controlling motor speed. They can also be used in battery management systems that do not require very high accuracy, such as in hybrid electric vehicles. These sensors use AEC-Q100 qualified integrated circuits to meet high quality and reliability requirements.

The Honeywell CSHV series open loop sensor. (Image: Honeywell.)

Honeywell’s CSNV 1500 has both closed loop and open loop functionality. This enables the sensor to meet an accuracy requirement of 1%, and is designed for applications that require high accuracy. The CSNV 1500 is used for similar EV applications as the CSNV 500, as well as stationary energy storage systems and industrial operations.

Flux gate current sensors

Flux gate current sensors measure changes in the magnetic flux of a current as it passes through a magnetic loop, from which it can derive current measurements. The Honeywell CSNV 700 is designed for applications that fall between 500 A and 1,000 A requirements. It has a better zero-offset and higher sensing range than 500 amps sensors—but it also has higher power consumption than a closed loop sensor. The CSNV 700 has similar accuracy rating as the CSNV 500, at 0.5%, and it also uses AEC-Q100 qualified integrated circuits.

As with closed loop sensors, the flux gate sensor is best used in BMS settings that require high accuracy. When using flux gate sensors, however, engineers need to be mindful of their higher power requirements, which could consume more battery energy.

Honeywell’s CSSV 1500 is a combination open loop and flux gate sensor. It was designed to meet Automotive Safety Integrity Level C (ASIL-C) requirements for safety-critical applications where customers desire a higher level of reliability and performance. While many 1500 A sensors consume more power, the combination of open loop and flux gate technologies uses less power while still meeting the accuracy and functional safety requirements. It meets Automotive Safety Integrity Level C (ASIL C) requirements for safety critical applications. This requirement is typical of battery electric vehicles (BEV).

Shunt current sensors

A shunt current sensor measures the voltage drop across a sense resistor placed in the conduction path between a power source and a load. It is an inline current sensor connected directly to the busbar; closed loop, open loop and flux gate sensors are non-contact sensors that don’t have that direct connection.

One of the benefits of a shunt sensor is that it can provide an instantaneous measurement of current. However, it generates more heat and contributes to power loss in the circuit. This creates parasitic energy waste. Fowowe says that advancements in shunt technology is increasing its attractiveness in high voltage systems and Honeywell is actively researching additional value that can be derived from the application of shunt technology such as the potential combination of current and voltage measurements into one sensor to reduce the overall cost of the BMS.

Other key considerations for EV current sensors

In addition to considering which sensor to use in which application, engineers will also need to factor in other variables. Since the sensor needs to work properly in a magnetized environment, its capacity to handle magnetic interference is important. For BMS applications that rely on a high level of accuracy, engineers will need to consider the sensor’s zero-offset, which is the amount of deviation in output or reading from the lowest end of the measurement range.

Ease of integration is also important to consider. EVs can use either controller area network (CAN bus) standard or analog outputs. CAN communication is more common in the BMS. CAN bus communication speed is limited by the CAN protocol to 10 milliseconds, which is acceptable for the BMS. For more immediate measurements, motor control functions use analog outputs, which can respond in microseconds.

Another factor to be mindful of is the EV’s driving environment. EVs need to be able to function properly in any conditions, from a heat wave in Arizona to a snowstorm in New York. Therefore, the sensor’s operating temperature range needs to be factored in. According to Fowowe, Honeywell’s sensors are built to maintain performance in temperatures ranging from -40 to 85 degrees Celsius; the sensors feature a Honeywell patented multi-point temperature compensation algorithm to ensure the sensors can deliver very high accuracy and performance under any driving condition.

To learn more about current sensors for EVs, visit Honeywell at TTI.

The post What engineers need to know about current sensors for EV applications appeared first on Engineering.com.

High above contested airspace, a military aircraft locks onto its objective, preparing to engage. But then, systems flicker—navigation data skews, a communication signal drops out. The culprit is electromagnetic interference (EMI), disrupting key onboard functions at the worst possible time.

Electromagnetic compatibility (EMC) refers to a device’s capacity to operate as intended in the presence of external electromagnetic interference, while also avoiding emissions that interfere with nearby systems via conducted or radiated paths. When that balance is off, the result is non-EMC: systems that suffer malfunction, data corruption, or outright failure.

Managing EMI is becoming increasingly difficult as electronics grow more complex and densely integrated. Today’s systems rely heavily on high-speed processors, wireless technology, RF/microwave components, and compact power supplies—all of which are more sensitive to interference while simultaneously being more likely to generate it.

As the potential for EMI increases, it becomes all the more important for engineers to address it early in the design process.

How Spectrum Control Tackles EMI Challenges

Spectrum Control designs EMI solutions for a range of sectors, including military and aerospace, medical and measurement, and telecom, industrial and energy.

“It might be a control circuit for electronic warfare, or it might be an MRI machine—the problems end up being the same no matter what industry you’re in,” says Jeff Chereson, Director of Engineering at Spectrum Control.

The company produces a wide range of EMI mitigation components, including board-level filters, panel mount filters, filtered connectors, and chassis mount power line filters. “The focus is on putting our EMI solutions at the point of entry into the system, where you get maximum effectiveness,” says Chereson.

While Spectrum Control offers off-the-shelf solutions, customization plays an equally important role in their business, if not a larger one.

“A lot of our custom work is derivative of our catalog offerings,” says Matthew McAlevy, Engineering Manager at Spectrum Control. “For example, if you have a D-sub and want selective load filtering—where our catalog D-subs will have the same filter value on all lines, we can customize that and put different circuit values on individual lines. We can do mechanical customizations for different mounting configurations and higher-end sealing or ruggedization.”

These tailored solutions are shaped not just by customers’ preferences, but by their EMC requirements. “Depending on the industry, you’ll get a whole plethora of specs you have to meet,” says Chereson. “We try to get customers to meet their EMC requirements via a filter, but often there are also power, size, safety and ruggedization constraints to work within. Some of the time, we get the specification at the eleventh hour because people don’t realize they have an EMC issue.”

“Doing EMC at the tail end—now you’re trying to shoehorn in a filter, and it’s not costed in your budget,” says McAlevy. “Like with most things in design, the earlier you do it, the better.”

How to Avoid Late-Stage EMI Issues

Spectrum Control recommends taking the following steps during the initial stages of development to avoid EMI headaches down the line:

1. Know your EMI profile and specs: Understand the standards you need to meet, whether it’s MIL-STD-461 for defense, DO-160 for aerospace, FDA for medical devices, or FCC for telecom.
2. Filter at the entry point: Place filters where power or signals enter the system.
3. Design application-specific signal line filters: Tailor the filter response—i.e., the pass band and reject band.
4. Match and balance impedances: Prevent reflections and EMI by ensuring proper system impedances.
5. Apply shielding where necessary: Shield noisy or noise-sensitive modules and interfaces.
6. Use proper grounding techniques: Add low impedance ground planes—avoid large loops.
7. Wrap cables with ferrites to suppress common mode currents: Choose ferrite materials with high loss at EMI-relevant frequencies.
8. Use twisted pair wiring: Twisted pairs reduce magnetic pickup and crosstalk.
9. Limit chassis openings: Keep enclosure apertures small enough to block high-frequency emissions.
10. Use appropriate transient suppressors: Choose components based on energy level and response time: TVS diodes for fast, low-energy events; varistors for medium energy; gas discharge tubes for high-energy pulses like an Electro Magnetic Pulse (EMP).

Designing for Smaller, Faster Systems

As systems evolve, miniaturization is becoming an emerging trend. “When you go up in frequency, things naturally get smaller,” says Chereson. “Because things are faster, they create more EMI, and there’s a need for higher frequency filtering.”

Spectrum Control is addressing these demands with two new standout products: the dual-line coaxial filter and the power circular connector. The dual-line coaxial filter combines the functionality of two single-line filters and a common mode choke within a compact, hermetically sealed panel-mount design, while the power circular connector incorporates a traditional power filter circuit into a form factor traditionally only used for control line filtering. Both products help customers meet SWaP-C goals by reducing size, weight and complexity.

To keep pace with changing systems, Spectrum Control continues to adapt its filtering solutions. “Every year, different platforms and configurations come out,” says Chereson. “We do all kinds of unique shaped filter elements, capacitors, and inductors to fit into different connector sizes.”

To learn more, visit Spectrum Control at TTI.com.

The post How to design with EMI in mind appeared first on Engineering.com.

As industries race to automate, the machine vision market is undergoing rapid change. Advances in artificial intelligence and machine learning — combined with faster processing speeds — are enabling vision systems to come closer to replicating the function of the human eye. High-resolution cameras combined with ultra-fast computing systems are working to capture images and convert them into data that machines can respond to in real time.

But as the technology evolves, so do the demands placed on it. From robotics and drones to autonomous motion equipment, new applications are pushing machine vision systems into more dynamic, high-stress environments.

“Not only is the industry driving higher speeds, but also ruggedness,” says Dave Nyberg, global portfolio manager for industrial electronics at 3M.

Consider a robot in a hospital environment, navigating hallways to deliver prescription medications while carefully avoiding staff and patients. Using machine vision, it continuously scans its surroundings, identifies obstacles, and adjusts its path accordingly. The robot experiences constant movement, and the odd bump or two.

Or it may be a drone surveying crops or delivering products, capturing visual data while experiencing endless vibration and the occasional hard landing.  These types of conditions can prove challenging for internal components, including the cables and connectors that link the vision system to processing units.

“When you think about robotic applications, a typical one is the auto assembly line,” says Nyberg. “You’ve got the big robotic arms moving big parts on the assembly line and riveting or welding the panels onto a car. That arm movement is replacing what a human arm would do; you’ve got to have a lot of range of motion with the shoulder, elbow, wrists — a lot of twisting, movement, flex. Whenever you have robotic applications with a lot of movement, the cables have to withstand that movement because they’re going out to sensors on the end of the arm.”

Similar motions are seen in agricultural automation. “There are a lot of applications now where you think of the movement of a robotic arm and hand that’s picking an apple or strawberry,” says Nyberg. “It’s moving, reaching, grabbing, twisting, turning, pulling it back.  This creates significant stressors such as torsion and flexing of the cable assembly.”

These factors are driving manufacturers to create more robust cable assemblies — and that’s where 3M’s USB3 Vision and CoaXPress solutions come into play.

“Our 3M USB vision cable assemblies are very heavy-duty, high industrial strength cables — very durable,” explains Nyberg. “3M has USB cable assemblies that are tested for over 100 million cycles. What that means is these USB cables have been on drag chain equipment being flexed over and over again, 24 hours a day, seven days a week, for a few years uninterrupted.”

3M’s designs include screw locks at the connector ends to help secure cables during operation. “With vibration concerns, you cannot afford these cables coming loose from a board-mount connector,” says Nyberg.

Right-angle and other connector configurations help address tight space constraints common in compact equipment like drones. Length is another area where 3M’s cable assemblies stand out, offering a level of versatility not typically associated with USB cables.

“Longer length is a real attribute and feature of our products, particularly our USB cables,” says Nyberg. “Typically, USB cables are thought of as — you go over 4 or 5 meters, and most users are thinking of a different interface. But we have passive USB cables that can transmit signals up to 5 Gbps at over 11 to 12 meters, and that’s very unusual in the market. We have a very strong long length cable in our USB product line that meets that requirement.”

Customization of these cable lengths is another key differentiator. Nyberg elaborates: “If you have a piece of equipment and you’re trying to save weight and you have space constraints, we can ship you a 2.5-meter cable, but you may have extra cable there that you can’t afford because of a weight or space issue. We can cut that down and customize to a 2.35-meter cable. You can get any of our cable assemblies customized to an individual length—right down to a fraction of a meter, to a centimeter, or whatever the requirement might be.”

3M’s CoaXPress cable assemblies provide many of the same advantages as the 3M USB cables. Like their USB counterparts, they incorporate features such as secure screw-on or quarter-turn, key-lock mechanisms. They come in various connector types, including Micro-BNC right-angle versions for tight spaces. Their dynamic bending durability is tested to 50 million cycles, a high standard for the coaxial cable industry.

In addition to the applications mentioned earlier, machine vision is transforming healthcare. Through high-resolution imaging, fast interface standards and reliable cable solutions, cameras can now capture extremely high-resolution images of blood samples and cell tissue — enabling remote diagnosis.  In surgical settings, doctors can control robotic arms from thousands of miles away, performing procedures in real time without ever being in the room.

“As this technology continues to evolve, lives will be saved,” says Nyberg. “People will be able to access the medical community that was out of reach before, due in large part through the advancements in machine vision and highly durable cable assemblies.  It’s really quite amazing.”

To learn more, visit 3M at TTI, Inc.

The post Navigating the demands of modern machine vision movement appeared first on Engineering.com.