Electronics Grounding (EE Tip #107)

Whether you are professional electrical engineer or part-time DIYer, before you start your next project, read through this primer on grounding. This short survey covers one of the most fundamental topics in electronics: grounding.

Electronics Signal Ground or Circuit Common

Signal ground is the current return to the power supply. Current leaves the power supply, passes through the various electronic components, and then returns to the supply. The typical symbol for signal ground is shown in Figure 1.EE107-F1-2

 Chassis Ground or Earth Ground

Chassis ground is an electrical safety requirement to prevent an electrical or electronic device’s chassis from delivering an electrical shock. A long copper rod is driven into the ground outside of the building, and a wire connects the metal chassis to the rod which is at the approximate 0 V potential of the earth. The symbol for earth ground is shown in Figure 2.

Ground Details

Consider the following two details about ground. First, ground is not exactly 0 V. And second, two physically different ground points will not be at the same voltage potential.

Ground Loop

By definition, current will flow in an electrical conductor connected to a difference in voltage potential between two points. Because two physically different ground points are not at the same potential, current will flow through an electrical conductor connected between those two points. This is a ground loop.

Notice this current flowing between these two different ground points is not related to or correlated to any electronic data or message signal. This is noise or garbage that will interfere and distort any information contained in the electronic system.

Note: While “noise” can be added to systems on occasion, it is specifically controlled and the exact quantity is regulated.

Example

Given: A ground loop producing 610 μV of ground noise. It’s a very small quantity. You have a 16-bit A/D converter with a 0- to 10-V input. The smallest voltage it can resolve is:

= 10 V/16 exp 2

= 10 V/65,536

= 152.5ìV

Note that the ground loop noise is four times greater than the actual data, so that A/D converter loses two bits of resolution, and it is now a 14-bit converter.

Connect with Single-Ended/Unbalanced Amps

In Figure 3 the two grounds exist at different potentials, so some current will flow between the grounds. EE107-F3

This ground current has nothing to do with any signals being amplified, and it is noise decreasing the accuracy of the system. Figure 4 is a complete schematic.EE107-F4

Connect with Transformers

When connecting with transformers, keep the following in mind:

  • There is no ground connection, so there can be no Ground Loop.
  • Common-mode rejection of RF interference.
  • Signals are AC coupled, so of limited use for circuits with DC data such as accelerator focus and bend magnets (see Figure 5).EE107-F5

Connect with Differential Amps

Refer to Figure 6 for connecting two systems with differential amplifiers.

  • There is no ground connection, so there can be no Ground Loop.
  • Common-mode rejection of RF interference (see Figure 7).
  • Signals are DC coupled, so this is the perfect solution for circuits with DC data.EE107-F6EE107-F7

—Dennis Hoffman

Note: This article first appeared in audioXpress  (June 2011). It is from a class that Dennis Hoffman teaches at the SLAC National Accelerator Laboratory (Menlo Park, CA). Like Circuit Cellar, audioXpress is Elektor International Media Publication.

3-D Printing with Liquid Metals

by Collin Ladd and Michael Dickey

Our research group at North Carolina State University has been studying new ways to use simple processes to print liquid metals into 3-D shapes at room temperature. 3-D printing is gaining popularity because of the ability to quickly go from concept to reality to design, replicate, or create objects. For example, it is now possible to draw an object on a computer or scan a physical object into software and have a highly detailed replica within a few hours.

3-D printing with liquid metals: a line of dollsMost 3-D printers currently pattern plastics, but printing metal objects is of particular interest because of metal’s physical strength and electrical conductivity. Because of the difficulty involved with metal printing, it is considered one of the “frontiers” of 3-D printing.
There are several approaches for 3-D printing of metals, but they all have limitations, including high temperatures (making it harder to co-print with other materials) and prohibitively expensive equipment. The most popular approach to printing metals is to use lasers or electron beams to sinter fine metal powders together at elevated temperatures, one layer at a time, to form solid metal parts.

Our approach uses a simple method to enable direct printing of liquid metals at room temperature. We print liquid metal alloys primarily composed of gallium. These alloys have metallic conductivity and a viscosity similar to water. Unlike mercury, gallium is not considered toxic nor does it evaporate. We extrude this metal from a nozzle to create droplets that can be stacked to form 3-D structures. Normally, two droplets of liquid (e.g., water) merge together into a single drop if stacked on each other. However, these metal droplets do not succumb to surface-tension effects because the metal rapidly forms a solid oxide “skin” on its surface that mechanically stabilizes the printed structures. This skin also makes it possible to extrude wires or metal fibers.

This printing process is important for two reasons. First, it enables the printing of metallic structures at room temperature using a process that is compatible with other printed materials (e.g., plastics). Second, it results in metal structures that can be used for flexible and stretchable electronics.

 

Stretchable electronics are motivated by the new applications that emerge by building electronic functionality on deformable substrates. It may enable new wearable sensors and textiles that deform naturally with the human body, or even an elastic array of embedded sensors that could serve as a substitute for skin on a prosthetic or robot-controlled fingertip. Unlike the bendable polyimide-based circuits commonly seen on a ribbon cable or inside a digital camera, stretchable electronics require more mechanical robustness, which may involve the ability to deform like a rubber band. However, a stretchable device need not be 100% elastic. Solid components embedded in a substrate (e.g., silicone) can be incorporated into a stretchable device if the connections between them can adequately deform.

Using our approach, we can direct print freestanding wire bonds or circuit traces to directly connect components—without etching or solder—at room temperature. Encasing these structures in polymer enables these interconnects to be stretched tenfold without losing electrical conductivity. Liquid metal wires also have been shown to be self-healing, even after being completely severed. Our group has demonstrated several applications of the liquid metal in soft, stretchable components including deformable antennas, soft-memory devices, ultra-stretchable wires, and soft optical components.

Although our approach is promising, there are some notable limitations. Gallium alloys are expensive and the price is expected to rise due to gallium’s expanding industrial use. Nevertheless, it is possible to print microscale structures without using much volume, which helps keep the cost down per component. Liquid metal structures must also be encased in a polymer substrate because they are not strong enough to stand by themselves for rugged applications.

Our current work is focused on optimizing this process and exploring new material possibilities for 3-D printing. We hope advancements will enable users to print new embedded electronic components that were previously challenging or impossible to construct using a 3-D printer.

Collin Ladd (claddc4@gmail.com)  is pursuing a career in medicine at the Medical University of South Carolina in Charleston, SC. Since 2009, he has been the primary researcher for the 3-D printed liquid metals project at The Dickey Group, which is headed by Michael Dickey. Collin’s interests include circuit board design and robotics. He has been an avid electronics hobbyist since high school.

Collin Ladd (claddc4@gmail.com) is pursuing a career in medicine at the Medical University of South Carolina in Charleston, SC. Since 2009, he has been the primary researcher for the 3-D printed liquid metals project at The Dickey Group, which is headed by Michael Dickey. Collin’s interests include circuit board design and robotics. He has been an avid electronics hobbyist since high school.

Michael Dickey (mddickey@ncsu.edu) is an associate professor at the North Carolina State University Department of Chemical and Biomolecular Engineering. His research includes studying soft materials, thin films and interfaces, and unconventional nanofabrication techniques. His research group’s projects include stretchable electronics, patterning gels, and self-folding sheets.

Michael Dickey (mddickey@ncsu.edu) is an associate professor at the North Carolina State University Department of Chemical and Biomolecular Engineering. His research includes studying soft materials, thin films and interfaces, and unconventional nanofabrication techniques. His research group’s projects include stretchable electronics, patterning gels, and self-folding sheets.

 

 

 

Q&A: Jeremy Blum, Electrical Engineer, Entrepreneur, Author

Jeremy Blum

Jeremy Blum

Jeremy Blum, 23, has always been a self-proclaimed tinkerer. From Legos to 3-D printers, he has enjoyed learning about engineering both in and out of the classroom. A recent Cornell University College of Engineering graduate, Jeremy has written a book, started his own company, and traveled far to teach children about engineering and sustainable design. Jeremy, who lives in San Francisco, CA, is now working on Google’s Project Glass.—Nan Price, Associate Editor

NAN: When did you start working with electronics?

JEREMY: I’ve been tinkering, in some form or another, ever since I figured out how to use my opposable thumbs. Admittedly, it wasn’t electronics from the offset. As with most engineers, I started with Legos. I quickly progressed to woodworking and I constructed several pieces of furniture over the course of a few years. It was only around the start of my high school career that I realized the extent to which I could express my creativity with electronics and software. I thrust myself into the (expensive) hobby of computer building and even built an online community around it. I financed my hobby through my two companies, which offered computer repair services and video production services. After working exclusively with computer hardware for a few years, I began to dive deeper into analog circuits, robotics, microcontrollers, and more.

NAN: Tell us about some of your early, pre-college projects.

JEREMY: My most complex early project was the novel prosthetic hand I developed in high school. The project was a finalist in the prestigious Intel Science Talent Search. I also did a variety of robotics and custom-computer builds. The summer before starting college, my friends and I built a robot capable of playing “Guitar Hero” with nearly 100% accuracy. That was my first foray into circuit board design and parallel programming. My most ridiculous computer project was a mineral oil-cooled computer. We submerged an entire computer in a fish tank filled with mineral oil (it was actually a lot of baby oil, but they are basically the same thing).

DeepNote Guitar Hero Robot

DeepNote Guitar Hero Robot

Mineral Oil-Cooled Computer

Mineral Oil-Cooled Computer

NAN: You’re a recent Cornell University College of Engineering graduate. While you were there, you co-founded Cornell’s PopShop. Tell us about the workspace. Can you describe some PopShop projects?

Cornell University's PopShop

Cornell University’s PopShop

JEREMY: I recently received my Master’s degree in Electrical and Computer Engineering from Cornell University, where I previously received my BS in the same field. During my time at Cornell, my peers and I took it upon ourselves to completely retool the entrepreneurial climate at Cornell. The PopShop, a co-working space that we formed a few steps off Cornell’s main campus, was our primary means of doing this. We wanted to create a collaborative space where students could come to explore their own ideas, learn what other entrepreneurial students were working on, and get involved themselves.

The PopShop is open to all Cornell students. I frequently hosted events there designed to get more students inspired about pursuing their own ideas. Common occurrences included peer office hours, hack-a-thons, speed networking sessions, 3-D printing workshops, and guest talks from seasoned venture capitalists.

Student startups that work (or have worked) out of the PopShop co-working space include clothing companies, financing companies, hardware startups, and more. Some specific companies include Rosie, SPLAT, LibeTech (mine), SUNN (also mine), Bora Wear, Yorango, Party Headphones, and CoVenture.

NAN: Give us a little background information about Cornell University Sustainable Design (CUSD). Why did you start the group? What types of CUSD projects were you involved with?

CUSD11JEREMY: When I first arrived at Cornell my freshman year, I knew right away that I wanted to join a research lab, and that I wanted to join a project team (knowing that I learn best in hands-on environments instead of in the classroom). I joined the Cornell Solar Decathlon Team, a very large group of mostly engineers and architects who were building a solar-powered home to enter in the biannual solar decathlon competition orchestrated by the Department of Energy.

By the end of my freshman year, I was the youngest team leader in the organization.  After competing in the 2009 decathlon, I took over as chief director of the team and worked with my peers to re-form the organization into Cornell University Sustainable Design (CUSD), with the goal of building a more interdisciplinary team, with far-reaching impacts.

CUSD3

Under my leadership, CUSD built a passive schoolhouse in South Africa (which has received numerous international awards), constructed a sustainable community in Nicaragua, has been the only student group tasked with consulting on sustainable design constraints for Cornell’s new Tech Campus in New York City, partnered with nonprofits to build affordable homes in upstate New York, has taught workshops in museums and school, contributed to the design of new sustainable buildings on Cornell’s Ithaca campus, and led a cross-country bus tour to teach engineering and sustainability concepts at K–12 schools across America. The group is now comprised of students from more than 25 different majors with dozens of advisors and several simultaneous projects. The new team leaders are making it better every day. My current startup, SUNN, spun out of an EPA grant that CUSD won.

CUSD7NAN: You spent two years working at MakerBot Industries, where you designed electronics for a 3-D printer and a 3-D scanner. Any highlights from working on those projects?

JEREMY: I had a tremendous opportunity to learn and grow while at MakerBot. When I joined, I was one of about two dozen total employees. Though I switched back and forth between consulting and full-time/part-time roles while class was in session, by the time I stopped working with MakerBot (in January 2013), the company had grown to more than 200 people. It was very exciting to be a part of that.

I designed all of the electronics for the original MakerBot Replicator. This constituted a complete redesign from the previous electronics that had been used on the second generation MakerBot 3-D printer. The knowledge I gained from doing this (e.g., PCB design, part sourcing, DFM, etc.) drastically outweighed much of what I had learned in school up to that point. I can’t say much about the 3-D scanner (the MakerBot Digitizer), as it has been announced, but not released (yet).

The last project I worked on before leaving MakerBot was designing the first working prototype of the Digitizer electronics and firmware. These components comprised the demo that was unveiled at SXSW this past April. This was a great opportunity to apply lessons learned from working on the Replicator electronics and find ways in which my personal design process and testing techniques could be improved. I frequently use my MakerBot printers to produce custom mechanical enclosures that complement the open-source electronics projects I’ve released.

NAN: Tell us about your company, Blum Idea Labs. What types of projects are you working on?

JEREMY: Blum Idea Labs is the entity I use to brand all my content and consulting services. I primarily use it as an outlet to facilitate working with educational organizations. For example, the St. Louis Hacker Scouts, the African TAHMO Sensor Workshop, and several other international organizations use a “Blum Idea Labs Arduino curriculum.” Most of my open-source projects, including my tutorials, are licensed via Blum Idea Labs. You can find all of them on my blog (www.jeremyblum.com/blog). I occasionally offer private design consulting through Blum Idea Labs, though I obviously can’t discuss work I do for clients.

NAN: Tell us about the blog you write for element14.

JEREMY: I generally use my personal blog to write about projects that I’ve personally been working on.  However, when I want to talk about more general engineering topics (e.g., sustainability, engineering education, etc.), I post them on my element14 blog. I have a great working relationship with element14. It has sponsored the production of all my Arduino Tutorials and also provided complete parts kits for my book. We cross-promote each-other’s content in a mutually beneficial fashion that also ensures that the community gets better access to useful engineering content.

NAN: You recently wrote Exploring Arduino: Tools and Techniques for Engineering Wizardry. Do you consider this book introductory or is it written for the more experienced engineer?

JEREMY: As with all the video and written content that I produce on my website and on YouTube, I tried really hard to make this book useful and accessible to both engineering veterans and newbies. The book builds on itself and provides tons of optional excerpts that dive into greater technical detail for those who truly want to grasp the physics and programming concepts behind what I teach in the book. I’ve already had readers ranging from teenagers to senior citizens comment on the applicability of the book to their varying degrees of expertise. The Amazon reviews tell a similar story. I supplemented the book with a lot of free digital content including videos, part descriptions, and open-source code on the book website.

NAN: What can readers expect to learn from the book?

JEREMY: I wrote the book to serve as an engineering introduction and as an idea toolbox for those wanting to dive into concepts in electrical engineering, computer science, and human-computer interaction design. Though Exploring Arduino uses the Arduino as a platform to experiment with these concepts, readers can expect to come away from the book with new skills that can be applied to a variety of platforms, projects, and ideas. This is not a recipe book. The projects readers will undertake throughout the book are designed to teach important concepts in addition to traditional programming syntax and engineering theories.

NAN: I see you’ve spent some time introducing engineering concepts to children and teaching them about sustainable engineering and renewable energy. Tell us about those experiences. Any highlights?

JEREMY: The way I see it, there are two ways in which engineers can make the world a better place: they can design new products and technologies that solve global problems or they can teach others the skills they need to assist in the development of solutions to global problems. I try hard to do both, though the latter enables me to have a greater impact, because I am able to multiply my impact by the number of students I teach. I’ve taught workshops, written curriculums, produced videos, written books, and corresponded directly with thousands of students all around the world with the goal of transferring sufficient knowledge for these students to go out and make a difference.

Here are some highlights from my teaching work:

bluestamp

I taught BlueStamp Engineering, a summer program for high school students in NYC in the summer of 2012. I also guest-lectured at the program in 2011 and 2013.

I co-organized a cross-country bus tour where we taught sustainability concepts to school children across the country.

indiaI was invited to speak at Techkriti 2013 in Kanpur, India. I had the opportunity to meet many students from IIT Kanpur who already followed my videos and used my tutorials to build their own projects.

Blum Idea Labs partnered with the St. Louis Hacker Scouts to construct a curriculum for teaching electronics to the students. Though I wasn’t there in person, I did welcome them all to the program with a personalized video.

brooklyn_childrens_zoneThrough CUSD, I organized multiple visits to the Brooklyn Children’s Zone, where my team and I taught students about sustainable architecture and engineering.

Again with CUSD, we visited the Intrepid museum to teach sustainable energy concepts using potato batteries.

intrepid

NAN: Speaking of promoting engineering to children, what types of technologies do you think will be important in the near future?

JEREMY: I think technologies that make invention more widely accessible are going to be extremely important in the coming years. Cheaper tools, prototyping platforms such as the Arduino and the Raspberry Pi, 3-D printers, laser cutters, and open developer platforms (e.g., Android) are making it easier than ever for any person to become an inventor or an engineer.  Every year, I see younger and younger students learning to use these technologies, which makes me very optimistic about the things we’ll be able to do as a society.

Multi-Tasking Robot Platform

Fisnar F9960N

Fisnar F9960N

The F9960N multitasking robot is designed to dispense applications including miniature SMT circuit boards, large PCBs, and finished assemblies. The robot’s dispensing and coating system can be integrated within a conveyor-dependent inline manufacturing environment or installed as a stand-alone module.

The enclosed environment provides access for fume extraction systems, which creates a safeguard from potentially hazardous substances. Access to the working area is through a security door, which is locked while in operation but accessible during programming.

The robot includes a 178-mm touchscreen display that enables you to program a dispense path with unique characteristics (e.g., continuous path and point-to-point routing).

Contact Fisnar for pricing.

Fisnar, Inc.
www.fisnar.com

Low-Cost, High-Performance 32-bit Microcontrollers

The PIC32MX3/4 32-bit microcontrollers are available in 64/16-, 256/64-, and 512/128-KB flash/RAM configurations. The microcontrollers are coupled with Microchip Technology’s software and tools for designs in connectivity, graphics, digital audio, and general-purpose embedded control.

The microcontrollers offer high RAM memory options and high peripheral integration at a low cost. They feature 28 10-bit ADCs, five UARTS, 105-DMIPS performance, serial peripherals, a graphic display, capacitive touch, connectivity, and digital audio support.
The PIC32MX3/4 microcontrollers are supported with general software development tools, including Microchip Technology’s MPLAB X integrated development environment (IDE) and the MPLAB XC32 C/C++ compiler.

Application-specific tools include the Microchip Graphics Display Designer X and the Microchip Graphics Library, which provide a visual design tool that enables quick and easy creation of graphical user interface (GUI) screens for applications. The microcontrollers are also supported with a set of Microchip’s protocol stacks including TCP/IP, USB Device and Host, Bluetooth, and Wi-Fi. For digital audio applications, Microchip provides software for tasks such as sample rate conversion (SRC), audio codecs—including MP3 and Advanced Audio Coding (AAC), and software to connect smartphones and other personal electronic devices.

The PIC32MX3/4 family is supported by Microchip’s PIC32 USB Starter Kit III, which costs $59.99 and the PIC32MX450 100-pin USB plug-in module, which costs $25 for the modular Explorer 16 development system. Pricing for the PIC32MX3/4 microcontrollers starts at $2.50 each in 10,000-unit quantities.

Microchip Technology, Inc.
www.microchip.com

Industrial Double-Break Switch for IP67/68

The MP220 is an industrial high-current switch. It features a single-pole, double-break electrical circuit that enables it to double break two independent electrical circuits.

The MP220 switch utilizes a fast snap action mechanism to transition from break-to-make position. The switch in its standard version is currently IP67 sealed. An improved sealing for IP68 applications with a strengthened cable exit is also available.

The MP220 is available in standard microswitch-size housing with a range of possible actuators and a four-wire potted cable. The switch is EN61058-certified for 250 VAC/10 A with a –40°C-to-130°C operating temperature range.

Contact Microprecision for pricing.

Microprecision Electronics SA
www.microprecision.ch

CC276: Not a Hockey Fan?

Hockey can be fun, unless you’re building a surface-mount device (SMD) prototype and the “puck” is one of the tiny components getting away from your soldering iron. In an article appearing last month in Circuit Cellar, “DIY?Surface-Mount Circuit Boards: Tips and Tricks for Building SMD Prototypes,” engineer James Lyman inadvertently sparked a bit of debate on the magazine’s website. Readers posted various alternatives to Lyman’s approach to the “SMD hockey” challenge. Here’s how Lyman’s article describes the problem and his solution:

“When I built my first few surface-mount boards, I did what so many amateurs and technicians do. I carefully placed each minute component on the circuit board in its correct position, and then spent several minutes playing ‘SMD hockey.’ With nothing holding the component in place, I’d take my soldering iron and heat the pad component while touching the solder to the junction. Just as the solder was about to melt, that little component would turn into a ‘puck’ and scoot away. Using the soldering iron’s tip as a ‘hockey stick,’ I’d chase the little puck back to its pads and try again, which was maddening…

“It slowly occurred to me that I needed something to hold each part in place while soldering—something that would glue them in place. Commercial houses glue the components down on the boards and then use a wave soldering machine, which does all the soldering at once. That’s exactly what I started doing. I use J-B Weld, a common off-the-shelf epoxy.”

Here is a sampling of alternative solutions readers posted to circuitcellar.com.

  • From Bill: “If you must use epoxy, then the cheapest fast-setting epoxy from Poundland will do the trick.
    “Personally, I’ve always used a tiny spot of cheapo CA superglue, which gives you 20-60 s to position the component. If there are a lot of SMDs on the board, you might want to use an accelerant spray to reduce the CA cure to 5 or 10 s. If you can’t afford proper CA accelerant, then isoprop or a gentle waft above the board with a cloth soaked in a little household ammonia will do the trick.”
  • From Trevor: ”I did a lot of hand soldering of SM parts years ago and agree that it is best to fix the parts before soldering.
    “I used an adhesive made for SM parts from RS Components, which comes in a syringe and is really easy to use. ” (Trevor’s post provides a link to his preferred Electrolube brand.)
  • From Kevin: ”Crikey, epoxying all the components first is a bit brutal. What if you want to change one? Melt the solder and twist, all at the same time?
    “Much easier to tin one pad, then place the part on it with tweezers and touch it with the iron, one end soldered fine, now solder the other end.”

Feel free to visit circuitcellar.com to weigh in or take some of the advice offered there.

DIY Surface-Mount Circuit Boards

James Lyman, an engineer with degrees in Aerospace, Electrical Engineering, and Systems Design, has more than 35 years of design experience but says he was “dragged” over the past decade into using surface-mount devices (SMD) in his prototypes. He had a preference for using through-hole technology whenever possible.

“The reasons are simple,” he says in an article appearing in the June issue of Circuit Cellar magazine. “It’s much easier to use traditional components for building and reworking prototype circuits than it is to use wire to make the connections. Plus, the devices are large and easy to handle. But time and technology don’t leave anyone at peace, so my projects have gradually drifted toward surface-mount design.”

In his article, Lyman shares the techniques he developed for designing prototypes using SMD components. He thought sharing what he learned would make the transition less daunting for other designers.

This accompanying photo shows one of his completed circuit board designs.

Lyman’s techniques developed out of trial and error. One trial involved keeping small components in place during the building of his prototype.

“When I built my first few surface-mount boards, I did what so many amateurs and technicians do. I carefully placed each minute component on the circuit board in its correct position, and then spent several minutes playing ‘SMD hockey,’ ” Lyman says. “With nothing holding the component in place, I’d take my soldering iron and heat the pad component while touching the solder to the junction. Just as the solder was about to melt, that little component would turn into a ‘puck’ and scoot away. Using the soldering iron’s tip as a ‘hockey stick,’ I’d chase the little puck back to its pads and try again, which was maddening. Finally, I’d get a drop of solder holding one end of the puck in place, usually with the other end sticking away from its pad. Then I could reheat the solder joint while holding the puck and position it correctly. I would have to start over with the next component, all the while yearning for that wonderful old through-hole technology.

“It slowly occurred to me that I needed something to hold each part in place while soldering—something that would glue them in place. Commercial houses glue the components down on the boards and then use a wave soldering machine, which does all the soldering at once. That’s exactly what I started doing. I use J-B Weld, a common off-the-shelf epoxy.”

Using an easy-to-get epoxy is just one of the tips in Lyman’s article. For the rest, check out his full article in the June issue of Circuit Cellar.

 

Modify & Test a Phase Meter Calibrator

Charles Hansen described a DIY phase meter calibrator using all-pass, phase-shift filters in a November 2006 article published in audioXpress magazine. Being able to measure phase angle is often helpful, so I’ll begin by quoting from the beginning of his article:

“A phase angle meter is useful in audio work to determine the phase angle between a reference signal and a phase shifted signal, both having identical time periods. Typical uses include: Finding the phase angle between voltage and current to determine the phase shift and impedance of a loudspeaker over its frequency range. Finding the phase shift between the input and output of a tube amplifier to establish the HF (high frequency) and LF (low frequency) cutoff points needed to avoid instability in feedback amplifiers.”

In addition to these, there are other uses—for example, measuring the phase shift through any active or passive filter which includes equalization networks.

In his design, he chose a set of five calibrations frequencies: 10 Hz, 100 Hz, 1000 Hz, 10 kHz, and 100 kHz. He relied on an external oscillator to drive the calibrator at these input frequencies. I first built the calibrator as described, and then I made some modifications that better suited my needs. But first I will describe how the calibrator works. I think it’s best to just provide a bit more  from Hansen’s article:

“The Phase Angle Calibrator makes use of an op amp filter circuit called the all-pass circuit, which takes a sine-wave input and produces a constant amplitude phase-shifted sine wave output. The lag output version was used in Fig. 1. The theory behind the all-pass filter is available in many reference books and texts, but I found one by Walt Jung [1] that I believe is the easiest for a novice to understand. The phase shift angle is varied by the parallel combination of R3 and R9 through R19 with C3 through C7 in accordance with the formula:

θ = -2 arctan (2ΠRC)

where θ is the phase angle, and f is the frequency. After selecting a suitable value for C, you can solve for R by rearranging the formula:

R = tan(-θ/2) / 2ΠfC

This is hardly a linear relationship. Large changes in resistor value produce very little change in phase angle as you approach 0 or 180 degrees. It’s much easier to apply the input signal to both inputs of the phase angle meter for zero degrees, and use an op-amp inverter to generate the 180 degree signal.”

MODIFY THE CALIBRATOR

I added an internal Wein-bridge oscillator to simplify using the calibrator and I changed the set of frequencies to cover just the audio range: 20 Hz, 100 Hz, 1000 Hz, 10 kHz, and 20 kHz. (This range is also easier to cover with a single-range oscillator, the capacitor values stay reasonable.) The actual frequencies, shown in Table 1, vary somewhat from the ideal frequencies because I used standard 1% resistors and 5% capacitors.

Table 1: These are phase calibrator phase-shift measurements. The column labeled “305” refers to the Dranetz model 305 phase meter with 305-PA-3007 plug-in. The column labeled “5245L” refers to the Hewlett-Packard model 5245L frequency counter with a model 5262A time interval unit plug-in. Phase shift measurements at 19.6 kHz are not useful from the HP-5245L counter because the 10-MHz timebase does not provide enough significant figures.

Selecting and matching the capacitors would give closer results, but it’s more important to know what the frequencies are. Because I had already built a circuit board for the calibrator circuit, I used a second circuit board for the oscillator. Figure 1 and Figure 2 are the two circuit diagrams.

Figure 1: Phase angle calibrator using all-pass phase-shift filters. This is a Charles Hansen design, 2005, with circuit board design by the author (PHASECAL.PCB).

Figure 2: Wein-bridge oscillator with lamp amplitude stabilization. Adjust R6 for minimum harmonic distortion. (MAIN115.PCB)

Tables 2 and Table 3 are the parts lists. Please note that the calibrator circuit is unchanged from Hansen’s design, except for the values of the capacitors C3 and C7. In the original, C3 was 470 nF (for 10 Hz) and C7 was 47 pF (for 100 kHz). I put both circuit boards and a ±15-VDC power supply (any regulated supply will suffice) in a Wolgram MC-9 enclosure.

Table 2: Calibrator parts list

Table 3: Wein-bridge oscillator parts list

The completed calibrator is shown in Photo 1 with a Dranetz Phase meter. (More about this later.) The unlabeled knob, lower left in the photo, is an oscillator output level control (R8 in Figure 2), which I added after making the front panel label.

Photo 1: A Dranetz automatic phase meter, model 305, is at the top. The phase meter calibrator is below. The calibrator, a Charles Hansen design, is not a TDL product, but construction details are included in this article.

Wein-bridge oscillator theory is discussed in many textbooks and is rather mathematical. I will describe it as simply as possible. In Figure 2 the oscillation frequency is set by the value of R and C connected between the op-amp non-inverting input (pin 3), the op-amp output (pin 6), and common. For a frequency of 1,000 Hz, C = 22 nF (C3 and C10) and R = 7235 Ω (the series combination of R1 + R2 and R3 + R4). The equation is:

f = 1/(2ΠRC) = 1/(2Π(22 x 10-9) (7235)) = 1000 Hz

For amplitude-stable oscillation to occur, the gain of the op-amp circuit must be 1/3. This is set by the impedance of the RC network and resistors R5, R6 and R7 and the incandescent lamp. The lamp is important because it stabilizes the gain at 1/3. If the output voltage (pin 6) tries to increase, the lamp’s resistance decreases and the output voltage decreases. This works very well, but it takes the output amplitude a small of amount of time to stabilize, especially at low frequencies. The CM6833 lamp is very small, so its thermal time constant is very low and stability happens very quickly. The trimmer pot, R6, is adjusted for minimum distortion in the output signal. You can get rather close by looking at the waveform with a scope, but it’s better to use a distortion analyzer or spectrum analyzer. Spectrum analysis software on a PC is fine, just adjust R6 to minimize the height of the sidebands or use a program that directly displays harmonic distortion.

At 1000 Hz, TrueRTA shows the second harmonic (2000 Hz) down 80 dB (0.01% distortion) with the higher harmonics even lower. AudioTester shows a total harmonic distortion of 0.0105% using the first ten harmonics.

TrueRTA is a spectrum analysis program available from True Audio. Demo versions and a free version (level 1) are available on its website. AudioTester is another spectrum analysis program.

TEST THE CALIBRATOR

The calibrator should be reasonably accurate when built using the 1% resistors and 5% capacitors in the parts list. But as with any other piece of test equipment, it would be satisfying to make some measurements to be sure. I will describe two methods that I used: all the measured values are presented in Table 1. As you can see, the calibrator is very satisfactory.

One method is to use a calibrated phase meter with an accuracy better than the calibrator. I used a Dranetz model 305 (five-digit phase angle display) with a model 305-PA-3007 plug-in.(The Dranetz phase meter is no longer manufactured but used units may be found on eBay or from used electronic instrument dealers.) This plug-in provides automatic operation for input amplitudes of 50 mV RMS to 50 V RMS and frequencies from 2 Hz to 70 kHz. Automatic operation means there are no operating controls. The plug-in scales the input voltage to the mainframe and provides the correct frequency compensation.

Another method is to use a time interval counter to measure the time between an amplitude zero crossing of the reference signal to the amplitude zero crossing of the phase shifted signal. Phase shift can be calculated from the time interval as:

θ = 360τf/1000

where θ is the phase shift in degrees, time delay τ is in milliseconds, and f is the frequency in hertz.

I used a Hewlett-Packard (HP) model 5245L frequency counter with a model 5262A time interval unit plug-in (see Photo 2).

Photo 2: Hewlett-Packard model 5245L frequency counter with a time interval plug-in unit below and my dual zero-crossing detector above.

The 10-MHz counter timebase gives a time resolution of 0.1 ms. The time interval plug-in has trigger-level controls for each channel but they are not calibrated and can’t accurately set the zero crossing with a sine wave input. The smaller “box” above the counter in the photo is a two-channel zero crossing detector. I designed and built this detector to output a pulse whose leading edge coincides in time with the input zero crossing. The counter measures the time between the leading edges of the two pulses: the reference and the phase shifted signal. The detector circuit diagram (see Figure 3) and parts list (see Table 4) are included. I packaged the Detector circuit board with a simple ±5-V regulated power supply in a Wolgram MC-7A enclosure.

Figure 3: The ual zero-crossing detector circuit board

Table 4: The two-channel zero crossing detector's parts list

Looking at one of the detector’s channels in Figure 3, U1 is an input buffer. Resistors R5, R6, and D1 clip the negative-going half of the input sine wave. The comparator circuit (U2) outputs a very short pulse at the input zero crossing. This pulse is “stretched” by the monostable multivibrator in U3 to about 12 ms as set by the time-constant of C1 and R19. Two front panel toggle switches select either the positive-going or negative-going output pulses. The reference and shifted pulses—45° at 10 kHz—are shown in Photo 3.

Photo 3: The digital storage scope display of reference pulse (above) and phase shifted pulse (below) for 45 degrees of shift at 10 kHz. The pulse width is 12 us. The pulse amplitude is 5 V. Pulse baselines are shifted for clarity.

FINDING A PHASE METER

New phase meters are expensive but used models can sometimes be found on eBay or from used electronic test equipment dealers, just try a Google search. In addition to the Dranetz 305 (which I found on eBay), other useful models include:

  • Aerometrics model PM720 phase meter, 5 Hz to 500 kHz, analog meter display. Aerometrics  denies any association with this unit but it is often listed under this name.
  • Hewlett-Packard model 3575A gain-phase meter, 1 Hz to 13 MHz, four-digit display
  • Wavetek model 750 phase meter, 10 Hz to 2 MHz, four-digit display

In addition, you can find application notes and magazine articles that describe how to build your own phase meter. These are usually fairly simple designs. The following appear to be useful: Intersil Application Note AN9637 (This is identical to Design Idea #1890 that was published in the July 4, 1996 issue of EDN); Elliott Sound Products Project 135; and Salvati, M. J., “Phase Meter Profits From Improvements,” Design Idea, Electronic Design magazine, April 11, 1991.

TAILOR THE DESIGN

I sent a copy of this article to Hansen for comments. He agreed that having the oscillator built-in is a good feature. He also commented as follows:

“A problem with my phase meter calibrator design is that the distortion increases with phase shift, and the amplitude drops as well. It might be possible that the zero-crossing detector might be fooled by the higher order distortion harmonics. I’d be interested in what you find out in this regard.”

So, I measured the amplitude drop and distortion at 150°, which should be worst case. I set the 20-Hz variable output to an arbitrary 2.00 V. Keeping the output level control unchanged, I measured what you see in Table 5. This amount of drop seems acceptable.

Table 5: I set the 20-Hz variable output to 2 V, and I kept the output level control unchanges as I measured these.

I used a Hewlett-Packard model 3581A wave analyzer to measure the harmonics. Refer to Table 6. These numbers look acceptable and the zero-crossing detector output at 20 kHz and 150 degrees measures 22 ms on an oscilloscope with a calculated 21.5 ms at the actual frequency of 19.61 kHz.

Table 6: I used a Hewlett-Packard 3581A wave analyzer to measure the harmonics. These numbers are acceptable and the zero-crossing detector output at 20 kHz and 150 degrees measures 22 ms on an oscilloscope with a calculated 21.5 ms at the actual frequency of 19.61 kHz.

I am very satisfied that the calibrator is suitable to troubleshoot and calibrate any phase meter you are likely to find, either new or used. Without overdoing the math, there is enough design information here to allow you to tailor the design to a specific frequency range, keeping in mind the 1000:1 practical frequency range of the Wein-bridge oscillator, without using range switching.

The circuit board designs listed in the parts lists are available in CIRCAD format and are posted on the TDL website. (CIRCAD is a circuit board design program available from Holophase. The boards in the pmcalpcb.zip file were designed with Version 4, a free download of which is available on the Holophase website.) The physical boards are not available.

Ron Tipton lives in Las Cruces, NM. Visit the TDL Technology website for more information about his audio designs and services.

REFERENCE

[1] Jung, W. and Sams, H., Audio IC Op-Amp Applications, 2nd Edition, Sams Publishing, 1978.

Editor’s note: audioXpress, like CircuitCellar.com, is an Elektor International Media publication.