No longer anything to fear from wireless charging

No longer anything to fear from wireless charging

Loek Janssen - Project Engineer, Sensor Systems

By: Loek Janssen
Project Engineer, Sensor Systems

1st March 2017

Home » Product Development

Recently, I have been working on a project where we really wanted wireless charging for the device. Being only in the prototype phase, it was considered more a “nice to have” than a complete requirement, with several engineers putting it into the “requires a lot of work” category. However, after doing some research, I decided that the standards (such as Qi and PMA) were developed enough, with excellent support, to give it a try.

So how does it work?
Most wireless charging is based around the idea of inductive charging, i.e. inducing a current using a magnetic field. In many ways, the idea is similar to that used in transformer, current in one coil, induces current in a second coil. However, transformers share the same magnetic core and when the two coils are separated by air the method is horrendously inefficient with most of the power wasted instead.

When the two coils are, in combination with capacitors, used to form resonant circuits, the efficiency over air increases dramatically. The range is only a few centimetres but reasonable amounts of power can be safely and easily transferred.

As only alternating signals can be used in the inductive coupling, any receiver then needs to rectify the signal, filter and provide the output to power the system or charge a local battery. A control loop usually exists in the receiver to limit the overall current drawn. Additionally, the transmitter and receiver can also communicate by gently modulating the signal, while only a small amount of data can be transferred; it is more than enough to allow basic control messages between the two devices.

wireless

Standards
Several standards now exist using this idea of inductive charging, using different frequencies, voltages and modulations for power transfer and communication. While none have yet won out, multiple mature ICs, which handle much of the heavy lifting of the receiver (and transmitter) side, happily exist for the QI, PMA and Airfuel standards. The Qi standard, as an example, uses a frequency of 100-200 kHz and ASK (amplitude shift key) modulation for communication.

Getting it into a product
While the datasheets require some maths to determine the correct component values, it is fairly straightforward, and once a suitable charging coil has been chosen, I quickly got around to prototyping the design. With the right evaluation board, I was able to get a nice receiver working over short distances, allowing me to test various coils to see which would work best through different plastic materials. I had chosen a QI receiver IC, as the QI protocol seemed very mature, and looking ahead to the future, plans were in place for extensive improvements from the recently released V1.2 standard.

Now the prototyping was done, I designed the IC, components and charging coil into our system and once the final PCB was back, tested the wireless charging system. As expected, power could be drawn through the system, with the IC internally controlling the voltage to produce a nice 5V DC signal to power to the system.

Despite not having any personal experience in the area, the wealth of information and useful components allowed me to quickly prototype and design a circuit with wireless charging. It is a seriously useful component of the system, allowing devices to be sealed and protected, while still being easy to recharge.

Regularly, people are surprised by the addition of successful wireless charging to the system but I think this is a hangover from the past; when implementing such a design was difficult, complex and required a lot of effort to get working. Now we have the addition of excellent standards and mature ICs in mass production, it is definitely time to stop being afraid of wireless charging.

Save

Further Reading

50 Years in Engineering

Fifty Years in Engineering

Stewart Da'Silva - Senior Designer, Product Design

By: Stewart Da’Silva
Senior Designer, Product Design

22nd February 2017

Home » Product Development

pcb_layouttoolsIn 1966 whilst serving my apprenticeship as a mechanical design draughtsman, I was assisting a senior draughtsman on a design that required a simple power supply. This product was going to be manufactured in medium volume and he suggested that maybe I would like to investigate the possibility of using a Printed Circuit Board to connect it up instead of using wire to make the connections. This was my first introduction to PCBs.

He had really set me a challenge as no-one in my company had used one before. Anyway, suffice to say that that very first layout of mine was constructed using an ink pen, rule and compass using Indian black ink on white Bristol board. To increase accuracy of the finished PCB, it was drawn 4:1 scale. As is probably obvious to the reader, mistakes whilst drawing the PCB usually necessitated starting from scratch again. The next stage was arranging for an industrial photographer to generate a 1:1 positive film from the 4:1 artwork that could be used by a printed circuit board manufacturer to fabricate the PCB.

I finished my apprenticeship in 1968 and not long afterwards started work as a mechanical designer in a ’Contract Office.’ This was essentially a design house offering design capabilities to companies that did not have the necessary skills or had to outsource projects due to a high volume of work.

taped-artwork_black-tapeIt was here that I learnt the basis of PCB design. Things had moved on, although designs were still only single or double sided. Instead of ink on Bristol board, the initial design was drawn, again at a scale of either 2:1 or 4:1, on a stable semi-transparent plastic film that was placed over a similar transparent film with a 0.1inch matrix printed on it that was fixed to an A0 drawing board. This grid was used as a guide for the PCB layout.

If the PCB design was double sided, the usual convention used was blue pencil for the component side and red pencil for the solder side. Once completed and checked this pencil layout was flipped over and secured over another grid that was in turn attached to the surface of an A0 size light box. A translucent film was positioned over.

lightbox1Using pre-cut adhesive backed-tapes and pads of various sizes and following the red colour of the layout that was under this sheet as a guide, the solder side of the PCB took form as the designer built up the artwork. When the solder side artwork was complete it was removed from the light box together with the pencil layout. The artwork was flipped over and secured once again to the light box, another plastic sheet was placed over this and again, using the pre-cut pads, the designer aligned these with the pads on the completed solder side artwork. Once all the pads were positioned, the solder side artwork was removed and this ‘pads only’ component side was again placed over the now turned over pencil layout and the blue colour followed to tape up the component side. The two sides of the finished and checked artworks were then sent to an industrial photographer who generated a 1:1 artwork from the originals.

red-blue-artworkThe next step that the industry took was to use only one piece of stable plastic sheet instead of two. The pre-cut black pads were still used but instead of black tapes, transparent blue and red tapes were used and were placed on opposite sides of the sheet. The industrial photographer would then attach filters such that only the red or the blue traces appeared as black when he created the 1:1 artworks. This may seem a small step but it did mean that alignment of both sides of the PCB artworks was guaranteed as exactly the same pads were used.

Read part 2 of Stuart’s blog here.

Save

Save

Save

Save

Save

Further Reading

Save

Save

Save

Save

Save

Save

Not every snake is the same

Not every snake is the same…

Glenn Wilkinson - Senior Consultant, Sensor Systems

By: Glenn Wilkinson
Senior Consultant, Sensor Systems

15th February 2017

Home » Product Development

They live behind cupboards, skulk under the bed, lurk at the back of desks and hide in car glove boxes. They lay in wait across the world. Coiled and ready to unleash their potentially deadly power in an instant. And we wouldn’t have it any other way.

USBPhoneThe now ubiquitous USB charger lead has increasingly littered our lives for nigh on 10 years now. Adopted by portable device manufacturers near unanimously as soon as their handheld marvels had a need for a power and wired data connection. So given how common the USB charger is – it’s surprisingly misunderstood.

Stay with me now – it might stop you getting bitten…

Start looking into the subject and you’ll find forums, reviews and help boards littered with common complaints, comments and questions (sad face emoji’s removed to avoid repetition);

“I bought a 2 amp charger, but my phone doesn’t charge any faster.”
“I left it connected overnight and it was still flat in the morning.”
“Will this phone charger work on my tablet?”
“My Galaxy charges faster on my PC than with this supply.”
“I get a ‘charger unrecognised’ message when I connect it.”

So what’s going on?

First off, your charger doesn’t determine how much power it is going to supply – your device decides what it is going to consume. It’s a self-protection mechanism. The batteries you are charging need a limited charge current to prevent overheating and maximise cell lifetime, so your device contains a charge circuit which implements this limit. Go and buy the beefiest charger you want, but it will only charge your handheld quicker if the device allows it.

So if the charging smarts are in your device, why can it perform differently with some chargers, and not at all with others, even though they have sufficient power capability?

We could start digging into history here, pointing fingers and quoting standards – but let’s not go there. Simplistically put; somewhere between standards being insufficient or optional, and manufacturers implementing in a manner to suit their own needs, we end up with the ‘mish-mash’ of solutions we have now amalgamated over time.

USBslotsStrictly a USB-compliant host port can support a number of power modes (USB2.0 high/low, BC, PD for those curious). The host USB port and device need to indicate to each other across the USB data lines and agree the most suitable mode to operate in.

Some devices ignore any form of convention completely and just draw a fixed current. This is relatively common in legacy smartphones and is usually around 500mA, not unintentionally the same as the original USB2.0 high power spec.

Others create their own hybrid language solutions, predominantly by bias voltages on data lines to indicate whether faster charge modes beyond USB low or high power current limits can be used.

It has to be said that things are getting better with the widespread adoption of the USB battery charging spec (BC1.2), and particularly the simplicity of the direct charge port (DCP). But the delay in it being published means there are still a huge number of devices out there with their own niche charging characteristics and languages.

Why am I telling you all this? Because it’s another one of those hidden little hurdles that us engineers have to overcome. Plextek recently created an award winning commercial device featuring USB charging ports. The brief required those ports to support charging of any common smart device at a rate equal to the manufacturer’s charger. This means speaking every language, and knowing which to use at any particular time.

Of course, the main fun of this is in the testing. After a good day swamped with more tech than a Gadget Show giveaway, and kind co-operation of a very nervous man at PC World (after some negotiation) – job done, a truly multilingual solution.

So hopefully you may have gained some small appreciation of the engineering behind something the vast majority of people take for granted – and there may just be fewer sad faces on those internet message boards.

Save

Save

Save

Save

Further Reading

Save

Save

Save

Why Circles Are Better Than Squares

Why Circles Are Better Than Squares – An Introduction to Geometric Tolerancing

By: Polly Britton
Project Engineer, Product Design

8th February 2017

Home » Product Development

Tolerances save on costs
A tolerance on a technical drawing describes the boundary between what is acceptable in a part and what is not.  When a part is made with even one measurement out of tolerance it has to be reworked or scrapped, at cost to the manufacturer. Additionally, an accurate manufacturing process is usually more expensive than the less accurate alternative. Therefore, great savings can be made to component prices by designing tolerances to be as lenient as they can be without impairing the function of a part. One easy way to do this is using circles instead of squares in your tolerances, as illustrated by the following example:

The square way
A simple way to define a tolerance for a hole’s centre position looks like this:

Squares6.1
There is nothing incorrect about this tolerance, but it is probably not a good description of what deviation is allowable. Consider the shape of the tolerance zone for the hole’s centre, shown in orange:


The problem with a square tolerance zone is illustrated above. The green and red dots represent two possible centre positions. Although both dots are an equal distance (√2 or 1.41) from the “true” centre position, a hole centred on the green dot would pass a quality inspection and the red dot would fail. For most applications the red and green centres are equally acceptable, so how can you communicate this in your technical drawing?

The circle way
This is a drawing of the same hole, with a geometric tolerance.

Squares6.4
The two “10” dimensions have become “true dimensions”, as depicted by the boxes around the numbers, and the two datum edges they are measured from have been marked “A” and “B”. The annotation pointing at the circle means something like “The position of the centre can lie anywhere within a diameter 2.83 (that’s 2 x √2) circle of the true centre, with respect to A and B.” This defines the tolerance zone shown below, and allows the centre of the hole to be at both the red and green positions.

Squares6.5
Which is better?
Using geometric tolerancing to describe a circular tolerance zone may not allow you access to cheaper manufacturing methods, but it may reduce the number of scrapped and re-worked parts without having to loosen the tolerance at all. In fact, a circular tolerance zone will be 57.1% bigger in area than a square zone would have been. This additional allowance could make quite an impact when added up over many holes in one part, when just one non-conformance means scrappage.

Squares6.62

I would not insist that geometric tolerancing is essential for every application. There are many cases where it would just be a waste of time to define reference datums (or “data” if you prefer) and struggle with your CAD package’s tolerance tools, especially if the intended audience of your drawing is not trained in geometric tolerances. Even if you never have and never will use a geometric tolerance, I hope this has helped you think about what tolerances can say about your part.

Further Reading

Save

Sensor Systems – Getting to Market… Quickly!

Chris Roff - Head of Smart Sensors

By: Chris Roff
Head of Smart Sensors

1st February 2017

Home » Product Development

As an engineer, when somebody comes to see you with an exciting technical idea, you are always keen to listen. But, when the first meeting is on Halloween, and the business case requires complex prototypes deployed before Christmas you’re allowed to worry that it could be a little scary! But, with careful planning, lots of experience and a team who know what’s what, it’s possible to deliver without getting spooked.

This particular project was in the smart sensors area and required an array of about 20 different environmental sensors (some of which needed creation from scratch), a pair of backhaul radio options (cellular and ISM) as well as a secure backend interface. Security was extremely important and the device had to operate in a harsh environment. Oh – and it all had to run from a battery with tough lifetime specs and sit inside a unique custom enclosure…

We responded to this design and timescale challenge by bringing the customer’s technology leads into our offices for a week-long workshop. The week’s agenda started at high-level aims and finished with electronic schematics being sketched out – progress!

The next phase of the programme saw high-risk sensors being prototyped in the laboratory whilst the electronic schematics were being fine-tuned. Meanwhile Printed Circuit Board (PCB) layout engineers were creating component footprints and doing everything possible to be ready to route the schematic as soon as it was released. Having a good view of the components that exist in the sensor market is vital to allow rapid down-selection of appropriate devices (from the often intimidating range of options).

At the same time, software engineers were devising the message protocol for the radio communications and working on low-level drivers for the chosen components. Embedded device software suitable for battery powered sensor systems requires working with limited memory space and real pressure to control current consumption.

Another parallel effort was looking at the mechanics, designing enclosures, IP68 seals, cable assemblies and the like. Modern 3D printing and soft tooling facilities allow large complex designs to be realised within even the most aggressive schedules. And watching the machines in action is a sight to behold.

Once the electronic hardware is nearly designed in CAD, the supply chain must jump into action, securing components “kitting up” and arranging access to pick and place assembly lines. Supply chain guys are an engineer’s best friend when time is tight – they always manage to locate that rare component just when you think the lead time is unworkable.

After assembly, when the product returns from the factory, it can be a tense time. Everyone is proud to see their hard work made real, but equally tense that one mistake overlooked during design could render the boards useless. Usually there’s a deep sigh of relief when the first LEDs flash and no smoke appears!

The next stage is to bring the board up, test hardware functionalities and software control. Typically a plan for these activities is made beforehand and a burndown list is worked through to ensure nothing is missed. Ticking off working functionalities is very satisfying for engineers but the thrill is perhaps greatest for the project managers!

Inevitably along the way some modifications will be required when bits of circuitry don’t work quite as they did in the simulator or on paper. The trend for smaller and smaller electronic components means that this work requires sharp eyes and a steady hand with soldering irons or hot air guns. When you have a deadline, having technicians or “wire men” with nerves of steel on your side is invaluable. Having patient ones is a bonus too!

Once the hardware is functioning as desired and the software has been developed to beta level you can begin system tests to verify that the product will do what it needs to in the field. The key here is to plan a suite of test conditions that cover both normal operation and all likely corner cases. At this stage modifications may again be required to arrive at the stable end product.

Once the design passes test it’s ready for showtime. Final assembly, shipping and installation of prototype units before Christmas? No problem for this team! Now there’s just time to monitor that sensor data over another portion of Christmas pudding…

Save

Save

Save

Save

Save

Further Reading