electrodesigns.net

I just started to use Cadsoft Eagle for PCB design some weeks ago. I tried many times in past but somehow left in between due to lack of interest. This time, however, I had a project in hand and the client requirement was to do it in Eagle. I am more used to the Diptrace software and I have reasons to like it. But Eagle is also a very popular PCB design software so I desire of using it.

In Eagle, I specially liked the schematic section as it looks very pretty at the end. It looks surely much more professional than drawn in Diptrace. Most of the hobby projects on Internet with pretty looking schematics you will see are designed in Eagle. Eagle saves its library files in XML format. So one can copy/paste a chuck of XML in the file to create a new footprint. Then can edit the part in the library editor of Eagle. (continue reading…)

PCB designing is an art as well as a science. This is science is for very obvious reasons as electronics is essentially one the most important fields of science. One really needs lots of knowledge of Physics and Electronics to make the components populated on a PCB operate properly. But art plays a role in the look and feel of PCB as well as that of the overall product. Here are a few tips:

The scientific objective:

• The PCB meets design requirements i.e. it actually does for what it is meant for. The DRC is well-defined and checked thoroughly.
• The PCB has shortest possible trace lengths
• Minimum vias and keep-outs
• Minimum number of layers
• Maximum trace with to reduce impedance
• Minimum cross talk and static capacitance. Your DRC should address the possible EMI issues as seen before or as pointed out by the circuit designer.
• The design meets the manufacturing constraints such minimum via size, trace width etc.

The artistic objectives:

• The design meets the mechanical layout requirements. The shape of PCB will fit into the previously decided mechanical casing. There are enough mounting holes with right sizes, placed on right places. (continue reading…)

Open source hardware model

The open source hardware initiative is very interesting because open source, in most of the cases, means something for free as free lunch without any royalty or earning. This is especially interesting because unlike software, where one usually needs to donate time and effort only, here you need to spend considerable money to bring-up a working prototype. There are several things, however, which make this logic practical:

• Today there is bundle of CAD tools which help to simulate hardware before going to fab. Tools like Altium have everything built-in from PCB design to FPGA design to firmware development. There are also some cheap alternatives available like Diptrace or Eagle, though with a subset of functionality. This reduces risk of repetitive prototyping and saves time and money.
• It is possible that initial effort of design be shared to open source community and let somebody else develop the actual hardware. Like some of the designs on this website which have not been tested by building a working prototype but at least they provide a very good ground for starting a product development from something which is valuable and can save lots of time, especially, for the beginners like young electronic engineers.
• The people who use the open source hardware design to build their gadget are very important because they are the ones who actually debug the hardware. These are typically called application engineers who design for service providers. Their feedback is extremely important. Feedback plays very important role which open source community uses in the improvement of a design. The feedback also helps to get new ideas popped up and hence new designs and expansion in open source base.
• Though the hardware may be free, the components like ICs, capacitors, inductors are never free. The people who manufacture these parts must support the open source hardware initiatives. Infact many companies like Coocox, Adafruit, Sparkfun and alike work on the very same idea. Look at the success of Beagle board; it could not have been possible without the support of TI and Digikey. After all who spent probably hundreds of thousands of dollars to bring down the price of Beagle board with so many hefty features? Coocox site tries to depict the actual open source cycle which is a win-win situation for everyone, the chip manufactures, the open source hardware developer, the hobbyist or application engineer, the solution provider company that hires this application engineer and, not to forget, the consumer or end-user.

Why not the application engineer designs directly from the reference design of chip manufacturer? Basically this is what widely happens and it is happening right from the beginning. But open source hardware has some advantages which this conventional model does not offer. Like the product standardization is lot easier and natural as lots of people use the same design or the derived ones from the same hardware. The end product is more likely to bug free as larger number of people build and test and report the bugs. The end product design life-cycle drastically reduces as the application engineer utilizes pre-tested open source hardware components. The hobbyists and young engineers just entering this field greatly benefit by having lots of free stuff available to play with. And sky is the limit.

A few days back I had a chance to work on an IR transmitter/receiver circuit. It was long time since I was engaged in a real circuit design. Most of the previous time I spent was on PCB design and firmware development. The TX part worked like a charm but when the RX part came I really found problems reducing the noise at the input. This noise caused to trigger a false alarm and hence crippled the very purpose of circuit. This led me to look into the very physics of optical sensor. First, I read through the data sheet, which revealed that it is a PIN IR junction diode. Then I further dug into the PIN junction and this recalled me that no biasing voltage is required for it as it generates its own sufficient potential to feed into an amplifier. This reminded me the days of my solid state electronics courses at college and university. This was the subject I really liked because of interesting concepts of semiconductor Physics. I redesigned the whole circuit and it really improved the performance. This made me think that electronics engineers must not forget those Physics lessons back in college and university which we often do due to much brainstorming in the application development.

I wonder there are so many things to learn for an embedded systems developer. This field is rapidly changing. What to learn to keep oneself up to date? Generally this industry is segmented into two portions. One the people who which design advanced DLD systems. These usually work on FPGA based SoC designing. Their tools and skills may include learning an IDE like that of Xilinx. The other guys usually used MCUs and microprocessors (like me). They write the firmware for these device and (usually) design the PCB as well. My focus is on the second type of people as this relates to me better.

Here are the things which I think, are necessary to learn:

1. Hands on experience of at least two MCU families. One on the higher end and the other on the lower end. The higher end may be the ARM architecture. The low-end party should be the one which has tiny-miny MCUs like PIC or 8051 or may be AVR. There are times when you don’t need to use the horse power of ARM, so you need a small companion for that.
2. Good PCB design tool like Diptrace (please forgive if you don’t like it). The Altium designer is a complete design suit not just for a single developer but for a whole team to system design, PCB design, software development and mechanical assembly in a single unified tool. Some people still are in love with older tools like OrCAD. I just prefer Diptrace over OrCAD.
3. At least 2 good RTOSs
4. Embedded Linux. Well, you at least need to have know-how of Embedded MPUs like Cortex-M8/9. Linux seems to me, the future and it would be norm to use Linux in most of the designs.

I have been designing PCBs for around 8 years now. My first experience was with OrCAD 9.x. OrCAD is very powerful tool and widely used in electronics industry. However, somehow, I was a little tired of it as it seemed to me based on old technology. Working in the schematic capture section was not very easy and layout was even difficult. Components library creation could easily lead me to head scratching.

It was some 3 years back. I was really looking for a software which is easy and just does the job. One day I found Diptrace. I learned it very quickly as it was very easy. The user interface was very intuitive and seems to be a really modern interface built-in “this era”. The PCB layout was also very easy, nothing less nothing extra, just every necessary tools available there. The component and footprint library creation was also very easy. Most importantly the cost of software was affordable. It increased my efficiency and reduced the headache of PCB design.

Diptrace has very good import/export options. One of the great option is Route-> Electra/Specctra Interface-> Export Autoroute DSN and Import Autoroute SES. I use this option for complex designs to export my unrouted board to Specctra, which is a great autoroute tool, route in it and import back the SES file into Diptrace for fine tuning. This is one of the great features of Diptrace I really missed in Eagle.

There are, however, a few things which I wish could be improved.

• First is to improve the 3D viewer section. The graphics rendering should be improved to make it look more closer to the real PCB.
• Diptrace currently does not support built-in 3D object creation, which I think, can be a big bonus.
• The third and last is the back annotation. There come times when we need to design the layout first and schematic later. In this case Diptrace is not going to help you much.

All in all, Diptrace is a very good tool for beginners as well as experienced engineers. It saves times, makes PCB design an enjoyable experience and does the job which you expect from a professional designer.

PIC32MX1 and PIC32MX2 are the families of microcontrollers recently introduced by Microchip. These are the cropped down members of PIC32MX3/4/5/6/7 MCUs with lesser power (40MIPS) and lower number of pins. The chips host the same M4K core but with smaller program and data memories.
This all free, open source, very small (1.5″x2.0″) and simple board is an attempt to introduce a ready to use board for easy prototyping. It has a small SOIC-28 PIC32MX1 MCU. SOIC-28 is chosen because it’s easier to solder. The breakout pins can be soldered with standard 0.1″ male/female header to be easily used with veroboard or breadboard. Diptrace source files, a schematic in PDF format and Gerbers are there to download.

Download Files:

In the world of embedded systems designing, there are some main design objectives which are desired by client as well as design house. These are almost universal:

• The design meets the user requirements
• The end-product is low-cost
• Minimum time in development process
• Minimum development cost

To meet these requirements in current competitive environment, the designer must be smart enough. There are several techniques which effectively reduce time and cost of development while managing to fulfill all user requirements.

1. Lock the requirements before starting the design: carefully read-out the exact client requirements. It often happens that non-technical managers take the requirements from client and pass on to the designer. When the designer comes up with the final prototype, he comes to know that client was looking for something else, huh! This has happened to me several times. So if management doesn’t mind, try to interact with the client at the very early stage of project to get the exact picture.
2. Re-use previous designs: Try to reuse what you already have and used in other designs. You better know the pros and cons of your previous designs and try to the best options you have. This may include schematic parts, the components and layout libraries etc.
3. Use available and tested firmware components: Often firmware libraries are available for every vendor of main MCU that you are going to use. If there are no exclusive requirements otherwise, use the available code. It is tested and trusted.
4. At least initial design should be on a tested hardware:  This may be a development kit or a PCB of previous design (Keep the royalty thing in mind haan..). Now a days, demo kits are cheap, tailor-able to some extent and have some great features. If you don’t have kit(s), you can also take help from reference design of the vendor. Every vendor has ref-designs of almost all hot topic in industry. Like Zigbee is a hot field today and every vendor like TI, Atmel, Microchip etc. have their demo boards and firmware stacks to download and develop upon.

There may be other things besides mentioned above, what is your experience?

During my PCB design experience of around seven years, I have learnt some very simple tips to better PCB design irrespective of the EDA tool used. Here are these:

1. Use existing libraries as much as possible. Why re-invent the wheel, it saves lots of time and helps to market product quickly. Search around and you will find libraries like those hosted on this website.
2. When you design your component library, keep the pin layout as per datasheet. Often the component pin-out arrangement is difference in schematic then the real physical layout. Try to design components according to the real layout. Why? Because when you place components around that component, say an IC, you can easily copy that arrangement in your PCB layout. His will help you to better components placement hence better PCB.
3. Arrange your designed library for good reference. By this you will quickly located the parts and hence re-use the library for your future designs.
4. Arrange components on PCB as in your schematic. What components are near, to say an MCU, keep them close in layout as well. Components like coupling capacitors need to be near the ICs, they are meant to be near. This will reduce the physical length of the traces on PCB. This will also help you to understand where different parts of a “system” are located and help while troubleshooting. In-fact placing components in the PCB layout is the key to PCB design. This is the “soul” of PCB design skill.
5. Use hand-routing as much as possible. CADs are not more intelligent than human mind. Unless it is very difficult, like multi-layer PCB with dozens of ICs, use hand-routing. It will be a much cleaner and neat design at the end.
6. Keep housing of final product in mind when designing layout. Often we design PCB when no housing is decided. But at the end of the day they need to be housed. So keep mounting holes there, and better choose some readily available and simple housing and design around that. It is possible that later-on the same housing gets approved and the re-designing is avoided.

You may be interested to further read: PCB designing, science and art.

There a few things which every technical business owner or a manager must know in order to run his/her business efficiently. Following tips are written with intention to guide small start-ups with limited resources.

1. You must know the available resources and their capabilities. Resources mean the personal as well the equipment, required services and most importantly the time.
2. The constraints. Every manager must know the constraints under which his/her work-force is working. For example a small embedded design company must not expect its team of merely two people to develop in-house GHz processor board which contains several BGA packages to be soldered. Ofcourse it needs special technology and for that some outsourcing is required. The manager must understand the technical capabilities and his manpower’s scope of work before indulging into some serious business. The project must be in-lined with company’s technical capabilities.
3. Choose the target market which can earn you more with less investment. This is crucial because the difference between investment and return (ROI)  is the profit.
4. Choose the market carefully which can be captured under small resources. For example trying to design something which is useful but cheaper to develop like electronic sensors etc.
5. Try to remain within your domain of expertise. This will ease up the job.
6. Try to re-use the work already done. This is extremely important w.r.t. timely marketability and catching the window of opportunity. However, this is not a very simple thing to do as it seems. The hands-on technical experience of several years is required to master this “art of science”.
7. Try to sell services as much as possible instead of products because services cost almost nothing and earn handsome while there is always a risk in the success of product after all the investment done. This can be untrue for large companies but for small start-ups, it is a wise thing to do.
8. Convert services into products. The experience during the services period is the greatest asset of a technical person. This a very slow and careful process.  It is like joining small pieces to bring up something useful. So keep your data protected and well organized. This will greatly help to reuse it.
9. Technical documentation has always been a weak point of, especially, the engineers. Technical documentation help to understand something which human memory can not keep-up. Whether it is C code, a PCB design or a CAD library, everything must be “written-in” why it is, how it works and its dependencies.