Choosing the right electronic components for your project can make a big difference in its performance, reliability, and cost. But how do you decide which ones to use among the many options available? In this article, you will learn some of the best practices for selecting electronic components based on your design goals, specifications, and budget.
Before you start looking for components, you need to have a clear idea of what your project needs. What are the functional, electrical, mechanical, and environmental requirements of your circuit? What are the constraints and trade-offs that you have to consider? For example, you may have to balance power consumption, speed, size, and temperature. You also need to know the standards and regulations that apply to your project, such as safety, compatibility, and certification.
Sometimes it is helpful to think outside the box. I was working on a medical device once that used an MSP430 microcontroller and supporting analog electronics and switches to deliver a biphasic trapezoidal waveform for electrical stimulation. There was a lot of software involved, and because it was a medical device, a lot of software documentation required. I thought to myself, there's got to be an easier way to do this, and was able to perform the same waveform delivery using two tiny mixed-signal ICs costing < 0.50 each. Best of all, the system was 'programmed' using a schematic-like description, so there was no 'code' involved.
The business end is as important — actually more important — than any technical requirement. What good is specifying the perfect part of you can’t get it? You need to understand how your organization purchases components. If your company only purchases through its preferred distributors, that limits your options. You need to understand your estimate annual usage (EAU), when you will start buying production quantities FCS), in what quantity increment will you order (LOT), if you expect usage to grow over time (FOR), and how long you expect to use the part (LIF). Will you need bonded inventory? If so, where held? Will you contract for deliveries? Do you need surge capacity? Check the supplier’s end-of-life (EOL) practices too.
The best tool is a parametric selector to filter the huge range of candidate components down to a manageable shortlist. I got so fed up with the restrictions of the parametric selector tools available from manufacturers and distributors I wrote my own. It's taken a long time to build up tables of parts, but the earlier you start the sooner you get the payback. Some years down the road the chosen part will go obsolete, so having a fallback plan early will also pay off.
For projects with low quantity devices to sell, components with rich functionality are preferable, because they add flexibility to design process. It reduce time for designing.
Once you have defined your requirements, you can research the different types of components that can meet them. You can use online databases, catalogs, datasheets, and reviews to compare the features, specifications, and prices of various components. You can also look for reference designs, application notes, and tutorials that can help you understand how to use them. You should also consider the availability, quality, and support of the components, especially if you are planning to scale up your production.
This step can be and usually is very time consuming. Finding documentation (such as datasheets) and comparing component functions, parameters and performance curves can only be done within a project for a limited amount of options, even though component manufacturers make a lot of content available. To get it done, consider previous experiences or well-known components on the function you are trying to implement. A common example is the usage of 7805 voltage regulators for 5V supply. There are new engineering tools for electronics that can suggest and generate designs based on requirements and the best components and their topologies.
After you have narrowed down your options, you should evaluate and test them to see how they perform in your circuit. You can use simulation tools, prototyping boards, and testing equipment to measure the parameters, functionality, and behavior of the components. You should also check for compatibility, interference, and stability issues that may arise when you combine them with other components. You should also consider the ease of assembly, soldering, and maintenance of the components.
Simulations are great, but they are only as good as the models they are using. Unfortunately, designing a good model can take a long time. If possible, testing the real component, be it on an evaluation board or on a small prototype is always more revealing.
When you have selected the components that work best for your project, you can optimize your design to improve its efficiency, reliability, and cost. You can look for ways to reduce the number, size, and power consumption of the components, as well as to simplify the layout and wiring of the circuit. You can also look for alternatives or substitutes that can offer better performance or lower price. You should also review your design for errors, bugs, and potential failures that may affect the components.
The final step in selecting electronic components is to document your decisions and the reasons behind them. You should keep a record of the components you have chosen, their specifications, sources, prices, and suppliers. You should also document the testing results, design changes, and optimization steps that you have done. This will help you to track, verify, and troubleshoot your project, as well as to share, reuse, and improve it in the future.
For some devices, such as transistors and opamps, document precisely *what* parameters were key to the selection. There are dozens of parameters, only a few of which are usually applicable in a given circuit. If a part needs to be replaced later (obsolescence, availability) then the person given the task will know what to look for.
Documenting circuits can also be very time-consuming. Opt for the most accessible tools for your use-case. For schematics, for example, there is a range going from free and open-source up until very complex and expensive. You can also leverage a few tools that can generate the documentation automatically for you.
What I prefer to do is select the component(s) based on the requirements/ parameters and then based on experience, select several substitutes (not just the best one preferred) for the BOM, for each component number- IC 101, for example, based on availability Let's say a dual op-amp. The first one you choose on Digikey has a 36 week delivery time or the second manufacturer on Mouser is also too long. Choice 2 or 3- a 5532, or TLO72. etc.,etc. say would give the purchasing people several choices for each component on the BOM. This requires building your BOM as IC101a, IC101b, IC 101c. All based on availability and time frame to manufacture. Even accounting staff will get this reality aspect, because it's "Time to Manufacture", isn't it?
The field of electronic engineering is constantly evolving, and so are the components that are available. You should keep yourself updated on the latest trends, developments, and innovations in the industry. You should also follow the feedback, reviews, and recommendations of other engineers and experts who have used the components that you have selected. This will help you to learn from their experiences, discover new opportunities, and enhance your skills.
In my experience, the choice of a vendor tends to be a sticky decision. Many times, it comes down to how prompt the vendor is in sharing information and samples, how easy it is to get CAD drawings and models, etc. However, it is important to spend the time and research different vendors to find the best component choices for your design.
At MIT/KAVLI/LIGO, I designed a circuit that no place else would pay for, but one on a billion dollar observatory was insignificant, and had 10 times the dynamic range. I also spent more of my man hours fixing the ground loops, which made it half as noisy. All they've been able to do since was to quantum entangle pairs of photons, making the instrument 1dB quieter.
Check a would-be supplier’s technical and sales support, what their Product Change Notice (PCN), end-of-life (EOL) notice, and Lifetime Buy (LTB) policies are.