Component Selection: How Do You Do It?

When it comes to selecting a component for your new design, this can really be an overwhelming task. You know what you want and what you want it to do, but need to find the right components to get to the ultimate goal. You want the satisfaction of watching your design flow down the production line and out the door for the world to see.

What do you do when you are testing, evaluating, and working on the bill of materials (BOM)? How do you choose the components you need? Do you just copy and paste from previous designs and go with the parts you have used for years? Or maybe select three or four competitive parts and create a pros-and-cons part matrix? You see that you need an op amp, voltage reference, ADC, and a boost converter. But you know that there are more suppliers out there making functionally equivalent parts.

There are advantages and disadvantages to most any option of component selection. The veteran engineers may use a particular part just because they have used it in every previous design since day one and are simply comfortable with it and refuse to look elsewhere. However, that eliminates the chance for finding something that may be better performing, more efficient, or lower cost. It's great to go with what you know, but at the same time you might be missing out on the latest technology.

Are you afraid to go with a newer part, perhaps because it has had less market exposure? Do you just compare data sheets and pick the part because the numbers look good on paper? I think many of us know that datasheets are not always good at taking the actual part and testing it in a way that shows the real, true performance — or shows how it will perform in our design.

Maybe you choose the component supplier that is No. 1 in the market. The assumption is that because it's No. 1, it must be the best. Maybe you narrowed down to two suppliers because they tested and performed nearly the same and have similar costs. Do you take time to listen and meet with suppliers or supplier reps and see what they have to offer?

How do you select that particular component and finalize that BOM? How did the parts end up on that schematic, on the BOM, and in the PCB flowing down the production line? Are you really picking the best components for your design? Are you opening up your eyes, ears, and mind to what else is out there? What is your main driver of component selection? Is it quality, cost, delivery, function, package, history of the part, ease of use, or performance? Let us know what you go through to pick the parts. And let us know about the consequences.

Related posts:

7 comments on “Component Selection: How Do You Do It?

  1. RichQ
    August 27, 2013

    Starting out as a designer, I was told to discount spec sheet numbers by a factor of 2 or 0.5, whichever was worse. And to discount sales claims by 10 or 0.1.

    Spec sheets are notoriously difficult to obtain application-specific information from, so there is seems to always be a need to choose a few candidates and run detailed tests or create prototypes to be sure you have an item that will work for you.

    I'm not sure how accurate the models supplied to EDA tools by parts vendors are, but if the models are good, simulation can help in choosing.

  2. Davidled
    August 27, 2013

     Price is a major factor to decide which component is selected, unless there is a difference among components. Purchasing department will get a quote from vendor and decide component.  In the engineer viewpoint, performance might be focused to meet the requirement. In some case, there is a compromise between cost and performance in the system integration.

  3. Netcrawl
    August 27, 2013

    @Daej you're right companies ideally want a technology that is low cost and reliable, purchasing departmet get quote from various vendors, and decide what to acquire. Companies decide based on products performance and pick the best. Differentiation will always be a key factor in this game.   

  4. eafpres
    August 27, 2013

    Hi Jason–from my experience managing design teams, parts that were readily available from distributors next day had a much higher chance of getting spec'd onto a BOM than others.  For some highly specialized parts, this of course isn't true, but for a large % of BOM, what went into the prototype got onto the first production release if it worked.  We left it up to supply chain to negotiate good pricing for production; our emphasis was on fast completion of design and meeting performance criteria.

  5. RedDerek
    August 28, 2013

    I look at price for most designs. But I cannot say I have started a design with the intent to reuse a past part.

    For commodity items such as resistors and capacitors, it is mostly about price and basic specs. Derating is 50% or just a size that is easy to prototype.

    For ICs, there are a handful of opamps and comparators I use, but I do double check the specs for my application. More specialized ASIC needs put me on the hunt for the ideal part. I have found myself spending up to a week looking for that perfect part to fit the bill. Here I start with key manufacturers and then just go through looking at data sheet after data sheet until I find a part that fits. Sometimes it could end up that no one has what I need and thus I have to start going discrete.

  6. BillWM
    August 28, 2013

    Ability to meet demanding requirements, availability, price —

  7. jkvasan
    August 29, 2013

    “our emphasis was on fast completion of design and meeting performance criteria.”


    I fully agree. Once the design is completed in time and performance met, the impact of BOM could become insignificant at the production stage where the volumes are higher. To worry about small volume prices would only reduce the design team's efficiency.

    Of course, this applies to high volume products.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.