It's the same with circuits. If you need a compact, affordable circuit tomorrow that performs a common function, say, voltage doubling, then you'll likely find a dozen or more affordable ASICs (application specific integrated circuits) that will satisfy your requirements — with no or few external components.
However, if your goal is to learn how a particular circuit works and how it can be modified to suit your particular needs, then you should consider a circuit made of discrete components, transistors, and generic ICs. The end product will likely cost more in terms of component prices, printed circuit board real estate, and your time. However, you'll have an opportunity to learn more.
Of course, you have to make the most of this. If you simply insert the parts into a breadboard, solder everything into place, and then apply power, you'll waste the moment. For example, let's say you have a voltage doubler circuit based on the LM555 timer chip. You can change the frequency and duty cycle and external component values to assess how the changes affect output voltage and current.
The process vs. product tension isn't the only reason to consider single chip solutions over discrete component circuit design and construction. Some of the newest chips have exceptional specifications: higher efficiency, lower voltage requirements, less potential for interference, better thermal regulation, lower noise floor, etc. Sometimes it comes down to weight. For example, I've been working with quadcopter controllers, and every extra ounce of circuitry means about five minutes of less flying time, less maneuverability, and less space for something else.
Often, it's a matter of focus. Let's say you've built a LIDAR that requires a voltage doubler. If your focus is to learn how to apply LIDAR to robot navigation, then don't waste your time designing or building a low level supply circuit. Pick up a voltage doubler chip and get to work on the LIDAR proper.
I've found the major hurdle in using an ASIC over a handful of discrete components is determining if my wish-chip exists. My first go-to resource is Digi-Key. They maintain a good (albeit incomplete) catalog of special purpose chips at reasonable prices. My second stop is Mouser — another online supply company. Prices are usually a tad higher than Digi-Key, but they also handle a slightly different IC product line. Failing to find something in my top two list, I turn to the chip suppliers directly, such as TI and National Instruments. Most of these companies provide powerful search engines indexed to their products.
In closing, there's nothing inherently wrong with using discrete component circuits in your projects. It simply depends on what's best for your budget, the components you have on hand, your experience with new and old ICs, and whether the circuit is for your personal use or destined for the masses. However, make an informed choice. Check out which ICs or components best fit your needs. If, after a first exploration, you can't find the chip you're looking for, keep looking. New ASICs are announced daily. SV
Posted by Michael Kaudze on 02/21 at 02:54 PM