Buy One, Get One Free

Many components come in both single-device and multi-device packages. The 555 timer is available as both a single timer in an 8 pin IC and as the 556, with two timers in a 14 pin IC.


Frequently, these both cost the same price, or so close to it that the difference is negligible. The same is true for op-amps, comparators, and many other devices.

So, if you can afford the extra space on your PCB, stocking the multiple device versions of the parts is more economical, even if you often only use one of the devices. Each time you use the second device, you're getting it practically for free.

Also, with the various uses that most devices can be put to (op-amps can take the place of many logic gates when wired properly, for example), they give you extra "wiggle room" in your circuits for growth or for changes to the design.

Old-school Components Keep It Cheap

Keep costs down by using simple non-digital, non-IC components for simple jobs. Instead of using a microcontroller that costs a couple of dollars to blink an LED, use an astable built out of a couple of transistors, caps, and resistors. The cost is pennies rather than dollars. The same goes for invertors (a couple of resistors and a transistor), one-shots, simple encoding/decoding, etc.


Most of the circuits you need are already online or available in various electronic circuit cookbooks.

This also saves I/O lines in circuits where you have microcontrollers doing other tasks. If you use bulk buys for your components, you can cut costs dramatically by using simple analog circuits in place of microcontrollers, even the small cheap ones.

Microcontroller Project Cost Control: Develop on a Big One, Implement on a Small One

Save money on your Microcontrollers. Just as you can save money by replacing expensive standard logic with a microcontroller, you can save money on microcontrollers by using smaller, less expensive versions in your projects.

But if you pick a chip that's too small--too little memory, not enough I/O--for your project, you can end up stopped midway in, held up because you didn't start with enough to finish the job. How do you avoid this?

The trick is to develop for a larger version of your favorite microcontroller first. Then, once you've got your application working, you can see exactly what you need and move the app to a smaller, less expensive IC that has enough memory, I/O, and the right peripherals for what you're doing.

For example, I often develop small applications on the ATMega8 AVR microprocessor. It has 8K of flash, 1K of SRAM, and 512 bytes of EEPROM as well as a good selection of peripherals like UART, ADCs, and timers.

This gives me a lot of room for a small application. Then, when the app is done, I see what I need and put it on the appropriate ATTiny controller. The cost of an ATTiny can cut the cost of the chip to a fraction of the cost of an ATMEGA. The shift is quick and easy, I usually end up with a better version of the code when I move it since it gives me a chance to make some changes I thought of after writing the app on the big chip.

Then the big chip gets wiped, ready for the next app to be developed.