Do designers care about supply chain issues?

If they don’t, they’d better get onboard very quickly. OEM designers need to pay attention to supply chain issues including availability and pricing, particularly now in the midst of an ongoing shortage of passive and discrete components. Some components, including capacitors, MOSFETs, and diodes, are at historic high lead times, and some parts are on allocation, according to an EPSNews report.

Some of you may remember the tantalum capacitor shortage of 2000/2001 primarily caused by a materials shortage of tantalum ore, which led to some double ordering in the industry. At that time, some OEM designers had to replace tantalum capacitors with multilayer ceramic capacitors where possible, and that caused some design hiccups. The same thing could happen now but in reverse.

Do designers have to get more creative in their parts selection during extended component shortages? Do they need to work more closely with the purchasing team? A few articles in this September’s issue made me think about how designers deal with long-term component shortages. For example, the article on silicon IPDs indicates that these devices are great for applications that have extreme space constraints and they eliminate the need for discrete capacitors, resistors, and inductors — all parts in short supply.

So can designers move to IPDs en masse? I doubt it for a couple of reasons: Doing so requires additional design resources and time to design, especially if they are custom devices, and will require a change in board layouts. In addition, IPDs are not suitable for all applications. But they could help in some niche applications.

Kim Eilert, principal engineer and IPD product line manager at ON Semiconductor, told me that designers can consider IPDs as replacements for some capacitors in short supply, but they have to consider their operating voltage. “Some values of capacitance are too large to fit well into an IPD. Folks looking for high values for powerline decoupling, for example, are unlikely to be happy with what they can get out of an IPD, but anyone using values for filters will find those same values in IPDs. You can use IPDs, but it will be for a niche market.”

What about power topologies? Analog Devices Inc. noted a couple of options for designs using buck- and boost-mode operation in “Buck-boost DC/DC converters: MVP for power supply design .” These include using multiple power converters (which have multiple drawbacks) or a SEPIC that requires two inductors and a high-current coupling capacitor. Besides adding to the bill of materials, it adds further design complexity. Then there is the buck-boost DC/DC converter that replaces two ICs and uses only one inductor.

With such widespread shortages across so many passive components, does it make sense for designers to start considering alternative technologies like IPDs or other power topologies and architectures that can either reduce or eliminate some passive components? What kind of trade-offs can designers live with to get their product designed, manufactured, and shipped out the door on time?

I’d be interested in getting your thoughts and ideas.