When I first started managing procurement for an electronics manufacturer, I thought a bill of materials was just a list of prices. Lowest quote wins. That's how you buy office chairs. It is not how you buy a capacitor that might end up inside a blood pressure cuff.
I was wrong. Took me about two years and one expensive rework order to figure that out.
You think it's a price problem. It's not.
Every costing review I've sat through starts the same way. Finance points to a line item. 'The Kemet T520 is 80 cents. The alternative is 28 cents. Why are we paying three times more?'
If you're in that meeting, you know how hard it is to explain ESR, ripple current, and reliability data to someone whose job is to make the number smaller. You walk them through the engineering. They nod. Then the email arrives with a quote for a cheaper generic cap.
This is the surface problem. It feels like a cost problem. It's actually a category problem.
The deeper problem: comparing the wrong 'same thing'
Here's the thing: we use the same word for different things without realizing it. At my company, someone once asked about 'switches vs cisco' because they heard we were looking at switches for a new product line. But a Cisco switch is a network switch. An electrical switch in a power circuit is a different animal. The word is the same. The failure modes are not.
That kind of category confusion is exactly what happens when you compare passive components by price alone. A Kemet T520 is a polymer tantalum capacitor. A generic electrolytic cap is also a capacitor. Both smooth voltage. One is expected to handle higher ripple, tolerate harsher conditions, and keep stable capacitance over a decade. The other is a commodity part that might work fine in a desk lamp.
I used to think I could tell the difference from a datasheet. Not anymore. The difference shows up in the field, after the design is frozen and the boards are built.
What a blood pressure cuff taught me about component trust
Nothing exposes this faster than a medical device. Consider a blood pressure cuff. Not the clinic version with a nurse standing next to it—the home unit someone orders online. It has a pump, a valve, a sensor, a display, and a small power supply. That power supply has a capacitor that smooths the voltage. If that capacitor fails, the display goes blank in the middle of a reading.
'This brand is unreliable.' That's what the user thinks.
The same thing applies to a popular model like the Duraforce Pro 2. It's a well-regarded home blood pressure monitor. People trust it. That trust doesn't live in the firmware or the cuff—it lives in the components that keep the device alive for years.
Maybe the Duraforce Pro 2 doesn't use Kemet parts. I don't know if it does. The point still stands: trustworthy medical products are built with parts that don't fail. And that's rarely the cheapest part.
The cost of getting this wrong isn't the part cost
Let me give you numbers from my own spreadsheets. In 2023, I audited $180,000 in cumulative component spending across four product lines. I wanted to know how much of our so-called savings from switching to cheaper passives was eaten by other things. The answer: 38%.
Not because the cheap components were all junk. Because switching a component is never 'just swap the part.' You have to re-qualify it, run the board through thermal testing, update documentation, and wait longer for delivery. Those costs aren't on the purchase order. They show up in engineering hours and test reports.
Then there's the field failure.
One of our products had a $1,200 redo after a cheap capacitor failed a power-on test. The board was one of a hundred. We saved four cents per part. The redo cost more than the entire annual volume of that component. Simple math. But finance didn't see it until I built a total cost of ownership calculator.
Here's what I learned: the 'cheap' option is often the expensive one, just delayed.
I also had to make a decision with 24 hours, once. Our usual distributor couldn't deliver a Kemet T520 variant in time for a prototype run. A manufacturer I didn't know had a cross-reference at half the price. Normally I'd do a proper qualification. There was no time.
In hindsight, I should have pushed back on the schedule. But with the customer waiting, I made the call to source the Kemet part from a slower authorized distributor and pulled a board from another project. It wasn't a perfect decision. It was a 'best available information' decision.
The cheaper part might have worked. But 'might' isn't a sourcing strategy.
What I recommend now: total cost of ownership
Look, I'm not saying you need a Kemet T520 in every design. That would be lazy. The T520 series is expensive for a reason, and if your application doesn't need that performance, buy something else.
This is where I'll be honest: sometimes a generic ceramic capacitor is enough. Sometimes an unshielded inductor is fine. But if you're building something that sits on a person's arm and gives them medical data, or something that will run for 15 years in a factory, you need to think in total cost, not unit cost.
The same logic applies to Kemet inductors. They cost more than a bare drum-core coil. In a space-constrained power circuit, that extra cost buys shielding and predictable saturation. I once spec'd a cheaper inductor because the datasheet looked fine. Our EMI test failed. We added shielding tape, copper cans, and two engineering iterations. The tape and copper cost more than the Kemet inductor we should have used in the first place.
In my procurement spreadsheet, I track four numbers:
- Unit price (obviously)
- Qualification and testing cost per new part
- Failure and rework cost per field return
- Supply chain risk cost—how likely is this part to disappear?
When I put those numbers together, the Kemet T520 and Kemet inductors start to look more reasonable. You're not paying for a logo. You're paying for consistency: same part, same specs, same reliability, delivered through a channel that can trace the lot back to production.
Authorized distributors matter because counterfeit and grey-market parts are a red flag. Buy through an unauthorized source and you save money until the day the part fails and nobody backs it. That's a risk I won't take.
And if you're selling a product that touches health, remember the FTC. Per FTC guidelines on advertising claims, if you make a claim about a product, you need evidence. Your component selection is part of that evidence. I'm not a lawyer, but I'd rather explain why I chose a rugged component to a design review than to a regulator.
A quick test before you downgrade a component
Before you approve a cheaper alternative, ask three questions. First, has this part been through your design's electrical stress conditions? Not the datasheet's nominal conditions. Your conditions. Second, can the supplier guarantee lot traceability for five years? Third, what happens to the product if this part fails in the field? If the answer to that third question is 'a recall,' don't buy the 28-cent cap. That's the kind of decision that should be a no-brainer.
No brand is right for every bill of materials. I respect that.
But when someone asks me if the Kemet T520 is worth it, my answer is simple: it's probably worth it for the designs where failure isn't an option. For everything else, do the math. Use the total cost, not the line item price.
Money is a constraint. Bad decisions are more expensive.
