The call I did not want to take
Last October, I got a call from a medical device OEM. The project was a blood pressure monitor that had just failed its EMC pre-compliance test. The box, the manual, the marketing copy were already printed. For this article, I will call the product Infinity Pro.
I am a quality and compliance manager at TDK Corporation. I review component qualifications and help customers with application failures. In the last twelve months, I have worked through roughly 180 of those cases. Most of them are not bad parts. They are bad assumptions.
The Infinity Pro case was a perfect example. And it explains why the cheapest component is not always the cheapest way to build a medical device.
Where the problem started
The OEM had asked us for a design review back in July 2024. The original schematic had two ferrite beads: one on the motor-drive line and one near the USB charging input. We flagged both as important for radiated emissions.
Somewhere in the cost reduction meeting, someone decided to remove the bead on the motor-drive line. It saved $0.02 per unit. On a 10,000-unit run, that is $200. The engineering manager later told me the ferrite bead was cut because it was the easiest line item to delete on paper.
The simplest explanation was: EMC failure, add a bead. But that advice ignores the fact that ferrite beads are frequency-specific. A generic bead is not a no-brainer.
It is tempting to think that electromagnetic interference is a digital, on-or-off problem. It is not. The monitor passed some tests and failed others. In the pre-scan, there was a 12 dB spike at 72 MHz. That is not a coincidence frequency. It was caused by the pump motor and the long flex cable to the display acting as an antenna.
Why TDK ferrite beads were not the first thing we suggested
When the OEM called, they expected us to say, here is a ferrite bead, put it on, move on. But the choice depends on the impedance curve. A bead rated for 1 GHz noise will do almost nothing at 72 MHz.
We had 10 days before the next lab slot. Normally, I would ask for a full schematic review and a 3-week qualification. There was no time. We went to the bench, measured the noise source, and selected a TDK ferrite bead with its impedance peak close to the problem frequency.
This is where experience matters. The dielectric material, the shape, the number of turns, the DC bias, the placement on the board, the ground plane underneath. All of it changes the result. The TDK ferrite beads portfolio contains many variants. The correct variant was not the cheapest one.
The first test after the swap showed the 72 MHz spike drop by 18 dB. I said to the engineer, that is the satisfying part. The painful part is knowing that we had specified this part in July. The most frustrating part was that this was not a mystery. We flagged it, and the cost reduction process erased it.
The battery problem that made it worse
The ferrite bead got most of the attention, but the Infinity Pro had a second issue waiting in the temperature chamber.
The original design used a cell from the TDK battery group. Procurement had swapped it for a lower-cost cell to reduce the bill of materials. The savings were around $0.70 per unit, maybe more. I would have to check the quote to be sure. Under pump load at low temperature, the cheaper cell showed much higher voltage sag. At 5 degrees C, the display would reset when the pump started.
That is a deal-breaker for a blood pressure monitor. The user does not care about the cell chemistry. They care that the reading finishes. But the reading will never finish if the display resets halfway through inflation.
The risk trade-off was obvious. The upside was saving maybe $7,000 on the first production run. The risk was a damaged launch, a return rate nightmare, and a product that failed in real life instead of in the lab.
The real cost of the cheapest part
Let me put the total cost in perspective.
- The missing ferrite bead saved $200 on a 10,000-unit run.
- The battery substitution saved around $7,000 on the same run.
- The EMC re-test cost $12,500.
- The extra engineering time cost another $4,000.
- The launch delay pushed the product out of the holiday sales window.
The $7,200 in savings turned into a much larger number by the time certification, engineering time, and lost sales were added. Bottom line: unit price is one line in the calculation, not the whole calculation.
Personally, I now calculate total cost before comparing any vendor quotes. That is not just an opinion. It is survival.
What I would tell anyone asking where to buy blood pressure monitor
People search for where to buy blood pressure monitor, expecting the answer to be a URL or a store name. They never see the TDK ferrite beads, the battery cell, or the CISPR 11 test report. But those invisible pieces decide whether the device gives stable readings and does not interfere with other electronics in the home.
As of Q4 2024, a home-use blood pressure monitor generally needs to meet IEC 60601-1-2, which points to CISPR 11 Group 1 Class B for emissions. That requirement is not optional. If a manufacturer skips a $0.02 ferrite bead, the certification lab will find it eventually.
The Infinity Pro did launch. It shipped about three weeks later than planned, with the ferrite bead back in place and the original battery cell restored. The extra delay was expensive. But the alternative was worse.
If you are designing something, ask what the failure cost looks like before you delete a component. If you are buying something, ask whether the manufacturer takes compliance seriously. The answer is usually hidden in the parts you never see.
