I handle incoming orders and field returns for power supply and electronic component customers. I've been doing this for seven years—or rather, eight this spring—and I've personally made and documented eleven significant mistakes, totaling roughly $18,000 in wasted budget. The mistake I want to walk you through is the one that still shows up in our review notes: a TDK-Lambda high voltage power supply that wasn't dead at all.
The Surface Problem: The Supply "Failed" on the Bench
A customer calls. Their TDK-Lambda high voltage power supply trips the instant they switch it on. No output voltage. Overcurrent LED lit. They're already asking for a replacement. I understand the urge—I did the same thing in September 2022, and it cost us about $890 in return shipping, restocking, and the week we lost waiting for a replacement. (That invoice still stings.)
The supply came back from the factory with a "No Fault Found" tag. I want to say the customer's cable was about 20 feet longer than the standard setup, but don't quote me on the exact length. The important part: the supply was fine. The cable was the problem. That's it.
A cable is a capacitor, an inductor, and a resistor in disguise.
The Deep Reason: Cables Are Part of the Circuit, Not Just Wires
From the outside, a high-voltage cable looks like a simple connector. The reality is different. A cable has parasitic capacitance between the conductor and the shield. It has inductance. It has insulation resistance. Under high voltage, a long cable can store enough charge to act like a capacitor in parallel with your load. That can make the output look like a short at startup, or make the power supply's current limit trip for what seems like no reason.
People assume cables are electrically invisible. What they don't see is that the cable becomes part of the power supply's control loop. Some TDK-Lambda high voltage power supplies use remote sensing or have tight regulation. Add too much cable capacitance, and the loop sees a different load than the one you intended. The symptoms can be subtle—startup current, output collapse, or an intermittent overcurrent trip.
That's why "use the right cable" is not a marketing line. When I order a replacement, I order a TDK cable designed for that supply and voltage class. Not because the brand is magic, but because length, connector, shielding, and ferrite behavior are chosen for the application.
What N93 Has to Do With It
Now the N93 part. If you've ever looked at a TDK cable or a power inductor and noticed a ferrite core, you may not have thought about which material it is. N93 is one of TDK's ferrite material designations used in power applications and chokes. I don't have hard data on the exact frequency and permeability curves from memory—the datasheet has those numbers, and you should verify the current version before ordering a custom magnetics part. This was accurate as of Q1 2025, but datasheets move. The key point is this: ferrite materials are not interchangeable. N93 is selected for specific flux, frequency, and temperature behavior. Swap it for a random "EMI ferrite" from a parts drawer, and you are not getting the same performance.
That's why a molded ferrite on a cable isn't decoration. It suppresses common-mode noise and helps keep switching transients from coupling into the output. Remove it, or use a generic cable, and you might introduce the kind of intermittent weirdness that gets blamed on the supply. (I have done exactly that, and the blame was mostly mine.)
TDK Technologies Working Together
TDK has multiple technologies—power ferrites, ceramic capacitors, sensors, and power conversion—that all show up in a high-voltage test setup. The power supply is the obvious part. The ferrite and the cable are the parts that keep the application from becoming a collection of problems. A TDK-Lambda high voltage power supply is designed to work with the right input conditions, output load, and cabling. Change any one of those, and you are no longer testing the supply. You're testing the setup.
One more thing: don't assume a new cable is good just because it's new. A molded cable can have a broken shield connection and still look perfect on the outside. A quick continuity check between the connector shell and the drain wire at the other end can save you another RMA cycle.
The Cost of Getting This Wrong
The cost isn't just a repair bill. When you return a no-fault TDK-Lambda high voltage power supply, you pay freight. You pay restocking. You lose time. And you still have the cable problem. In 2023, one customer missed a production deadline because both of us spent a week focused on the wrong part. The fix took twenty minutes.
I don't have hard data on how many returned high-voltage supplies are actually healthy. Based on our returns over the past five years, my sense is that 15-20% of first returns test as no-fault. That's a lot of expensive "surprise, surprise" moments.
Every no-fault return also teaches someone the wrong lesson. The technician who blames the supply today may not look at the cable tomorrow. The "failed" label stays with the part, even when it isn't failed. I've watched that happen more times than I want to admit.
A Short Checklist Before You RMA
Before you ship that supply back, do three things:
- Check the cable. Is it the exact length and type specified for your TDK-Lambda high voltage power supply? If not, test with the correct TDK cable.
- Check the load. Does the output behave differently with the load connected? Test into a known dummy load if possible.
- Check input and grounding. Verify line, neutral, and earth connections. A missing ground can let the output float to unexpected potentials and trigger protection.
If the supply still fails after those three checks, you have a real fault. If not, you just saved yourself a return.
I also avoid telling customers "this supply is guaranteed never to fail." Per FTC advertising guidance (ftc.gov), performance claims need evidence, and the evidence is exactly what a bench test gives us.
How to Unblock a Number on Phone: A Weirdly Useful Analogy
Someone in our office once asked me why I keep repeating "check the cable first." I told her it's the same logic as looking up "how to unblock a number on phone." The article says: open settings, find blocked contacts, tap remove. Simple. But the real work is figuring out that the number isn't blocked on the phone at all—maybe it's the carrier's spam filter. The phone was fine. Same pattern with a "bad" power supply. The unit isn't necessarily broken. The problem is in the part of the system that connects it to the load.
When the cable, load, and grounding are right, the TDK-Lambda high voltage power supply does its job. But no supply is a standalone miracle. It's part of a system—cables, ferrites, and all. Learn to check the part that looks innocent first. That's the lesson. Simple.
