The Magnet Wasn’t the Problem: A Rush-Fix Story From Steinert Blaustein

By Sören Steinert, Applications Engineer, Steinert Blaustein

In March 2024, at 3:45 on a Wednesday afternoon, I was in the Steinert Blaustein test area when the phone rang. A plant manager I’d known for years got straight to the point. “Our overband magnet has lost its pull. We’re seeing iron in the eddy-current fraction, and the smelter rejected one load already. We need a replacement by Monday morning, or we lose the next shipping window.”

I’ve handled maybe 200 urgent service calls during my time at Steinert. I’ve seen rush jobs from a simple field check to a six-figure project that had to be commissioned in a week. So the first thing I felt was pressure. The second thing was doubt. A magnet doesn’t usually lose its pull overnight. Something else had changed.

Let me be clear: I’m not a metallurgist and I don’t play one on LinkedIn. If you need advice on smelter chemistry, ask someone who studied it. What I know is how magnetic separation behaves in a real plant. And the behavior they described didn’t match a dead magnet.

Why sending a rush replacement was tempting

Normal lead time for a magnet that size is two to three weeks. The customer was asking for four days. Any service engineer knows that feeling: you want to say yes because a client is in trouble, and the pressure is enormous.

Here’s the thing, though. If we sent a new magnet and it got installed at the same distance, with the same burden depth and the same belt speed, it would have exactly the same problem. A different magnet wouldn’t fix a gap that was introduced by the conveyor setup.

Our policy since an expensive lesson in early 2023 is simple: before we quote an emergency replacement, we ask for a diagnostic call. It feels slower at first. It has saved us from more bad surprises than I can count.

The test that changed the next 36 hours

Instead of checking inventory, we asked the customer to send measured field values and a video of the line running. They did both within about two hours. I want to say the field reading on the magnet face was around 680 milli-Tesla—don’t quote me on the exact number, because the real story was what happened next.

On paper, the magnet was almost as strong as the day it was first installed. So we got on-site the next morning.

When I arrived, the plant was running, but I could see why the product was failing. I walked around the machine with a tape measure and a gaussmeter. The magnetic field at the belt surface was far below the design value. Why? Because the magnet face was about 310 mm above the belt. The layout drawing called for 220 mm. I’m fairly sure of those numbers, though the exact gap might have been a bit different.

A small gap change matters more than people think. Magnetic force doesn’t fade linearly with distance; it fades fast, because the field has to reach through a widening air gap. An extra 90 mm can be the difference between catching a stray bolt and letting it ride through.

The reason for the gap was boring: a few weeks earlier, a maintenance crew replaced a worn conveyor belt. They raised the magnet out of the way to do the splice, then set it back down one bolt hole higher. Nobody checked the original drawing.

From the outside, it looked like the magnet had failed. The reality was a process problem. The magnet itself was still strong—it was just too far away to do its job.

The temporary fix that became permanent

The client was skeptical. They wanted a new magnet, not a debate about mounting brackets. I suggested a test: lower the magnet back to its specified height and run a sample batch. If the contamination didn’t improve, I’d personally expedite a replacement.

We sent a small service crew that same afternoon. They adjusted the magnet, replaced two rubber wipers, and ran the line for about 45 minutes. If I remember correctly, the iron contamination in the non-ferrous fraction dropped from roughly 0.3% to below 0.1%. The rush delivery turned into a weekend service visit. It wasn’t glamorous, and I’m not going to pretend it was a technological miracle. There’s something satisfying about that, though: after 48 hours of panic, the answer was physics and a wrench.

What this taught me about quality and brand trust

I have mixed feelings about emergency calls. On one hand, I understand why customers ask for rush parts—they’re under contract pressure, and time is the most expensive thing they have. On the other hand, more of these calls are caused by deferred checks and small installation mistakes than by genuinely bad products. The most expensive equipment in the world won’t protect a brand if it’s installed wrong.

Quality perception doesn’t start when the final product reaches the customer. It starts upstream. If your magnetic separator lets tramp iron through, you don’t just have a contaminant; you have a customer who starts to see you as unreliable. That loss of trust costs far more than a spare part or an afternoon of service.

People talk about total cost of ownership as if it’s only a spreadsheet line. For me, it’s this: the cheapest up-front option can be the most expensive when it affects the quality of your output. And good service isn’t an expense—it’s brand insurance.

I’m not saying you should never buy a rush replacement. Sometimes parts genuinely fail, and replacement is the correct answer. But if you’re facing a deadline and your gut says “it’s the magnet,” stop and ask what changed before you order a new one. That hour of diagnosis might save you 48 hours of scrambled delivery—and a lot more than the price of the part.

At Steinert Blaustein, we still test urgent material samples in our lab before committing to a solution. That won’t make every emergency disappear. It does, however, make the fix more likely to work the first time—and that’s the thing customers remember when they decide whose equipment to buy next.