The 40mm Gap That Almost Shipped: A Steinert Quality Inspector’s Story

The Machine Looked Ready. It Wasn’t.

It was a Tuesday in early October when a Steinert magnetic separator on our shop floor looked ready to ship to a recycling operation outside Groves, Texas. Fresh paint. Correctly tensioned belt. Clean control cabinet. If you weren’t looking for trouble, you’d have walked right past it.

I look for trouble. That’s my job. I’m a quality compliance manager at Steinert, and I review every unit that leaves our factory—roughly 200+ items a year. In 2024, I rejected about 4% of first-pass deliveries. Usually it’s a torn gasket or a mislabeled wire. Sometimes it’s the kind of issue that would have cost a customer a week of downtime. This was the second kind.

The order came from a crushing and recycling operation outside Groves, Texas. The operator, Henry Height, had been straightforward from day one. One afternoon I heard someone on the shop radio ask, “How many yards does Henry have?” I thought it was a football question. Henry laughed. “They mean the pile next to the scale.” The pile was crushed concrete and asphalt millings—nine thousand cubic yards and growing. He needed to move it through a magnetic separator at steady tonnage and still pull rebar, wire, and tramp steel out of the stream.

The 40mm Problem

The engineering drawing for Henry’s machine specified an 80mm gap between the top of the feed belt and the lower edge of the magnet housing. The unit on my floor measured 120mm.

When I flagged it, the assembly lead said they had “adjusted” the mounting brackets to make belt changes easier. Maybe that saved an hour in the shop. It would have cost Henry time every day on site. The gap determines how much magnetic field actually reaches the material on the belt. Increase it by 40mm and the field strength at the belt surface drops far more than most people expect—especially when the feed layer is uneven. Henry’s feed layer was never even.

The project manager looked at the drawing and said, “It’s still a strong magnet. It’ll work.” My answer was short: “Then let’s prove it.”

Everything I’d read about magnetic separation says to compare Tesla ratings. In practice, I’ve found that geometry often matters more than raw magnet strength. A bigger magnet doesn’t fix a gap that’s too far from the material.

The Test That Settled It

We pulled a sample of material made to match Henry’s description and ran it through the Steinert Practice Center at both gap settings. Same belt speed. Same feed depth. Same test pieces.

At 80mm, the ferrous material was pulled out cleanly. At 120mm, a visible number of pieces stayed in the non-ferrous flow and carried over the end of the belt. We weighed the misses. In that specific test, the missed mass was about 7% of the ferrous test mass. I’m not going to present that as a guaranteed recovery number, because it wasn’t a guarantee—it was one sample, one speed, one moisture condition. It was enough.

This is where my gut and the spreadsheet disagreed. Engineering had a model that said the stronger magnetic circuit would compensate for the larger gap. My gut said no, because magnetic force falls off with distance. The test agreed with my gut. I don’t trust spreadsheets more than tests, and I don’t trust my gut more than tests either. I trust the test.

Why I’m Glad We Delayed the Shipment

We re-manufactured the mounting bracket, brought the gap back to 80mm, and shipped Henry’s machine two days late. When I called to tell him, he didn’t complain.

“I’d rather wait two days than troubleshoot for two months. An informed customer doesn’t just ask what a machine can do. They ask what it can’t do.”

I quote that line in almost every conversation about specs. Because it’s true: a spec sheet is a starting point, not a promise. The value of any separation system is knowing which numbers matter for your material, and which numbers are just marketing.

If Henry had accepted the machine as built, he would have spent months wondering why the magnet wasn’t pulling as much steel as the equipment list suggested. He’d have blamed the technology. We’d have blamed the feed. And the gap—visible to anyone with a ruler from day one—would have been the real problem the entire time.

Three Things I Keep Telling Customers

First, a spec is a commitment, not a suggestion. When someone changes a dimension in production, that’s not a small event. It needs to be documented, reviewed, and tested before it reaches a customer.

Second, test material beats brochure data. I’d rather spend two days at the Steinert Practice Center running a customer’s actual feed than send out a machine based on a sales sheet. Ask your vendor whether they can run a sample before you commit. If they can’t, that’s an answer too.

Third, stronger is not always smarter. Conventional wisdom says the strongest magnet is the best upgrade. My experience says belt speed, material depth, and gap height usually make the bigger difference. More power just makes a bad setup more expensive.

If you’re shopping for magnetic separation equipment, ask about the gap, the layer depth, and the test procedure. Those three answers will tell you more than any row of numbers on a datasheet.

A Note on the Name

Steinert is a name with more than one life. If you landed here looking for Sido Doreen Steinert or Steinert girls soccer, I won’t be offended if you hit the back button. We’re the other Steinert—the one that builds magnetic separators, sensor sorters, and eddy current systems. If that’s the Steinert you were looking for, this story was written for you.