Magnetic Separation Equipment: A Buyer's Guide Based on Real Mistakes

If you're searching for "Steinert" because you heard about Otto Steinert artwork or happened across Steinert High School football highlights, you might be surprised to land on an industrial equipment page. Honestly, I get it. The name shows up in weird places. But right now we're talking about magnetic separation for mining and recycling — and I can promise you, this is the kind of Steinert that actually saves you money if you pick the right system.

I've been handling orders for magnetic separation equipment for about six years. In my first year (2019), I made the classic mistake of recommending a drum magnet for a stream that was 80% non-ferrous. That order — about $12,000 — ended up in the customer's warehouse collecting dust. They needed an eddy current separator. The lesson cost us a client and a lot of embarrassment. Since then, I've documented 47 significant screw-ups across our team, totaling roughly $86,000 in wasted budget. Now I maintain our internal selection checklist.

Here's the thing: there is no single "best" magnetic separator. It depends entirely on your material, throughput, and contamination tolerance. Let me walk you through the three most common scenarios I've seen — and help you figure out which one you're in.

Scenario A: You're Removing Large Ferrous Contamination from a Coarse Stream

Think: construction and demolition waste, shredded cars, or primary crushing in mining. You want to pull out heavy iron pieces before downstream processing.

What works: A cross-belt magnetic separator or an overhead magnetic plate. These are workhorses. They handle big chunks, they're relatively cheap, and they don't care about belt speed variations.

What doesn't: A drum magnet with a shallow field depth. I once ordered a 36-inch drum for a client processing concrete rubble. The material was 6-inch minus. The drum simply didn't pull the rebar pieces off the belt — they just rode over the top. That mistake cost $4,800 in shipping and restocking. We ended up swapping to a cross-belt, and the client was happy. But I should have known: for large ferrous objects, you need deep field penetration, not pretty surface grabbing.

Pro tip: If your material is coarser than 4 inches, go cross-belt. If it's finer (under 2 inches), a drum can work — but check the exact gap and belt speed. I learned this the hard way in September 2022. A client named Lewis from a recycling yard in Ohio called me in a panic because his drum magnet wasn't catching 12-inch rebar. I felt terrible.

Scenario B: You're Separating Non-Ferrous Metals from a Mixed Flow

Aluminum cans, copper wire, shredded electronics — typical for e-waste or municipal recycling. This is where eddy current separators (like Steinert's EddyC or the newer ECS models) shine.

What works: An eddy current separator with a high-speed rotor (3000+ RPM) and a properly tuned magnetic pole pattern. But here's the catch: it doesn't work on stainless steel (unless it's ferritic), and it's terrible for particles below 3 mm.

What doesn't: A magnetic drum or a plate magnet. They only attract ferrous metals. Non-ferrous will just pass through. I can't count how many buyers ask: "Why can't I just use a magnet for aluminum?" Honest answer: you can't. That's just physics.

When to say no to an eddy current: If your material is mostly plastic with small metal contamination, an eddy current might actually be overkill. You might be better off with a sensor-based sorter that identifies materials by color or near-infrared. But if you're dealing with mixed non-ferrous metals (aluminum, copper, brass), eddy current is your best bet — assuming the particle size is above 5 mm. Below that, efficiency drops off a cliff. I learned this after a $3,200 order where the customer wanted to separate 2 mm aluminum particles. The unit basically did nothing. That was a rough call to Lewis's counterpart in Chicago.

Scenario C: You Need Ultra-High Purity Sorting — Like Removing a Specific Color of Plastic or Recovering Valuable Minerals

This is where sensor-based sorters (like Steinert's UniSort series) come in. They use cameras, near-IR, or laser sensors to detect material properties and then blast them with air jets. These are expensive — think $150k+ — but for high-value streams like wire chopping or mineral processing, they pay for themselves fast.

What works: A combination of magnetic separation (to remove ferrous) followed by sensor sorting. For example, remove iron first with a drum, then use an optical sorter to separate copper from brass. That's a two-step process. Don't try to skip the magnetic step — it will wreck your sensor windows. I had a client who bought a sensor sorter without a pre-magnet, and within a week, fine iron dust had scratched the camera lenses. $1,200 replacement cost.

What doesn't: Sensor sorters on wet materials. They rely on optical detection, and moisture causes false readings. Also, if your material is black (like shredded circuit boards), standard cameras struggle. You'd need a near-IR or XRT sorter. This is a niche but important caveat: don't assume a camera sorter works for everything.

How to Figure Out Which Scenario You're In

It took me three years and about 150 orders to understand that the "best" separator is highly context-dependent. Here's a quick self-diagnostic:

• What's your target metal? Ferrous only? Go Scenario A. Non-ferrous? Scenario B. Both? You probably need a two-stage system (drum + eddy current).
• What's the particle size range? Over 4 inches? Cross-belt. Under 2 inches? Drum or eddy current (depending on metal type). Under 5 mm? Eddy current won't work well — consider a sensor sorter or wet gravity separation.
• What's your throughput? Over 50 tons per hour? Sensor sorters get expensive and complex. Mechanical magnets (drum, cross-belt) are more scalable.
• What's your required purity? If you need 99.9% removal of a specific metal, you'll need multiple passes or a sensor sorter. If 95% is fine, a simple magnetic system will do.

I can only speak to my experience with mid-size B2B operations (typically 10-100 tons per hour). If you're dealing with a massive mine or a tiny artisanal setup, your calculus might be different. Also, this was accurate as of late 2024. The industry changes fast — especially with new sensor technologies. Verify current specs with your supplier before buying.

Honestly, the biggest mistake I see is people trying to use one machine for everything. It's like trying to ski every trail at Milano Cortina 2026 with a single pair of skis — you'll crash. Buy the right tool for your specific material. And if you're not sure, run a test. Steinert has a practice center where you can send your material for a free trial. I've personally witnessed a client save $60,000 by testing first instead of guessing.

Anyway, I hope this helps you avoid the same embarrassment I went through. If I could go back and tell my 2019 self one thing, it would be: "Stop assuming, start asking about particle size."