The Steinert Magnetic Separator: A Quality Inspector's Buying Checklist

Let me start with a quick disambiguation. If you're here because you searched for maladie de Steinert transmission, that's not the same Steinert. Maladie de Steinert is myotonic dystrophy, and 'transmission' in that context means genetic inheritance, not a gearbox. This article is about the Steinert magnetic separator from Steinert GmbH. It is also not about the LEGO Millennium Falcon, and it won't answer 'why is Henry not playing'—unless Henry is your lead operator. If he is, the answer is in the last section.

If you're still here, you're probably comparing magnetic separators and need a way to cut through the sales talk. This checklist is for anyone who's about to commit real money to a steel box with a big magnet and hopes it works. It is not a replacement for engineering; it is a way to make sure your engineering is asking the right questions.

I review equipment specifications for a living—roughly 40 technical dossiers a year in mining and recycling. I've rejected 15% of first vendor submissions in the last 12 months because spec sheets didn't match test data. This article is the checklist I run before I approve a Steinert magnetic separator purchase.

The Six-Step Checklist

Use this if you're buying a magnetic separator for scrap, mining, recycling, or waste processing; if you want a defensible purchase decision; or if you've been burned by a 'cheap' machine before. There are six steps. Don't skip step 4.

Step 1: Define your feed and target before you look at machines

A Steinert magnetic separator is not one machine. There are overhead belt magnets, magnetic drums, eddy current separators, and sensor sorting systems. If you don't tell the sales engineer what you're feeding, you'll get a long but fuzzy answer.

• Feed material: shredded e-waste, incinerator bottom ash, auto shredder residue, glass cullet, crushed slag, or something else.
• Target fraction: ferrous metal, stainless steel, aluminum/copper, or removing ferrous to protect downstream equipment.
• Particle size range: 'one to six inches' is not enough; you need the distribution and the percentage of fines.
• Moisture and bulk density: wet materials behave differently; low bulk density can limit throughput before magnet capacity does.
• Throughput target: tons per hour at the separation quality you expect, not at maximum belt speed.

Checkpoint: if you can't name all five, call your operations team before you call any vendor.

Step 2: Ask for the specification behind the brochure

The datasheet says 'up to 20 t/h' or '900 mT at belt surface.' Those are starting points, not guarantees. Here's something vendors won't tell you: the feed rate in a product data sheet is often the maximum the system can physically move, not the maximum at which it can separate cleanly. Those are different numbers.

Ask for:

• Recommended burden depth (layer thickness) at your target throughput
• Minimum particle size for reliable recovery
• Belt speed and magnetic rotor speed ranges
• Behavior with tramp ferrous metal and oversized pieces
• Calibration or sensor refresh intervals for sensor sorters

If the answer is 'it depends,' that's a good sign. If it's a single perfect number with no test report, ask for the report.

Step 3: Test with your material, not with a lab sample

The easiest way to avoid a bad decision is to run your own material through the machine. Steinert has a Practice Center and test facilities. Use them. A 'representative' sample does not mean the cleanest bucket from your stockpile. It means the wet, dusty, slightly oversized mix that actually shows up in the hopper.

In Q1 2024, I sent 500 kg of shredded PCB waste to a test facility. First pass recovered 89% of the target copper. After adjusting splitter position, belt speed, and rotor speed, it reached 96.5%. The test report is now the baseline in our purchase order.

If a vendor won't let you test with your own material, that's your answer. They don't want it on record.

A test also gives you the variables that matter. You'll see what the machine sounds like, what sparks fly, and which adjustment matters most. That's not in the brochure.

Step 4: Calculate total cost of ownership, not purchase price

This is where penny-wise, pound-foolish decisions happen. The purchase price is one line in a much longer ledger. I use a simple TCO list for every magnetic separator quote: purchase price, freight, installation, foundation work, electrical connections, commissioning, sample testing, spare parts, consumables, energy, maintenance labor, downtime risk, and operator training.

I remember a buyer saving €35,000 on a cheaper machine. The gap between the magnetic rotor and the belt was shorter, so material had to be processed twice to reach the target grade. The extra rehandling cost erased the savings in the first year. Don't hold me to the exact euro amount; the point is the pattern. The cheapest quote is only cheap at the moment it's accepted.

People think expensive vendors deliver better quality. Actually, vendors who deliver quality can charge more. The causation runs the other way. When you calculate TCO, you're not paying for a brand; you're paying for consistency.

Step 5: Write acceptance criteria into the purchase order

If you don't define 'pass,' the vendor will define 'pass,' and it will be more generous than yours. Acceptance criteria must include recovery rate, product purity, throughput, noise, and wear-related limits. But don't write 'recover >95% of copper' until you've agreed on the test method and feed conditions.

Put these numbers in the PO:

• Feed conditions: particle size, moisture content, throughput and belt speed during the acceptance test
• Recovery threshold: percent of target metal recovered, measured by hand-sorting or equivalent method
• Purity threshold: acceptable percentage of unwanted material in the concentrate
• Minimum continuous test duration at full load
• Tolerance on splitter position, gap settings, and conveyor speed

The tolerance part matters more than people think. In print quality, a Delta E of less than 2 is considered acceptable for brand-critical colors; above 4 is visible to most people (Pantone Color Matching System guidelines). For a separator, the equivalent is a splitter position: a 5 mm change may not look like much, but it can swing a product from accepted to rejected. Specify the tolerance, not just the target.

Step 6: Plan maintenance before you need it

The most expensive magnet is the one that's waiting for a spare part. Before you sign, ask:

• What is the lead time for wearing parts: belts, bearings, rotor components, sensor windows?
• Is the magnetic rotor proprietary, or can it be serviced locally?
• Are consumables available as standard parts or factory-only?
• How long is a service engineer on site if something breaks?
• Is there a recommended spare parts kit for the first year?

I've seen a line stop for 11 days over a $600 sensor cable that had to be flown in. The spare parts kit looked optional during the sales negotiation. It stopped looking optional on day 6.

Three Mistakes I See Again and Again

1. Chasing maximum gauss

A higher magnetic force at the surface doesn't always mean a deeper capture zone. It can also mean more ferromagnetic fines stuck to the belt. You need the field profile for your application, not the brochure headline.

2. Leaving the operator out of the project

If somebody asks 'why is Henry not playing,' it's usually because Henry wasn't told the machine arrives next week. Invite the operator to the test run, installation, and training. They'll hear a bad bearing before the vibration sensor does, and they know which materials used to hang up in the hopper. That knowledge is worth more than a manual.

3. Testing only the average material

Your input will change over time. What happens when the batch has more copper, less aluminum, higher moisture, or more small particles? Test the edge cases, not just the average, and plan an annual re-test with a new sample.

Run the test. Write the thresholds. Involve the operator. The total cost will take care of itself.