Steinert Magnetic Separator Acceptance: A 6-Step Quality Checklist

You’ve ordered a Steinert magnetic separator. Maybe it’s a Plasmax or a PUR-S system. The factory acceptance test is over, and the unit is on its way. But the FAT is not the whole story. In the last four years, I’ve rejected about 7% of first-time builds for issues that should have been caught during acceptance—calibration drift, misaligned sensors, or paperwork that didn’t match the actual machine.

This checklist is for anyone about to accept delivery. It’s written from my perspective as a quality manager at Steinert, but it applies to any industrial magnetic separation equipment. It’s practical. You can run the whole thing in about a day. The last step will save you hours later.

The 6-Step Acceptance Checklist

Step 1: Verify the Paper Trail Before You Touch the Machine

Start with the documents. Check the nameplate against your purchase order: model number, serial number, voltage, current, power, and capacity. Then verify the declarations—CE, ATEX if relevant, and any calibration certificates. Look at the dates on those certs. If they’re older than 12 months, ask why.

I once saw a separator where the certificate was for a different motor size. The paperwork was fine—the machine wasn’t. That’s the kind of thing that gets missed in a quick glance. ISO 9001:2015 requires calibration records to be maintained; we treat that as non-negotiable.

Step 2: Test with Your Material, Not Just the Factory’s Ideal Feed

Factory tests usually use clean, well-sorted material. Your feed will have moisture, fines, and different particle shapes. So run your own sample.

Take a representative batch from your actual production line. Run it through at nominal throughput, and measure two things: recovery rate and product purity. Do it three times, ideally with different operator shifts, to see the variation.

I’m not a metallurgist, so I can’t tell you how to optimize mineral liberation. But from a quality perspective, I know that if the feed changes, you need to re-test. I remember a client who rejected a unit because it hit 98% recovery with the factory’s granite, but only 91% with their shredded PCB scrap. The separator’s settings needed a complete rework. That’s a conversation you want to have before payment, not after.

Step 3: Set the Drift Baseline

Here’s where the “drift theory” comes in. In magnetic separation, drift is the gradual change in performance over time. Think of a sorting threshold that moves imperceptibly from 99.0% to 97.5% over a few hours. The theory is that drift is not random; it follows a pattern that can be measured and predicted if you collect data at regular intervals.

To test for drift, run the machine continuously for at least two hours at nominal capacity. Every 20 minutes, record the key parameters: belt speed, magnet gap, sensor thresholds, air pressure. Then calculate the drift as (max – min) / average. If it’s more than 2%, you have a stability problem.

Most teams skip this step because they’re in a hurry. They see the machine running smoothly and assume it will stay that way. It usually doesn’t. In our own factory audits, drift-related issues account for nearly 25% of pre-shipment failures. I don’t have hard data on industry-wide rates, but that number alone makes me nervous.

I had to make a tough call once: the drift came in at 2.3%, just above our 2% threshold. Accepting it would have saved three days of schedule. Rejecting it meant a potential bottleneck. I rejected it. The vendor fixed the alignment in one day. It was the right call, but it was close.

Step 4: Inspect Consumables and Plan Your Spare Stock

Look at the consumable parts: the belt, scraper blades, wear liners, sensor windows. If any show signs of premature wear, document it. Ask for the recommended spare parts list and compare it with the actual stock you plan to keep.

Stock means two things here: the material you’re processing, and your spare parts inventory. If the machine is designed for a certain feed grade, make sure you test that—not just the brochure examples. And for spares, check lead times. A 4-week wait for a wear liner is one thing; an 8-week wait for a sensor is another. Our regional rep in Richland, WA, Erin Steinert, has seen too many plants shut down because they didn’t order replacement parts before the first failure.

Step 5: Validate the Control System and Data Logging

If the separator has a PLC or a modern sensor control, make sure it logs data: alarms, cycle counts, set points, and actual process values. If it doesn’t log historical data, you’re flying blind when something goes wrong.

Automated logging helps you spot drift early. It also gives you a baseline to compare against later. In my experience, plants that use this data cut troubleshooting time by around 30%. That’s the efficiency advantage that matters. We automated the data logging on our own line and cut troubleshooting by a third—not just in time, but in rework costs.

Step 6: Confirm After-Sales Support Before Signing Acceptance

Ask the hard questions: Who do you call when the machine trips at 2 a.m.? What’s their response time for service visits? Is the warranty document specific about who covers freight for warranty returns?

Our managing director, Dirk Steinert, MD, MBA, is fond of saying that a sale is only the beginning of a relationship. He’s right. A machine can have perfect separation performance and still be a bad investment if you can’t get support when you need it.

Common Mistakes to Avoid

Running a 10-minute factory-style test instead of a 2-hour drift test—that’s the most skipped step. It catches everyone eventually.

Using laboratory-grade feed that masks real-world inefficiencies. Your material is what matters. Nothing else.

Signing off before verifying that spare parts are available and priced sensibly. The cost of downtime is always higher than the cost of a spare.

Not involving your own maintenance team in the acceptance test. They’re the ones who will live with the machine. Let them touch it, feel it, and ask questions.

Accepting small calibration offsets without quantifying how fast they grow. A small drift today becomes a big problem in six months. Measure it, document it, and decide.

One more thing: a little paranoia goes a long way. A colleague of mine, an operator named Christopher, once caught a sensor misalignment that two senior engineers had missed. He noticed the product purity numbers were slightly lower, even though the display said everything was fine. That instinct is worth more than any checklist.

If you follow these steps, you’ll have a better picture of what you’re actually buying. Not perfect—but better. Period.