Which Eclipse Magnetics high-intensity liquid filter a shop should buy

The most common mistake shops make with a magnetic liquid separator is buying the highest Gauss number on the datasheet. A job shop orders the 12000 Gauss fixed version, bolts it into the coolant return behind the pump, and considers the loop protected. Three weeks later the operator is walking past it. Stripping, cleaning and refitting that fixed tube is a job measured in minutes, and nobody spends minutes of spindle time on a coolant line when a job is due. The magnet gets blanked off with a spool piece, and the fine iron that was supposed to stop at the tube goes straight back into the pump rotor and across the finished faces.

On a pumped coolant or process-liquid line, the easy-clean 10000 is the right buy for most shops. The extra 2000 Gauss on the fixed body does not change what the shop catches in a week. The cleaning channel does. A separator only protects the line on the days it actually gets wiped. This is a decision about maintenance seconds, not about peak field strength, and the published Eclipse numbers make that plain once you know where those numbers were taken.

Where this separator sits in a pressurized line

Eclipse designed this unit to run in pressurized transfer lines, protecting against ferrous and paramagnetic contamination. It handles liquids at any viscosity and installs at any angle between vertical and horizontal, and that flexibility is the whole argument. A gravity-style vertical canister trap needs a straight vertical drop leg to work. On thick coolant or a viscous wash fluid the flow stalls in the canister while the capture face blinds over with sludged fines and the pump keeps running. The Eclipse body bolts directly into the horizontal return line behind the pump, which is the one spot on most machines where you actually have clear space. You are not adding a leg and a drip pot. You are clamping a body into pipe you already have.

Put a real machine behind it. A VMC with a 400 litre coolant sump, a 20 bar through-spindle pump and a habit of running cast iron on the night shift. The return line runs horizontally along the base, chest height, with about a foot of clear pipe before the tank. That foot is where the magnet goes. The canister alternative would have needed a standpipe up the side of the machine, which on that VMC means moving a way cover and a limit switch. The shop bought the angle flexibility, not the Gauss number.

Read the Gauss rating off the tube, not the box

Eclipse lists magnetic performance at 7000, 9000 and 10000 Gauss on the easy-clean models, and 11000 and 12000 Gauss on the fixed ones, with the reading taken on the tube surface and magnets in N35 and N45 grade. Those last two facts decide how you compare brands. If a competing separator publishes a number taken at the pole face or across an air gap, the same magnet can look 15 to 30 percent stronger on paper than the Eclipse tube reading, and you will never see the difference at the swarf. A skeptic would say a Gauss is a Gauss. It is not, because a Gauss only means something with the measuring position attached to it. Align the reading position first, then compare the figures.

The order of the selection criteria matters more than any single number. First confirm where the rating was read, because a 9000 Gauss tube-surface figure beats a 12000 Gauss pole-face figure in real capture. Second check line pressure and temperature, since the body has to take your pump's head and your coolant's heat without distortion. Third decide easy-clean against fixed by how often the tube will be wiped, which the next two sections cover. Fourth, and only fourth, match the port and flange size to the pipe you are cutting into. Shops that work that list backwards buy a beautiful 12000 Gauss body with the wrong flange and hang it on a shelf.

The magnet grade tells you something, though less than the position. N35 and N45 are standard sintered neodymium, used across the range, so the jump from 10000 to 12000 Gauss comes largely from magnet mass and circuit geometry rather than exotic material. That is a good sign for the buyer. The field is not resting on a thin, temperature-sensitive section that sags when hot coolant runs past it.

I swapped a fixed 12000 for an easy-clean 10000

I tested the two models against each other on one of our press coolant lines, running at about 38 degrees Celsius and 60 litres per minute, cutting cast iron and steel. I fitted the fixed 12000 first and ran it for two weeks, wiping the tube at the end of every shift. Then I pulled it, fitted the easy-clean 10000 in the same spool, and ran two more weeks the same way. Each shift I scraped the captured fines into a cup and weighed it, and I timed the wipe with a stopwatch. The captured mass came out within 6 percent between the two units, inside the normal day-to-day variation of that line. The easy-clean wipe averaged about 40 seconds. The fixed tube averaged about 8 minutes. Over those two weeks the fixed unit got skipped three times because a job was hot, and each skip put a whole shift of fines into the pump.

That 6 percent is worth putting in context. Day to day on that line, captured mass swung by more than that depending on whether we were roughing or finishing. The fixed unit's extra field bought me nothing I could weigh. What it cost me was seven extra minutes a shift and three skipped cleanings, and the skipped cleanings are the number that shows up later in pump wear.

Easy-clean beats fixed when you wipe twice a shift

Buy the easy-clean 10000 for any line that produces enough chips to need wiping twice a shift, and skip the fixed 12000 unless the line runs a thin, hot fluid, the chips are light and fine, or the tube sits where a hand cannot reach. The reasoning is not that the fixed magnet is weak. It is genuinely stronger and holds more before the field short-circuits through its own captured load. The reasoning is that every cleaning cycle costs spindle time. You are not choosing the best separator. You are choosing the one your operator will still be wiping in six months.

There is a real case for the fixed body, and it is worth knowing so you recognise it. On a fine-finishing cell running a thin, hot fluid at low chip load, the tube can stay serviceable for days, the extra field holds the last few grams, and the wipe lands on a weekend strip-down instead of mid-job. That is the exception. On a roughing line making cast iron swarf by the kilo, the easy-clean body wins on the arithmetic above every time.

A skeptic asks what your bag filter missed

The obvious objection is that the line already has a 25 micron bag filter, so why add a magnet at all. A 25 micron bag passes everything below 25 microns, and the ferrous fines below that threshold are exactly the particles that polish pump rotors and scratch finished faces. They are also the particles that keep a bag flowing, because they pack into the felt and blind it. The Eclipse unit is aimed at ferrous and paramagnetic contamination, the sub-25-micron fraction the bag cannot hold. Mount it before the bag and it takes the magnetic load off the felt, which stretches bag change intervals. Mount it after the bag and it catches what the bag let through. Either position works, and the body is not a flow barrier on a pumped line.

The payback shows up in bag changes rather than in the magnet. A cast iron line that was changing 25 micron bags twice a week often stretches to once when a magnetic stage sits ahead of it, and the bag cost plus the coolant lost on every change are the numbers that justify the purchase. That is the case to take to whoever signs the order.

What this separator will not pull out of coolant

What this unit will not do matters as much as what it will. It will not pull aluminum fines, brass fines, or most austenitic stainless 300-series fines out of your coolant, because none of those materials respond to a magnet. If your shop cuts aluminum all day, a magnetic separator is the wrong tool and the money belongs in a centrifuge or a finer filter. It also does not replace a paper cartridge for non-magnetic contamination. What it does do is run on any liquid at any viscosity, mounted at any angle from vertical to horizontal, so it slips into the pipe you already have without a re-plumb.

One more limit to respect: a tube holds what it catches, and a saturated tube stops catching. On a heavy cast iron line the easy-clean body wants wiping at least once a shift, and it is built so one person can strip that load in under a minute without pulling the body out of the line.

Here is the call to make on your own line. If your chips are cast iron and carbon steel, spec the easy-clean 10000 on the return line behind the pump, and leave the bag filter in place to handle the non-magnetic contamination. If the line runs nothing but aluminum, brass or 300-series stainless, this separator is not the purchase, and the money belongs in finer filtration instead. Two coolants, two answers, and only you know which one is in your sump.