We all know that oil is the lifeblood of any engine. As such, we’ve focused a lot of time and effort as an industry trying to improve and refine its chemistry for the better. A key component of the engine’s oiling system is the oil filter. Designed to keep that lifeblood clean, you can think of your oil filter as the engine’s kidney. So, it seems surprising that there hasn’t been a whole lot of change in the oil filter market over the past few decades, with traditional paper-filter media units getting recommended time and again, even for high-dollar, high-output race engines.

Never content with “good enough” or “because that’s the way it’s always been,” Total Seal‘s Lake Speed, Jr. and Shaver Specialties‘ “Dyno Don” McAskill teamed up to test a pair of oil filters; a traditional paper filter against a modern reusable media filter, and then analyze the results. “The only thing standing between your engine and failure is the oil filter,” says Speed. “The number-one cause of bearing failure is contamination.” McAskill continues, “We’re going to take our mule motor here, that we’ve gotten so much data from, and we’re going to try and break it.”

Now, there are tons of videos on YouTube of people just absolutely wrecking engines for destruction’s sake. But, that’s not what McAskill is talking about. No, the tests he and Speed are conducting at Shaver Specialties are actual, scientific tests, which happen to put the engine in peril. “Essentially what we’re going to be doing to this engine, is what they do in the filter test lab,” Speed says. “They use a special filter test dust that they contaminate the oil with. Then they count the number of particles in the oil before the filter and after the filter.”

On the left is a 35-micron stainless-steel mesh oil filter. On the right is a 25-micron paper-media oil filter. Neither a finer steel mesh nor a looser paper filter were available, so the mismatch was as close as they could get.

The Testing Procedure

Since Speed doesn’t have the magic dust they use in the actual filter testing, he’s using an analog that is not uncommon to find in an engine – iron. Specifically, the fine iron powder used when magnafluxing parts. “We’re going to take 5 grams of that powder, mix it into 5 quarts of oil, and then put it into the engine,” Speed explains. They will take a sample of that oil before it goes into the engine, to be analyzed at SPEEDiagnostix. Then, the engine will be run for one hour on the engine dyno, making a full-power sweep every ten minutes of the hour runtime.

From there, Speed will drain the oil and take another sample of the used oil. Once analyzed, that will show the efficiency of the filter and media. Then, the fun part. They will be running this test twice; once with a standard paper-media filter and once with a wire-mesh oil filter. “We weren’t able to get exactly the same micron rating for the two filters, but they are very close to one another,” says Speed. So at the end of the test, they will have five samples to reference in the conclusion.

5 grams of magnaflux iron powder was added to 5 quarts of fresh oil for each filter test. This resulted in a pretty heavy contamination load.

Running The Paper Oil Filter

With the oil mixed at a ratio of 5 grams of iron powder into 5 quarts of oil, Speed and McAskill fired up the engine on the dyno and started the test. After 10 minutes of running, they made a 3,000 to 6,000 rpm power sweep of the engine, and repeated the process every 10 minutes, resulting in six full-power pulls and 60 minutes of dyno time on the engine.

The first promising sign was that the each of the power pulls on the engine — with 5 grams of iron powder floating in the oil, mind you — the curves were essentially the same at the end of the test as at the start. So, no adverse impact on the engine or its output. “There was no real change in power. Oil pressure is hanging in there,” says Speed. He then pulled the oil sample after letting the engine cool down for exactly three minutes. That is more of a measure to ensure consistent sample collection than anything else.

Running The Wire Mesh Oil Filter

The oil and iron powder was mixed at the same ratio of 5 grams to 5 quarts of oil, and the same testing procedure was used. After the first pull, things looked about the same on the dyno graph. However, on pull two, the power graphs started to separate from one another, showing a loss of power. The same trend continued on the third, fourth, and fifth engine pulls.

After the sixth pull, there was a clear loss of power between the first and final pulls. However, there were no oil pressure anomalies during the test that would have indicated an imminent engine failure. Speed noted during the sample collection that the used oil in the second test was visibly contaminated. With that, the five samples were bagged up and sent to the Speediagnostix lab for analysis.

SPEEDiagnostix tested all five of the oil samples for a variety of things. The main parameter for this test being the number of particles (larger than 4 microns) per milliliter of oil.

The Test Results

The first test results came from the oil straight out of the drum came in at an ISO cleanliness rating of 16. For more on the ISO cleanliness scale, you can check this out, but Speed explains the short and easy way: “All oils have something in them. Every time the ISO number increases, the number of particles doubles.” So, 16 is the baseline rating. A rating of 17 means there are double the number of particles in the oil, and 18 would be double that number as well.

The first batch of oil mixed at 5 grams of iron powder to 5 quarts of oil came in at an ISO 22 rating. The actual numbers were 404 particles per milliliter (larger than 4 microns) for the clean sample and 21,425 particles per milliliter for the first test batch. “That is a massive difference. We really dirtied up the oil with that iron powder,” Speed exclaims.

The sample that had been run through the paper filter for an hour on the dyno, including the six power pulls, came back at a contamination rating of ISO 19 with 3,435 particles per milliliter. So the 25-micron paper filter trapped approximately 85-percent of the contaminants introduced into the oil. “Another data point to keep in mind is that there was zero degradation of power between the first and last runs with the paper filter,” says Speed.

Here are the two filters after the same test protocol. In this case, the paper media appearing a lot dirtier is a good sign.

The wire-screen filter didn’t fare as well in the test results. The base contamination level was 24,000 particles per milliliter, falling within the same ISO 22 rating as the first dirty sample. The post-dyno oil sample, through the wire-mesh filter, came in at a rating of 20,000 particles per milliliter, which is still the same ISO 22 cleanliness rating as the unfiltered contaminated oil. That works out to approximately 17 percent of contaminants filtered out of the oil. To be fair, the micron rating of the wire screen filter was slightly higher than that of the paper filter (35 micron versus 25 micron) but that was only because it was the smallest wire-mesh filter they could get, and the biggest paper filter they could find.

“From an engine standpoint, this is really, really important,” says McAskill. “I’ve been telling people this for 40 years. I used to use a screen filter, and I don’t recommend them simply because of what we’ve seen here today. The difference in filters was only 10 microns, but look at how many more particles larger than 4 microns the mesh filter let through.”

Also worth noting again was the 10 horsepower loss throughout the one-hour testing with the mesh filter due to the wear being imparted on the engine. Obviously, this is an extreme example, but is indicative of the differences you can see between oil filter types. And this also proves that “more expensive” doesn’t necessarily equal better, as the paper filter is usually the cheaper option.