Actually the title of this story sounds more exciting than the real testing process. Somehow, my working title, "SAE J726 - A uniform testing procedure for direct laboratory performance comparison of dry type and oil bath type air cleaners" just didn't have a snappy ring to it.

Our story actually begins back in May of 1998 when I spent 3 glorious days at Fluid Technologies Inc. (FTI), an independent laboratory in Stillwater, Oklahoma hired by our company (NEUSPEED) to contact back-to-back testing of conventional paper air filters versus synthetic foam filters and pleated, reinforced cotton gauze filters. FTI is one of a handful of test laboratories equipped to perform the Society of Automotive Engineers' J726 test procedure for air filters.

Huh? Automotive engineers? What's up with that? Well, unlike any of the air filter "tests" you may have seen on T.V. commercials which are basically shameless self promotions for the advertiser, the SAE procedure is an industry-recognized standard of testing air filters for effectiveness, restriction and dust holding capacity. Any qualified lab in the world with comparible equipment should be able to perform the J726 test on the same group of filters and come up with the same set of results. Those of you in the U.K. may recognize another test facility whose services we previously engaged: MIRA -- yes, the Formula One wind tunnel place. So, this is no lightweight test procedure.

How Does the Test Work?

For the record, the testing rig takes up most of the space in a room that's approximately 40 feet long by 20 feet wide. My hand sketch below makes the equipment appear much smaller than its actual size.

The ideal filter should trap a large amount of dirt over an extended period of time without creating an excessive pressure differential between the inlet side and outlet side of the airbox. Notice that the measurement of horsepower is not factored into this test. That's because the preservation of horsepower is secondary in importance to the collection of dirt.

During my 3 days at FTI, we tested filters of various sizes and shapes. However, the test results I am showing you here belong to an automotive panel filter similar, but not identical, in size and shape to the 2-valve 900SS/M900 panel filter. Our paper filters were manufactured by Fram, a division of Bendix. Our cotton gauze filters were from K&N. Finally, our three-stage foam filters came from Ramair, a British company that has since gone bankrupt. A comparible motorcycle product to the Ramair foam filter would be a multi-stage ITG filter. Here's a picture of each filter type as scanned from an old Ramair technical paper.

As you would expect, the Fram paper filter finished third. It captured 119.68 grams of dirt over an 11 minute run before the filter became clogged and registered the pressure differential change of 6 inches of water. The starting differential pressure was 17.84 inches of water, and the ending differential was 23.84 inches of water. Was this a bad filter? No. On the basis of dirt capturing ability, this filter was 95.62% efficient. However, compared to the others it clogged up pretty fast.

The K&N cotton gauze filter finished second. It captured 140.08 grams of dirt over a 22.8 minute run before the filter clogged up and registered the pressure differential change of 6 inches of water. The starting differential pressure was 16.80 inches of water and the ending pressure differential was 22.80 inches of water. On the basis of dirt capturing ability, the K&N filter was 98.89% efficient. The K&N filter was able to run for over double the test time as the Fram paper filter before it became clogged. So, K&N is making a valid claim when they tell you that the filter does not need to be cleaned very frequently. In addition, it captured significantly more dirt than the Fram paper filter. Finally, when the testing rig's air pump was first started, the clean K&N filter only caused a 16.8 inch water pressure drop across the two sides of the airbox. The clean Fram filter caused a 17.84 inch water pressure drop. Translation: the K&N filter was less restrictive than the Fram paper filter both when it was new, and even later when it became dirty.

The RamAir multi-stage foam filter finished first in dirt holding holding capacity. It captured a whopping 199.81 grams of dirt over a 24.21 minute run before the filter clogged up and registered the pressure differential change of 6 inches of water. The starting differential pressure was 18.21 inches of water and the ending differential was 24.21 inches of water. On the basis of dirt capturing ability, the Ramair filter was 99.69% efficient. However, to score a victory in the dirt capturing contest, the RamAir filter was slightly more restrictive than both the Fram paper filter and the K&N cotton gauze filter. Notice that the starting differential pressure across the airbox was higher with the RamAir than either the Fram and K&N. Translation: the RamAir multi-stage foam filter creates a little more restriction, but delivers much higher dirt capturing results.

What does this mean to you? If you are choosing between a multistage foam filter (Uni, ITG, etc) and a K&N filter, there's no clear winner. The multi-stage foam filters have the ability to capture more dirt overall, but may cause slightly more initial restriction during the early period of use. In contrast, with the K&N you get superior airflow ability, but will have to settle with slightly lower dirt holding capacity. Hey, just clean it more frequently then! Nonetheless, they both outperform the stock paper filters found on the 2V bikes, so the faster you get rid of the cheesy paper filter, the better off you are.