For those who may remember, quite a while back I decided to test some 300Tdi air filters. Well the results are finally in...

Note that these tests were NOT performed according to the ISO standard for air filter testing. The following tests used much more sensitive equipment to measure short-term filtration efficiency and pressure drop of clean and dirty air filter samples. The ISO tests load filters with large amounts of dust and allow gravimetric measurement of filter efficiency.

First some specifics on the equipment and procedure: The tests used;
A collison nebuliser to aerosolise polystyrene latex particles (a common laboratory test aerosol)
A diffusion dryer to remove residual moisture from the particles
Dilution of the aerosol stream with HEPA filtered air

Measurements upstream and downstream of the filter were perfomed using either a TSI 3010 Condensation Particle Counter and a Grimm 1.108 Optical Particle Counter.

Pressure drop across the filter was measured using a TSI Velocicalc Plus.

IN MY OPINION, these tests provide an accurate comparison of pressure drop and filtration efficiency of the filters tested - as good or better than the ISO-standard testing would provide. The only information which this test does not provide is information on filter lifetime (dust capacity).

First the filters:

Left - Right:
Discovery 300Tdi K&N (thanks SSmith)
Defender 300Tdi DONALDSON - used for 20k km on dusty roads (thanks rick130)
Defender 300Tdi Coopers (OEM) NEW (thanks rick130)
Defender 300Tdi K&N - initially supplied with 20k km worth of dust (thanks rick130)


And also a 3.9V8 (RRC) K&N (thanks Hobie18) - you will notice that this is about the same diameter as a 300Tdi filter, but about half the length. Yet a 3.9L V8 would flow A LOT more air than a 2.5L Tdi - what were the engineers thinking...

So now for the pressure drop results - should be self explanatory???


So - the first conclusion we can draw is that V8 owners should switch to a TDI filter!!!

The 2nd conclusion is that K&N filters definitely have lower pressure drop - but at what cost (wait for the efficiency results) - when they are clogged, that would be a different story, but 20k km worth of dust isn't a lot in the scheme of things.

However the most surprising conclusion, was that a Donaldson (cellulose fibre (paper)) filter element with 20k km worth of dust in it, has a LOWER DP than an OEM (Coopers) filter does when NEW!!!

This can be explained by 2 things:
(1) Notice in the picture above, that the Donaldson uses expanded metal for the outer mesh, which has a high open area, whereas the OEM filter uses a drilled plate, which has a lower open area, and even dome closed areas. NOTE - all filters were tested without the plastic cyclone, which may make the DP curves of the 2 filters slightly closer.

(2) The Donaldson has deeper pleats, which means there is more filter material in the element, therefore lower average velocities and less restriction. It also should mean that the donaldson filter has more dust capacity (so your Donaldson element still has life in it rick )

Now onto the efficiency data! The most important part!


The image above shows the filtration efficiency vs particle size for 4 different filters. It can be seen that (a) a "dirty" filter is more efficient than a clean one, and (b) the cellulose-fibre type filters are more efficient than the K&N type of filter. The diffenence between the two filter types may not look significant, but over a few thousand km it will amount to a very large difference in the amount of dust getting into the engine.

If anyone wants any more details of the test setup, feel free to ask.