Known best filter for subie FA engines?

Me, since I don’t do dumb things when the engine is cold, and therefore, by the time the engine is warm it has been filtered a couple hundred times on every startup
The example I gave with warm oil at high rpm is only one example of a worst-case scenario for bypassing. With cold oil, engine rpm doesn't need to be nearly as high before bypassing occurs. The FRAM Extraguard for example would likely be bypassing a lot of oil continuously even at cold idle. The figures I came up with are also based on brand new filters with no dP increase from dirt loading.

I’ve also read probably every single Subaru filter thread (after the third one, out of morbid curiosity alone; how board members can continually have this discussion at least 3x/month and kschachn only occasionally posts all the relevant threads) and I understand and agree with the underlying physics. BUT, aftermarket filters probably outsell OEM filters at 15:1 or better. That would imply tens of thousands of worn out/failed engines if your claims applied universally.
Subaru engines don't get worn out prematurely even with aftermarket filters because a filter bypassing on a regular basis should normally only cause a very small increase in engine wear. However, in certain conditions, the bypassing might be problematic.

Cleaner oil 99%+ of the time is going to outweigh a temporary circulation of particles while the bypass is open, and as soon as it closes, those particles are then captured by the more efficient filter.
I agree and I think the FRAM Endurance will result in less engine wear than an OEM filter even when if the FRAM is bypassing often, but only in normal conditions when there isn't a lot of large abrasives in the oil. I've seen multiple teardowns of failed Subaru engines that were using cheap aftermarket filters, and most of the wear surfaces throughout engines had deep scoring from large bits of bearing material circulating through the engine. A bypassing filter will allow orders of magnitude more of these large bits to circulate, and it's the large stuff that causes most wear in these conditions, and when wear rates are already really high.

In this specific example, if the engine is going to be replaced with a new engine anyway instead of being rebuilt, I supose it doesn't really matter, but in some cases it might make the difference between replacing a failed engine component and replacing a long block.

The other concern I have with bypassing filters is clogging of oil screens and oil control valves. This is a common issue on Subarus and I suspect oil filter bypassing is a contributing factor. A filter that isn't bypassing will capture nearly 100% of any chunks of carbon large enough to clog an oil screen, whereas a bypassing filter will let a lot it through.

The risks of filter bypassing may be low, but they need to be weighed against the benefits of more efficient filtration, which are small. Car manufacturers seem to prioritize bypass resistance, but do not seem to prioritize filtration efficiency. This might be because normal abrasive wear will never cause issues within the warranty period, whereas a bypassing filter could cause warranty claims or make the fix more expensive.

All that said, I have an OG FRAM Ultra on my shelf that I will be using on my WRX, but I feel that if there's a filter like the Purolator available that won't bypass and isn't much less efficient, it's a better option. If I had a naturally aspirated Subaru like yours, I might be inclined to use a FRAM Endurance, as it shouldn't be bypassing all that often.
 
The example I gave with warm oil at high rpm is only one example of a worst-case scenario for bypassing. With cold oil, engine rpm doesn't need to be nearly as high before bypassing occurs. The FRAM Extraguard for example would likely be bypassing a lot of oil continuously even at cold idle. The figures I came up with are also based on brand new filters with no dP increase from dirt loading.
Hard to say when the pump only puts out around 2 GPM at 600 RPM. Might be around 3.2 GPM at cold idle, say 1200 RPM. The oil would have to be pretty cold and thick. Even the OEM Subaru filter might be in bypass if the oil is that cold. If the oil is very thick, the pump will hit relief at pretty low RPM too.

BR showed 8.4 PSI dP on the Endurance at 3 GPM with 500 cSt oil. And 11.1 PSI dP at 11 GPM with 15 cSt oil. Anyone interested in the OEM Subaru filters should contact BR and ask them to test some OEM Subaru filters at least for just the flow vs dP testing to see how they do. Could be the bypass on the OEM Subaru filter is set relatively high because they have a relatively high flow vs dP curve.
 
  • Like
Reactions: twX
ah ha........Subie means Subaru.....learn something everyday


does oil spew out around filter at oil change?
 
Hard to say when the pump only puts out around 2 GPM at 600 RPM. Might be around 3.2 GPM at cold idle, say 1200 RPM. The oil would have to be pretty cold and thick. Even the OEM Subaru filter might be in bypass if the oil is that cold. If the oil is very thick, the pump will hit relief at pretty low RPM too.
Cold idle on this engine is 1,800 rpm, so pump output would be 5.9 GPM if it increases linearly with rpm at low revs. Of course, the oil pump will be in pressure relief if the oil is thick enough, but thicker oil also increases filter dP at a given flow rate.

What my model shows for 1,800 rpm with the oil just thick enough for the oil pump to hit its relief pressure (~40 cST), is a dP of 13.5 psi for the PH9688, down from 22 psi with warm oil at high rpm with the same oil pressure, though it's still above the bypass pressure, and dP would still increase with rpm due to the limited effectiveness of the pump relief. However, this model is not going to be as accurate when the oil is much different than the oil filter test viscosities of 13.5-24 cST.

The Brand Ranks data is interesting. I'd estimate that the oil pump on this engine would be flowing <<3 GPM even at higher rpm when the oil is 500 cST, which would mean that none of the filters tested would be bypassing when the oil is this thick. However, the flow meter that they used isn't rated for oil much thicker than 100 cST, so the flow rates may not have been very accurate. I've seen some other data from engine tests that indicate that filter dP on a running engine is higher when the oil is very thick. It could perhaps go either way, depending on the engine and filter.
 
does oil spew out around filter at oil change?
If you loosen the top mounted filter just enough to rock it back and forth a little, give it a minute or two and the oil will drain without flooding the little containment dish around the filter. Then remove filter the rest of the way. Pet peeve is seeing someone's car thousands of miles into its next oci with old oil still stagnant in that collection area from the previous oil change.
 
If you loosen the top mounted filter just enough to rock it back and forth a little, give it a minute or two and the oil will drain without flooding the little containment dish around the filter. Then remove filter the rest of the way.
Punching a hole in the dome end of the filter also works well if the filter is mounted base down. Been doing that trick for decades.
 
Punching a hole in the dome end of the filter also works well if the filter is mounted base down. Been doing that trick for decades.
You don't worry about the filter housing marweial going in to the engine?
 
You don't worry about the filter housing marweial going in to the engine?
To be clear, I'm talking about spin-on filters that are mounted base down. Punching a small hole in the dome end with a sharp awl or nail won't put any material in the engine. The whole filter can/shell is on the dirty side of the media, and a punched hole will just form a hole without breaking any metal off anyway.
 
To be clear, I'm talking about spin-on filters that are mounted base down. Punching a small hole in the dome end with a sharp awl or nail won't put any material in the engine. The whole filter can/shell is on the dirty side of the media, and a punched hole will just form a hole without breaking any metal off anyway.
Ah sorry. I misunderstood
 
The Brand Ranks data is interesting. I'd estimate that the oil pump on this engine would be flowing <<3 GPM even at higher rpm when the oil is 500 cST, which would mean that none of the filters tested would be bypassing when the oil is this thick. However, the flow meter that they used isn't rated for oil much thicker than 100 cST, so the flow rates may not have been very accurate. I've seen some other data from engine tests that indicate that filter dP on a running engine is higher when the oil is very thick. It could perhaps go either way, depending on the engine and filter.
If BR at least did the cold and hot oil flow vs dP tests on an OEM Subaru filter it would give a good indication how it "ranks" against others since they were all tested in the same test rig. It would be interesting to instrument a Subaru engine to measure the actual oil flow rate vs RPM with different oil temps/viscosity. Or just measure the actual dP across the filter as a few members here have done on their vehicles. The Subaru RPM vs P curves you posted earlier show that the flow going to the engine is being cut back pretty good from either pump efficiency loss (slip) and/or the pump relieving some flow as RPM increases.
 
I had an oil change done on a family member's Crosstrek last week at the local dealership, and they installed a Tokyo Roki filter. These were hard to find back when I owned a Subaru.
 
If BR at least did the cold and hot oil flow vs dP tests on an OEM Subaru filter it would give a good indication how it "ranks" against others since they were all tested in the same test rig.
Yeah, I'd like to see that, but it's a safe bet that the OEM Subaru filters have low restriction. They have more media area than any aftermarket filter, and more than any filter that BR tested as well, aside from the OEM Toyota.
 
Yeah, I'd like to see that, but it's a safe bet that the OEM Subaru filters have low restriction. They have more media area than any aftermarket filter, and more than any filter that BR tested as well, aside from the OEM Toyota.
More media area certainly helps reduce dP, with all other factors constant. The PureOne PL4006 that Purolator did a flow vs dP test on many years ago had ~105 sq-in of media (might be different today), and its flow vs dP curve was pretty low. But a flow vs dP test is the only way to really know the flow performace of the filter.

On a side note - I've tracked the media area on the Motorcraft FL500S over the years, and in 2016 it was 227 sq-in, and in 2018 it went down to 120 sq-in, due to the pleat depth going from 0.75 in to 0.375 in (center tube was larger diameter). That's a huge reduction in media area.
 
Back
Top