Subaru 15208AA160 and Mazda N3R1-14-302 cut open and compared

Joined
Jan 16, 2021
Messages
159
Most of you will probably already be aware that the Tokyo Roki 15208AA160 "black" filter that comes on Japan-built Subarus is not available in the USA. US Subaru dealers replace it with 15208AA15A "blue", made by Fram, at the first oil change. It's widely understood that the Roki filter is superior. Subaru fans have been buying Mazda N3R1-14-302 Tokyo Roki filters, intended for the RX-8, in the belief that they are the same filter as the Subaru 15208AA160. But are they? I decided to find out.

Here we have a disassembled Subaru black next to a disassembled Mazda filter.

20220409_125629.jpg


I don't have a fancy-pants filter cutter like the rest of you experienced fanatics; I'm an inexperienced fanatic, so I opened these with an angle grinder.

They certainly look identical at first glance. The external dimensions are the same; the baseplate is the same. Internal construction is the same, with the same anti-drainback valve, and the same steel spring plate holding the filter cartridge down. The filter cartridge is the same length and diameter, and contains the same number of pleats (55), joined with a metal clip.

The bypass valve is spot-welded into the filter end cap:

20220409_125727.jpg


So far so good. But we know that Subaru guys are laser-focused on the bypass valve pressure. How to compare that? Hmm.

I put each filter cartridge in a cup on a scale and zeroed it:

20220409_123748.jpg


I put a piece of blue tape on a driver:

20220409_123838.jpg


Then I pressed down the bypass valve with the driver until the tape touched the end cap, and noted how much force the scale showed. I don't have pictures of this, because I only have two hands. You'll have to take my word for it. The Subaru filter took 5lb 6.2oz of pressure to open the valve to the tape, and the Mazda filter took 5lb 5.7oz of pressure. I would consider those results within the range of measurement error or manufacturing tolerance. Looks to me like it is the same valve. (Please note that I am not saying these are ~5-6psi valves. These numbers don't mean anything except for comparison to each other.)

So are they the same? Well, there is one difference. The center core is different. Here's the Subaru:

20220409_125639.jpg


And the Mazda:

20220409_125641.jpg


I didn't disassemble them to count the holes, but my eyeball estimate says the Subaru filter has about 50% more holes than the Mazda. What I don't know is whether that makes any difference. I took a look at a few other filters I have on hand (Champ, Wix, Baldwin, Purolator Tech, and Subaru blue) and the perforations in the core vary, but none of them have as many holes as the black Subaru pictured here. So I'm going to guess it doesn't matter, but I'm sure one of you guys will set me straight if I'm wrong.

So, assuming the core perforations are a difference that makes no difference, then my conclusion is that the Mazda N3R1-14-302 is functionally the same as the Subaru 15208AA160, is built to the same quality, and is a good choice to replace the USA-unavailable Subaru filter if you want the closest thing to what the engine was built with.

Just for a fun comparison, here's a Subaru 15208AA15A "blue" filter cut open:

20220409_130045.jpg


That's a Fram or I'll eat dirt. The less said about it the better.

Comments are welcome.
 
Good comparison ... but:

What's the efficiency and delta-p vs flow of the black Tokyo Roki filter compared to the OEM Fram filter?

What's the efficiency and delta-p vs flow of the black Tokyo Roki filter compared to the Mazda filter?

Are the bypass valves on the Tokyo Roki and Mazda filter the same diameter - it looks like it, but were they measured to verify?

What data in the Subaru community concludes that the OEM blue Fram is inferior? Actual measured data, or just Fram hate? :unsure: Realize that the Fram built filter is per Subaru specifications, not Fram's.

You should do a "burn test" on the ADBVs to verify if they are nitrile or possibly "black silicone".

 
Good comparison ... but:

What's the efficiency and delta-p vs flow of the black Tokyo Roki filter compared to the OEM Fram filter?

What's the efficiency and delta-p vs flow of the black Tokyo Roki filter compared to the Mazda filter?
If you can think of a way I could have measured that with an angle grinder, a scale, and a stick with blue tape on it, I'm all ears. :D
 
Are the bypass valves on the Tokyo Roki and Mazda filter the same diameter - it looks like it, but were they measured to verify?
They are. Aside from the center cores, they appear entirely identical.

What data in the Subaru community concludes that the OEM blue Fram is inferior? Actual measured data, or just Fram hate? :unsure: Realize that the Fram built filter is per Subaru specifications, not Fram's.
Well, just look at them.

If the Fram filter meets Subaru's minimum specifications, all well and good; but the Roki is still better constructed. Certainly a case can be made for being better than the bare minimum.

You should do a "burn test" on the ADBVs to verify if they are nitrile or possibly "black silicone".
Maybe next time. I believe I have reached the end of my wife's patience for the moment.
 
They all look very good. The blue Subie filter looks better that a standard orange Fram. Looks closer to a tough guard to me minus the silicon ADBV..
 
If you can think of a way I could have measured that with an angle grinder, a scale, and a stick with blue tape on it, I'm all ears. :D
Wasn't saying you should do that, but rather what is the efficiency data of all those filters? If one was way more efficient than the others, would you want to use the efficient filter or one with worse filtration efficiency. After all, the job of a filter is to filter well. "Judging" filters solely on their construction is secondary, especially if the construction really doesn't pose any problems. Now if you say failures in one filter over the other, like torn media or broken eCore center tubes or cut ADBVs, then that certainly should be taken into consideration.

How many falure issues have the Subaru guys seen on the OEM blue filter ... like torn media, etc?
 
Well, just look at them.

If the Fram filter meets Subaru's minimum specifications, all well and good; but the Roki is still better constructed. Certainly a case can be made for being better than the bare minimum.
Like said above, judging oil filters just by how they look or how they are constructed is not the whole story. Kind of like lots of men initially judge their girlfriends just by "looks", and then later find out that there was so much more to it all than just that. Can't judge a book by it's cover as they say. If you were here for 10+ years gleaning info you would have seen many filters that look worse than the Fram built filters do after use. 😄
 
Wasn't saying you should do that, but rather what is the efficiency data of all those filters? If one was way more efficient than the others, would you want to use the efficient filter or one with worse filtration efficiency. After all, the job of a filter is to filter well. "Judging" filters solely on their construction is secondary, especially if the construction really doesn't pose any problems. Now if you say failures in one filter over the other, like torn media or broken eCore center tubes or cut ADBVs, then that certainly should be taken into consideration.

How many falure issues have the Subaru guys seen on the OEM blue filter ... like torn media, etc?
I don't know that there's hard data on that. Only Subaru would know the failure rate of the blue Subaru filters, and they don't publish that data. Everything else is anecdotal.

You don't have to look too hard to find people on this site who have had failures with Fram filters of this design, and it's reasonable to suspect that such failures do happen with the blue Subaru filters; it's not a question of "if" but "how often." Probably the answer is "not often enough to justify the cost of improvement."

You strike me as someone who knows there's no such thing as a solution, only a tradeoff. I don't doubt they're made to Subaru's specifications, but those specifications were almost certainly written with a cost/benefit analysis that said, "If we save a total of X on this component by accepting a failure rate of Y, then as long as the cost of replacing Y engines under warranty is less than the savings X, we're ahead." So overspec'ing the filter is not in their best interest even if it does cost them the occasional warranty claim, and there's the tradeoff.

Can I prove that? Not without access to Subaru's confidential data; but it seems at least plausible given what we see during the teardown.

The car owner's incentives are different. You probably also know what marginal utility is. He only has to worry about his own car, and the small amount of additional fail-safe provided by a higher quality filter may be worth it to him.

All that said, I'm going to stop using these Mazdas. They're too expensive. :) I'm not going to switch to the blue Subaru filters, but I'm sure Wix, Champ, or Purolator are fine.
 
Like said above, judging oil filters just by how they look or how they are constructed is not the whole story. Kind of like lots of men initially judge their girlfriends just by "looks", and then later find out that there was so much more to it all than just that.
Yeah, unexpected things happen when you open up their cans. :D
 
Great information. Thanks for sharing.

I've heard comments for some time that the Mazda filter can substitute the Subaru. But I'll probably stay with the Napa Gold. It has served me well.
 
I don't know that there's hard data on that. Only Subaru would know the failure rate of the blue Subaru filters, and they don't publish that data. Everything else is anecdotal.

You don't have to look too hard to find people on this site who have had failures with Fram filters of this design, and it's reasonable to suspect that such failures do happen with the blue Subaru filters; it's not a question of "if" but "how often." Probably the answer is "not often enough to justify the cost of improvement."
I've been here long enough to have a decent feel for what filter brands have more overall reports of failures, and frankly over the last 10 years Fram is not at the top of the list. I don't think I've ever seen anyone post up a Subaru OEM Fram made blue filter with any failures on this site.

How many Subaru OEM blue filters with failures have been posted on the various Subaru chat boards?

You strike me as someone who knows there's no such thing as a solution, only a tradeoff. I don't doubt they're made to Subaru's specifications, but those specifications were almost certainly written with a cost/benefit analysis that said, "If we save a total of X on this component by accepting a failure rate of Y, then as long as the cost of replacing Y engines under warranty is less than the savings X, we're ahead." So overspec'ing the filter is not in their best interest even if it does cost them the occasional warranty claim, and there's the tradeoff.

Can I prove that? Not without access to Subaru's confidential data; but it seems at least plausible given what we see during the teardown.

The car owner's incentives are different. You probably also know what marginal utility is. He only has to worry about his own car, and the small amount of additional fail-safe provided by a higher quality filter may be worth it to him.
All speculation. I could also "speculate" that the Subarus OEM blue filter could also be more efficient at filtering than the Tokyo Roki or Mazda filter, because if it's using say the Fram EG type filter media than it could be at least 95% @ 20u. Of course, we would need an ISO 4548-12 efficiency test on all of them to know for sure ... but I could still "speculate".

All that said, I'm going to stop using these Mazdas. They're too expensive. :) I'm not going to switch to the blue Subaru filters, but I'm sure Wix, Champ, or Purolator are fine.
LoL ... I guess you haven't seen the various failures of the WIX and Purolator filters posted here. Good luck. ;)😄
 
I've been here long enough to have a decent feel for what filter brands have more overall reports of failures, and frankly over the last 10 years Fram is not at the top of the list. I don't think I've ever seen anyone post up a Subaru OEM Fram made blue filter with any failures on this site.

How many Subaru OEM blue filters with failures have been posted on the various Subaru chat boards?
Don't know. I'm only on one, and not that active there. Not sure anyone knows, and even if they did, I'm not sure that data would be a representative sample.
All speculation. I could also "speculate" that the Subarus OEM blue filter could also be more efficient at filtering than the Tokyo Roki or Mazda filter, because it's using say the Fram EG type filter media than it could be at least 95% @ 20u. Of course, we would need an ISO 4548-12 efficency test on all of them to know for sure ... but I could still "speculate".
Absolutely granted it is all speculation, but since the manufacturers are not forthcoming with hard data, we don't have much else other than speculation and anecdotal evidence. We do what we can with incomplete information.

Say, you strike me as knowledgable enough about filter design to answer that question about the center core holes, not necessarily for the specific filters I disassembled, but at least in the abstract. It seems reasonable to hypothesize that the total area of the holes in the core should be at least the same area as the inlet holes in the baseplate, and it would make sense that you would actually want quite a bit more than that to make up for holes that might be blocked by pleat folds or clogged media. But, per the discussion of pre- and post-filter pressure we had in the other thread, it seems like beyond a certain point, you'd reach a number and size of holes that would provide maximum possible flow, or close to it, and opening it up any further would just weaken the components with little or no performance benefit. Is that right?
 
Don't know. I'm only on one, and not that active there. Not sure anyone knows, and even if they did, I'm not sure that data would be a representative sample.

Absolutely granted it is all speculation, but since the manufacturers are not forthcoming with hard data, we don't have much else other than speculation and anecdotal evidence. We do what we can with incomplete information.
Well, like said earlier the Fram filters with non-metal end caps really don't have many failures - and I highly doubt the Subaru OEM blue filter does either. If you can find a rash of Subaru OEM blue reported failures I'd like to see them. You might see some wavy media (but not torn) or slightly deformed end caps, but those things really don't make it fail from doing it's job. But we have seen plenty of filters with metal end caps with wavy and also torn media.

Say, you strike me as knowledgable enough about filter design to answer that question about the center core holes, not necessarily for the specific filters I disassembled, but at least in the abstract. It seems reasonable to hypothesize that the total area of the holes in the core should be at least the same area as the inlet holes in the baseplate, and it would make sense that you would actually want quite a bit more than that to make up for holes that might be blocked by pleat folds or clogged media.
I don't think any media can actually "block" a hole or louver in a center tube "liquid tight" to actually cut off any oil flow. Besides, the oil must go through the media before it can go through a center tube hole - so even if the media pleat was up tight on a hole, the oil still goes through the media and then the hole ... so it's a non-concern. Remember, the PD oil pump is going to force the same volume of oil through the system as long as the flow resistance of the whole system (engine + filter) doesn't put the pump in pressure relief. It's actually pretty hard to put a PD oil pump in to pressure relief. A totally clogged oil filter with an undersized bypass valve might do it, but it the filter is that clogged then might be having some other major issues going on.

If you measure the center tube holes total area, and the base plate holes total area, both of those should be at least equal or greater than the area of the filter mounting gallery inlet hole - which is smaller than the threaded hole in the filter base. I think you'll find that any filter will meet that.

But, per the discussion of pre- and post-filter pressure we had in the other thread, it seems like beyond a certain point, you'd reach a number and size of holes that would provide maximum possible flow, or close to it, and opening it up any further would just weaken the components with little or no performance benefit. Is that right?
Again, it's not really about "flow" ... it's about the total delta-p across the filter as a function of oil flow. Look at the delta-p vs flow of the filters tested in this thread (link below). Yes, the center tube still needs to be strong enough to take max expected delta-p with a safety factor and still not add much delta-p to the flow restriction of the filter assembly.

Read from where this link pops you into the thread to get some insight to delta-p vs flow and efficiency testing per ISO 4545-12 (the current standard since year 2000).

 
Last edited:
Again, it's not really about "flow" ... it's about the total delta-p across the filter as a function of oil flow. Look at the delta-p vs flow of the filters tested in this thread (link below). Yes, the center tube still needs to be strong enough to take max expected delta-p with a safety factor and still not add much delta-p to the flow restriction of the filter assembly.
Okay, so, put another way, a filter with more restriction will have more delta-p and the oil pump will have to work harder to overcome it and move the required amount of oil; and a filter with less restriction will make it easier for the pump to do its job.

So let me rephrase the question in terms of delta-p. A center tube with insufficient porting would certainly increase delta-p. Adding more holes should reduce it. Does that stay true as far as you can physically take it? Or is it a case of diminishing returns, such that once you pass a certain amount of perforation, the additional reduction in delta-p from adding more perforations is zero or near-zero?

Put more colorfully, is there an amount of perforation where the center tube becomes basically invisible in the delta-p calculation?

Regarding holes being blocked by folds in the pleats, I didn't intend to say that they must be completely blocked. Perhaps "obstructed" would have been a better word. But as to this:

...even if the media pleat was up tight on a hole, the oil still goes through the media and then the hole.

I've always imagined filter media to be compressed at the fold point, and therefore more resistant to flow through the folded points than the unfolded areas between them. Is that not so? I picture basically this tight wad of compressed fiber, pushed together by being folded and then jammed up against one of the holes. I'd expect less flow through that hole than one that had nothing pushing up against it. That was what I had in mind; a restriction, not a complete plug.
 
Okay, so, put another way, a filter with more restriction will have more delta-p and the oil pump will have to work harder to overcome it and move the required amount of oil; and a filter with less restriction will make it easier for the pump to do its job.

So let me rephrase the question in terms of delta-p. A center tube with insufficient porting would certainly increase delta-p. Adding more holes should reduce it. Does that stay true as far as you can physically take it? Or is it a case of diminishing returns, such that once you pass a certain amount of perforation, the additional reduction in delta-p from adding more perforations is zero or near-zero?

Put more colorfully, is there an amount of perforation where the center tube becomes basically invisible in the delta-p calculation?
The base plate holes, and the center tube holes don't add much delta-p to the overall delta-p across the entire oil filter. Look at the delta-p vs flow rate data in the link in post #16 above. If the total delta-p across the oil filter is only 3 PSI when the oil flow rate is 5 GPM, then the delta-p from the base inlet and center tube holes is probably about around 0.5 PSI total. The filter media is by far the main restriction in an oil filter. It's really not anything to worry about unless you find a filter with louvers that are not opened up any more than a small slit - that's a whole other subject, lol.

If you wanted to feel better about the center tube not "restricting flow", then use filtes with an eCore center tube. But it's really hair splitting if the center tube + base plate gives 0.5 PSI or 0.3 PSI of delta-p.

Regarding holes being blocked by folds in the pleats, I didn't intend to say that they must be completely blocked. Perhaps "obstructed" would have been a better word.

I've always imagined filter media to be compressed at the fold point, and therefore more resistant to flow through the folded points than the unfolded areas between them. Is that not so? I picture basically this tight wad of compressed fiber, pushed together by being folded and then jammed up against one of the holes. I'd expect less flow through that hole than one that had nothing pushing up against it. That was what I had in mind; a restriction, not a complete plug.
The media is probably more dense at the pleat folds, but again even if a pleat fold was aligned right over a center tube hole I highly doubt it's going to block much flow through that hole. The oil is being forced through the filter, and it will find a way to all the holes in the center tube. If a hole flow it somehow reduced, then some other hole(s) will make it up by flowing more. What oil leaves the PD pump goes through the filter, so this to is something else not worth worrying about.
 
Last edited:
The media is probably more dense at the pleat folds, but again even if a pleat fold was aligned right over a center tube hole I highly doubt it's going to block much flow through that hole. The oil is being forced through the filter, and it will find a way to all the holes in the center tube. If a hole flow it somehow reduced, then some other hole(s) will make it up by flowing more. What oil leaves the PD pump goes through the filter, so this to is something else not worth worrying about.
Not worrying, just wanting to understand better. :) Always interested in learning something.
 
It's the same thing with Honda and Tokyo Roki and afterwards using the fram filters.
They also prefer the Tokyo Roki ,and now they are using anything that's available (purflux,Mahle tennex,...)
They look identical actually.
Whats the bypass pressure rating for the Subaru ?I just recently noticed the RX8 filter cross referenced with the s2000 filter and that was rated at 1bar (16psi).
 
Back
Top