Best brand of oil filters

I have said this for years ...

Filtering is very important, until it's not. Meaning once it's good enough, being better doesn't really matter nearly as much.

The goal should be to have a reasonably clean sump. However, the oil does not need to be clinically clean. If you have a filter which is 95% or better at 20um, that's more than good enough to make any improvements intangible. You can use a 99% filter, but there's no conclusive data from real world testing that indicates the gain in filter efficiency makes a real difference in wear control in one's garage.

Sure; I understand that lab studies have shown cleaner oil is better for the engine. I doubt anyone here has any disagreement with that. But what studies exist to determine a breakpoint for the ROI? Someone please point to a study which indicates, with conclusive statistically viable data, that filtration efficiency has a breakpoint for equipment lifecycle.

The daily variation of life, resulting in expected variation in"normal" wear, far exceeds the control a 99% filter can inflect into the equipment lifecycle. Do not forget that the lube filter is only one parameter of equipment lubrication wear control. Also consider these:
- OCI duration
- lubricant selection
- environment
- operational conditions
- air filtration (in the case of ICE applications)

We've seen countless examples of Honda and Toyota engines running 200k, 300k and more, all using OE fluids and filters. And let's be honest, those brands don't exactly set the highest bar when it comes to oils and filter efficiency.

Does filtration matter? Absolutely yes.
Does having the "best" filtration matter? Absolutely not.

In today's filter market, I actually would place a premium on construction and build quality over efficiency. The examples of poor quality have been astonishingly horrible from many brands of late. I'd take a well made 90% filter over a poorly made 99% filter every day of the week.
 
Last edited:
I have said this for years ...

Filtering is very important, until it's not. Meaning once it's good enough, being better doesn't really matter nearly as much.

The goal should be to have a reasonably clean sump. However, the oil does not need to be clinically clean. If you have a filter which is 95% or better at 20um, that's more than good enough to make any improvements intangible. You can use a 99% filter, but there's no conclusive data from real world testing that indicates the gain in filter efficiency makes a real difference in wear control in one's garage.

Sure; I understand that lab studies have shown cleaner oil is better for the engine. I doubt anyone here has any disagreement with that. But what studies exist to determine a breakpoint for the ROI? Someone please point to a study which indicates, with conclusive statistically viable data, that filtration efficiency has a breakpoint for equipment lifecycle.

The daily variation of life, resulting in expected variation in"normal" wear, far exceeds the control a 99% filter can inflect into the equipment lifecycle. Do not forget that the lube filter is only one parameter of equipment lubrication wear control. Also consider these:
- OCI duration
- lubricant selection
- environment
- operational conditions
- air filtration (in the case of ICE applications)

We've seen countless examples of Honda and Toyota engines running 200k, 300k and more, all using OE fluids and filters. And let's be honest, those brands don't exactly set the highest bar when it comes to oils and filter efficiency.

Does filtration matter? Absolutely yes.
Does having the "best" filtration matter? Absolutely not.

In today's filter market, I actually would place a premium on construction and build quality over efficiency. The examples of poor quality have been astonishingly horrible from many brands of late. I'd take a well made 90% filter over a poorly made 99% filter every day of the week.
I think I tried to write that. You did it so much better.

“Buy a good filter- don’t fret over small differences in filtration.”
 
At what point does an oil filter ..... filter so tiny and so much that it simply starts to just bypass?
When the media is so full of contaminates that it can no longer filter.

My understanding is a filter becomes "more efficient" as the media starts to load up... until it's not.
 
Amazing you don’t even understand or acknowledge what I wrote. Interesting really.- I never said “smoked or totaled” for example

Maybe I need to repeat and further simplify. Not sure.

Buy a good filter- don’t fret over small differences in filtration.
Post 15: "Oil filter choice will have zero bearing on the life of your car."
I replied "Depends on the efficiency of said "choice" ... that's the point."

Then you went down this typical rabbit hole in post 23.
"What I am saying, is the car as a whole will be long gone for some other reason, a reason other than oil filter choice alone, before the small difference in wear from oil filter choice causes the engine to grind down to nothingness."

So now you're basically saying oil filter efficiency doesn't really matter because the "whole car will be long gone for some other reason". Never did say what oil filter(s) you use on your own vehicles.

I'm surprised nobody posted the typical "The air filter is the most important filter" (of course it is) ... like that means people should ignore efficiency on the only filter left to keep the oil clean because oil cleanliness "doesn't matter", lol.

The debate is if a more efficient oil filter keeps oil cleaner and reduces engine wear, regardless of some other excuse for not caring about oil filter efficiency. The OP did focus on efficiency - something he was aware of and focused on. And like said before, over the long run a more efficient filter certainly could have a bearing on the running health and life of the engine.

A lot of people see some value in using high efficiency filters. Those who don't think it makes any difference or don't care about it seem to have the "it's going to get totaled before the engine wears out" mentality (my definition) - which is basically a binary viewpoint. Same with motor oil, if the engine doesn't fail or blow-up then the oil must be working pretty good (even though their oil choice may cause more wear over the long run) - that's a whole other debate. That kind viewpoint is fine since it's not my machine - but it makes for a good technical debate. I try to maximize wear reduction, and I do keep my stuff a very long time which drives that goal.

Everyone can do their own engine wear vs oil cleanliness self education because there is a lot of info available. Cummins did a lot of wear studies, and just like the conclusion in every engine wear study it shows that cleaner oil results in less wear with all other factors held constant. In fact, one of their studies found that just going from 99% @ 60u to a 99% @ 40u filter reduced ring and bearing wear pretty noticeably (using radioactive tracers in a real fired engine). They also added a bypass filter to the 99% @ 60u filter and it brought the wear down even more of course, why wouldn't it. And 0.1 gram per quart isn't what I'd call some crazy accelerated test. That would only be 4 grams of test dust in 10 quarts of oil ... enough to see the wear results based on oil filter efficiency and oil cleanliness.

1754873020504.webp
 
Last edited:
When the media is so full of contaminates that it can no longer filter.

My understanding is a filter becomes "more efficient" as the media starts to load up... until it's not.
Go read the big Ascent ISO 4548-12 testing thread - jump to post 368 and read from there. Shows that oil filers loose efficiency, some quite a bit of efficiency, as they load up. They may then gain some efficiency back right before they totally clog, but filters should never be ran long enough to cause near total clogging.
 
The goal should be to have a reasonably clean sump. However, the oil does not need to be clinically clean. If you have a filter which is 95% or better at 20um, that's more than good enough to make any improvements intangible. You can use a 99% filter, but there's no conclusive data from real world testing that indicates the gain in filter efficiency makes a real difference in wear control in one's garage.
That's because nobody here has the time and coin to conduct a well controlled test program on a real world test vehicle. It would have to be a dedicated vehicle that got a lot of use or it would take a long time to get the data. But there have been some real world test studies done, and again the conclusion is always that cleaner oil resulted in less wear ... regardless of how much reduction. I'll take even a small reduction since it really doesn't cost any more or maybe a few bucks more to achieve that goal. Same with oil viscosity ... I can bump up to a xW-30 from a xW-20 for no real added cost ... just use a different viscosity.

Sure; I understand that lab studies have shown cleaner oil is better for the engine. I doubt anyone here has any disagreement with that. But what studies exist to determine a breakpoint for the ROI? Someone please point to a study which indicates, with conclusive statistically viable data, that filtration efficiency has a breakpoint for equipment lifecycle.
Well, ROI isn't a thing most people maintaining there own personal vehicle is really concerned about. For a big fleet of vehicles used in a business, most likely. I'm certainly not going to use a filter that' 99% @ 40u when I can get a nice filter that's 99% @ 20u for the same price or a few bucks more. To me saving a few bucks isn't as important than knowing the oil is kept cleaner over the OCI.
 
Last edited:
That's because nobody here has the time and coin to conduct a well controlled test program on a real world test. But there have been some real world test studies done, and again the conclusion is always that cleaner oil resulted in less wear ... regardless of how much reduction.


Well, ROI isn't a thing most people maintaining there own personal vehicle is really concerned about. For a biPLXL6607g fleet of vehicles used in a business, most likely. I'm certainly not going to use a filter that' 99% @ 40u when I can get a nice filter that's 99% @ 20u for the same price or a few buck more. Saving a few bucks isn't as important and knowing the oil is kept cleaner over the OCI.
I got some Pentius PLXL6607 99% at 20 Microns $5.19 each
 
The daily variation of life, resulting in expected variation in"normal" wear, far exceeds the control a 99% filter can inflect into the equipment lifecycle. Do not forget that the lube filter is only one parameter of equipment lubrication wear control. Also consider these:
- OCI duration
- lubricant selection
- environment
- operational conditions
- air filtration (in the case of ICE applications)
With all of those factors held constant, a filter that's 99% @ 20u is going to benefit the sump, especially the longer the OCI is, than one at 99% @ 40 microns. Wear studies show it's the particles below 20u that do a lot of wear because they are small enough to get between moving parts in the oil film thickness. A high efficiency filter is going to catch and retain much more particulate that's 20u and smaller. Also, lower efficiency filters shed debris more than higher efficiency filters from dP surges and spikes that happen in a dynamic oiling system. Repeated shed and clean up cycles allows more debris to circulate through the system.
 
I think we do understand, and agree, ZeeOSix, that a certain level of filtration is necessary for a long, healthy life of the equipment. The question becomes one of where is that point of diminishing return? IOW, what is the "level of filtration" which provides a good ROI, and yet doesn't go into an abyss of the efficiency rabbit hole?

When I mention "ROI" in this case, it's a question of money spent vs wear control, not filtration efficiency. Filter efficiency is only one of many inputs into the desired output of wear control (I mentioned in a previous post). That's where the disconnect in many of these conversations exists.

I believe there is a concept of diverging curves; one which shows efficiency of the filter, versus actual control of wear. You can get ever tighter filtration, but you'll not be able to distinguish wear control past a certain level in real world use.

Yes, when you talk about 90% at 40um and 90% at 20 um, there's going to be a huge wear-control difference, and I accept and agree with that. But that's not the typical choices for off-the-shelf filters these days; most of them are rated at 20um (or 25um).

I've already stated that real world data cannot distinguish between a filter at 95% at 20um and 99% at 20um. I stand by that. The reason is that "normal" variation of daily use wear data is greater than the effect that any one minor filter efficiency delta can produce in wear control. In short, it's moot.

Let's not forget that most lab filter studies are HALTs (highly accelerated life tests). These tests are manipulated to advance wear in a VERY aggressive manner, so that the study can be conducted in a short duration of time. What is often misunderstood is that these HALTs, while good at showing disparity of performance in controlled variables, are poor at indicating a reasonable relationship with actual results from real world, "in the field" use. Further, they often focus on an efficiency delta that is very wide, so that the desired effect (that of wear disparity) is very clear. I have NEVER seen a study/test where two similar filters (say 95% and 99%; both at 20um) have ever been run. Most likely because those folks running the tests understand that such a minor disparity in efficiency isn't going to be discernible in test results, let alone in one's garage.

Let me give you an example of how HALTs can and cannot tell us ...
When I worked in the HVAC industry, we were doing testing on corrosion resistant coatings and paints. We had a large test chamber which we could induce and control environmental conditions; heat, humidity, salt-solution misting, etc. What we were able to prove is that paint A was better than paint B at resisting permeation (passing salt solution through the paint to the base metal). But that disparity in performance had no discernible relationship to a timeframe, which was the goal (for the establishment of a warranty period). The reality is that all the other factors of real world use (UV light, paint thickness control, damage from impacts such as hail, variation in base material metallurgy, etc) could not be overcome by the paint. Our testing proved that paint A was better than paint B, but we could not pin down a reliable duration that either paint would predictably stop rust from forming, even with controlled conditions in the chamber. The HALT was run over days, but real world use exists over years. The HALT is simply unable to establish any correlation to duration of exposure; that was the goal. The HALT proved one product better than another, but nothing more. If one paint failed early, both did so, but one slightly less than another. If one paint lasted a long time, both did, but one slightly more than another. But the GOAL was to establish a lifecycle prediction based on paint thickness for the purpose of warranty period; that relationship was never proven. We had plenty of historical test data to fall back on, relative to the real world warranty claims, and yet no distinct correlation was strong enough to say "X" mm of paint = "Y" years of corrosion resistance. The HALTs proved something we already knew; more paint lasted longer. Well ... duh! But the HALT failed to establish a real-world relationship of predictable life relative to paint thickness.

And I could give example after example of the same concept. HALTs are great at showing when one thing is better than something else, but they are really poor at showing how those results actually relate to the real world. So, in conclusion, I take lab studies with a large grain of salt. They are useful in establishing performance of one product over another. But they are very poor at actually showing how that disparity manifests in the real world. I enjoy reading SAE tests, but they often don't prove what many folks mistakenly surmise, because folks don't understand how to discern what is and isn't present in the data.


I most certainly agree that filtration is important. But there is a point where "good enough" is more than sufficient to establish a level of performance where efficiency is no longer the primary factor of wear control. Once the sump is clean enough, making it cleaner with filtration really doesn't make any discernible difference when the OCIs are of typical duration.
 
After not having used a Nissan OEM filter in years I threw one on an order I was already placing. The louvers were not open enough to my liking. I tossed it. There made by grupo Gohner in Mexico. I am surprised how many here are still using those - but to each his own.

Unfortunate because it used to be a decent filter with a silicon ADBV for a decent price. I remember when they were made in Japan by someone, and had holes not louvers.

I am done buying any filter with louvers anymore. Seems like none of them are any good.

Premium Guard or one of the variants, Hengst, or Denso - in that order - are the only ones that seem to have consistent construction at least.
I like Ecores more lately...
 
What's the least efficient filter that you would use on your vehicles?
Range of 99% at 20-30u is what I prefer. Of course I use Amsoil filters. I just figured you were being facetious. I have used some Honda and Toyota filters though.

Do you think using a 30u filter (not 40-60u) - the engine will wear out before the entire car will be just be sadly tired and not worth repairing?
 
Range of 99% at 20-30u is what I prefer. Of course I use Amsoil filters. I just figured you were being facetious. I have used some Honda and Toyota filters though.

Do you think using a 30u filter (not 40-60u) - the engine will wear out before the entire car will be just be sadly tired and not worth repairing?
What filter around here that anyone is going to use is 60um? Even cheapo filters seem to get pretty close to 99% at 40um.

Member @TC compiled a whole list of what was historically quoted on this site. There are some pretty lousy filters, but really most are reasonably good in the grand scheme.

I have concluded the filtration arguments are mostly academic - which I believe you and many others here have already stated.

Of course if the bypass doesn't seat or there is a tear, almost might as well have no filter really.

Lets not forget the best filter is a oil change - but short OCI gets no love around here much either.

None do a great job <10um and there are lots of studies in industrial that show particles down to 1um are still pretty abrasive.


 
What filter around here that anyone is going to use is 60um? Even cheapo filters seem to get pretty close to 99% at 40um.

Member @TC compiled a whole list of what was historically quoted on this site. There are some pretty lousy filters, but really most are reasonably good in the grand scheme.

I have concluded the filtration arguments are mostly academic - which I believe you and many others here have already stated.

Of course if the bypass doesn't seat or there is a tear, almost might as well have no filter really.

Lets not forget the best filter is a oil change - but short OCI gets no love around here much either.

None do a great job <10um and there are lots of studies in industrial that show particles down to 1um are still pretty abrasive.


Yeah I just took 60u from the graphic above.
 
One guy drove his 2 Toyota Tundras over a million miles each using just OEM Toyota Denso filters. So high efficiency filters are not needed to rack up millions of miles on a vehicle.
 
I am wondering if the reason that the major oil filter manufacturers are suddenly lowering the threshold of the size particles that they are able to filter out is somehow legal in nature as opposed to cost saving. For instance, if a company says that their filter will filter out 99% of 20 micron particles and someone runs a test using 10 grams of 20 micron iron particles and it doesn't meet that threshold, they could sue the company for false advertisement. You may laugh, but this kind of thing happens all the time with the way that prople are sue crazy nowdays. So, if the company lowers the threshold down to a 99% filtration rate at 40 microns, they know their filter will be able to accomplish that with no trouble. As an illustration, I read that Burger King was being sued because the whopper that they advertise on tv is bigger than what they actually serve you, so they are being sued for false advertisement. The second question that I have is whether or not a 99% 20 micron filter filters down to 20 microns at the expense of essential oil flow. Everyone knows that the more restrictive a filter is the slower the flow rate (or the more pressure needed to maintaon a given flow rate). For instance, has anyone ever verified if the Premium Guard EX has an acceptable flow rate with it's 25 micron rating. I would hate the put that filter on my car and have the motor periodically starving for oil longer than normal when I first crank my motor in the mornings because of a too restrictive oil filter.
 
I think we do understand, and agree, ZeeOSix, that a certain level of filtration is necessary for a long, healthy life of the equipment. The question becomes one of where is that point of diminishing return? IOW, what is the "level of filtration" which provides a good ROI, and yet doesn't go into an abyss of the efficiency rabbit hole?

When I mention "ROI" in this case, it's a question of money spent vs wear control, not filtration efficiency. Filter efficiency is only one of many inputs into the desired output of wear control (I mentioned in a previous post). That's where the disconnect in many of these conversations exists.
ROI shouldn't even be part of the decision, unless there was a huge price difference between high efficiency vs low efficiency filters. The cost difference between them can be anywhere from zero to a few dollars if you research and shop around, so for me at least the cost isn't a factor in deciding what efficiency filter to use. It all boils down to keeping the sump cleaner over the OCI to help reduce wear. Nobody can disprove what every wear study proves, that cleaner oil results in less wear. Regardless of by how much is irrelevant, because less wear for basically no significant added cost is going in the right direction. Same goes with motor oils ... the cost between the on-shelf oils are pretty close. Some people hear spend a lot more for "boutique" oils because they believe they are beneficial to use. Same goes with oil filters that keep the oil cleaner.

I believe there is a concept of diverging curves; one which shows efficiency of the filter, versus actual control of wear. You can get ever tighter filtration, but you'll not be able to distinguish wear control past a certain level in real world use.
The easy way to not make it a guessing game or conduct some elaborate "test program" in the garage is to simply use a high efficiency filter. Elaborate wear studies have been done for decades and every one of them conculdes that cleaner oil results in less wear. That's all I need ... I don't need to do my own test program to prove what's already been proven. Believe the science. Yeah, some people need some insane level of proof something makes a "meaningful" difference. I'm a filtration vs wear purist, so just the fact that cleaner oil means less wear is all the proof I need. I don't put a "has to be more than x% less wear" to decide to use a high efficiency filter.

Yes, when you talk about 90% at 40um and 90% at 20 um, there's going to be a huge wear-control difference, and I accept and agree with that. But that's not the typical choices for off-the-shelf filters these days; most of them are rated at 20um (or 25um).
I've never have argued about the differences between filters that are close in efficiency. I've also said in many of these debates that my personal efficiency "cut-off" is 95% @ 20 or 99% @ 25u. We've agreed on that a long time ago, and if you've put that same efficiency limit in your filter choice deisions, then you're also an avocate of high efficiency filters. I will always try to find a well constructed and quality manufactured filter that is ISO 4548-12 rated at 99% @ 20u. The OG Ultra was a good example. Filters with an ISO efficiency at that level also don't shed much debris as they load up which reduces the debris shed then reclean cycles, and they catch a good portion of debris below 20u, which is important for wear control.

My basic viewpoint has always been that better filtration results in cleaner oil which reduces wear, and regarless of the wear difference the fact that less wear is better ... regardless if someone believes it's meaningless because the car will fall apart or get T-boned before the engine "wears out" ... and everyone's definition of "wears out" is different. So if people want to define in thier own mind what a "meaningful" level of oil efficieny is, then that's for them to decide. Like said earlier, if engine wear can be reduced to some degree over the long run, then it's going to result in a more mechanically healthier engine, regardless if it's going to get totalled the next day. And yes, there are many factors involved in keeping an engine from wearing, and the goal of keeping the oil cleaner is one factor that can be achieved by better filtration.

I've already stated that real world data cannot distinguish between a filter at 95% at 20um and 99% at 20um. I stand by that. The reason is that "normal" variation of daily use wear data is greater than the effect that any one minor filter efficiency delta can produce in wear control. In short, it's moot.
Each variable has to be looked at with all other variables held constant. Anyone can argue that factor 1 out of 10 doesn't matter because factors 2 through 10 are always variable and changing and mask factor 1. That's how wear vs filtration studies can see that cleaner oil results in less wear. Even if everthing was held perfectly constant there would still be a detecible wear diffenece between a 95% @ 20u and 99% @ 20u. Sure, that difference would be small, and not be detectible if there were 9 other factors involved that were not constant.

Let's not forget that most lab filter studies are HALTs (highly accelerated life tests). These tests are manipulated to advance wear in a VERY aggressive manner, so that the study can be conducted in a short duration of time. What is often misunderstood is that these HALTs, while good at showing disparity of performance in controlled variables, are poor at indicating a reasonable relationship with actual results from real world, "in the field" use. Further, they often focus on an efficiency delta that is very wide, so that the desired effect (that of wear disparity) is very clear. I have NEVER seen a study/test where two similar filters (say 95% and 99%; both at 20um) have ever been run. Most likely because those folks running the tests understand that such a minor disparity in efficiency isn't going to be discernible in test results, let alone in one's garage. And I could give example after example of the same concept. HALTs are great at showing when one thing is better than something else, but they are really poor at showing how those results actually relate to the real world. So, in conclusion, I take lab studies with a large grain of salt. They are useful in establishing performance of one product over another. But they are very poor at actually showing how that disparity manifests in the real world. I enjoy reading SAE tests, but they often don't prove what many folks mistakenly surmise, because folks don't understand how to discern what is and isn't present in the data.
Regardless, every study shows cleaner oil results in less wear. That's all most people need to know ... they don't need to split the hairs to try and decide how that scales to "real world use" and if better oil filtration is "meaningful" enough or not. There have also been real world engine wear studies done, and just as the controlled fired engine HALTs the basic result is the same ... better filtration means cleaner oil and less wear. Most people only need that fact to decide to use high effiency oil filters.

I most certainly agree that filtration is important. But there is a point where "good enough" is more than sufficient to establish a level of performance where efficiency is no longer the primary factor of wear control. Once the sump is clean enough, making it cleaner with filtration really doesn't make any discernible difference when the OCIs are of typical duration.
As mentioned many times in these kind of debates, the longer the OCI the more benifical a high efficiency filter is. If the oil on a broke-in engne was changed every 1000 miles you wouldn't need much filtration. But again, why try to make it some complex science project when simpley using a high efficiency filter regarless of the OCI is going to cover everthing imaginable. And even still today filters that are 99% @ 20-25u are still pretty low cost ... sometimes even costing less than a less efficient filter. One thing about oil filters is that they keep the debates at a high level on BITOG. 😄
 
Range of 99% at 20-30u is what I prefer. Of course I use Amsoil filters. I just figured you were being facetious. I have used some Honda and Toyota filters though.
I like a bit better efficiency than 99% @ 30u (I'll always shoot for 99% @ 20u), but for short-ish OCIs it would work better than with a long OCI. I've stipulated many times the connection between OCI, filtration and wear. For me, I just shoot for 99% @ 20u regardless of OCI because it just covers all variables. ROI isn't even in my wheelhouse with oil filters because there isn't much if any cost difference.

Do you think using a 30u filter (not 40-60u) - the engine will wear out before the entire car will be just be sadly tired and not worth repairing?
My viewpoint is to keep the oil cleaner than not to reduce wear, regardless of by how much ... it's not based on if an engine will wear out before the entire car does - I don't operate like that with much of anything when I know doing certain things result in a positive. You could also say don't use other products that might keep the engine mechanially healthier because the rest of the car might not make it as far as the engine could.
 
Back
Top Bottom