Which oil filter filters best?

[fantastic]

Until all or most oil filter companies start sharing what each of their filters are ISO tested at we'll never really know "Which one filters the best". The best filter that filters is the one we have ISO testing data for but that info is severely lacking. In short the answer is not available at this time or may never be. However If you want to do several UOA w/ PC's to compare filters it could shed some information for your application if everything held similar between each filter. How many tests would you need to do? Some say 30 minimum while others say at least 4. Let us know if you do it.

 

[Fair enough]

Until all or most oil filter companies start sharing what each of their filters are ISO tested at we'll never really know "Which one filters the best". The best filter that filters is the one we have ISO testing data for but that info is severely lacking. In short the answer is not available at this time or may never be. However If you want to do several UOA w/ PC's to compare filters it could shed some information for your application if everything held similar between each filter. How many tests would you need to do? Some say 30 minimum while others say at least 4. Let us know if you do it.

Not a chance I would say one word. They should delete it all.

 

[Bkdc]

The most responsible filter companies do describe the balance between efficiency, flow, and capacity. It’s a matter of design compromises. I do wish every company posted specs for their filters.

 

[dnewton3]

... If an oil filter is used to filter oil that another oil filter has filtered a long time, and particles go down, the new filter may be better. More tests needed to verify. If no change seen, run filter longer. If the particles go up the first filter may be better. More tests needed to verify.

I'm going to continue to be blunt, so forgive me; it's not my intent to be rude, but you need to be set straight.

I don't really understand why you (and others) think that you can somehow usurp the ISO 4548-12 protocol with some home-brewed experiment. It's clear to me now that you have no idea how to develop and write a proper testing procedure; you have no DOE experience, or you'd not be asking these kind of noob questions in the first place. You can absolutely run the test in the manner you describe, and all you'll prove is that you have no idea what you're doing. Those of us who have spent years in ISO certified quality labs, as QEs doing statistical process quality control and writing DOEs, will brush off your "tests" as typical YT garbage not unlike PF and others.

But by all means, have at it; let's see what you can prove, or disprove, depending upon how your theorem is postured.

 

[ZeeOSix]

Until all or most oil filter companies start sharing what each of their filters are ISO tested at we'll never really know "Which one filters the best". The best filter that filters is the one we have ISO testing data for but that info is severely lacking.

Even if the ISO data says it's 99% >46 microns? There's enough ISO data out there to know which ones are probably better than the others. Ascent's ISO testing proved that many of the claims made from the manufacturers were pretty close to reality.

 

[dnewton3]

I lament the good ol' days when Wix had their real Beta data on display for all to see, and they were specific to each application. Several years ago, even before the M/H move, they began white-washing their data to just read "2/20=6/20" ... for most everything they make. Now on their marketing page, they claim "up to 99% dirt trapping efficiency", but have no qualification as to what that test is based upon, model wise or particle size. Great marketing for the average Schmoe to latch onto, but not helpful when actually trying to discern real performance relative to a reliable standard. 95% at 20um certainly isn't bad, but it's nothing to get wildly excited about.

For example, the 51515 (same as the old FL1A) and the 51348 (same as the FL910) both use the same media and both have the same beta rating now. Many years ago the beta data was different; slightly better for the larger filter because at the same flow, the larger filter would have a slower fluid velocity across the media per unit of square area.

I would imagine two things made it unreasonable to manage ...
1) anytime a filter was changed, they'd have to retest and certify; constantly managing changes and worry about someone "testing" their filters and then filing some sort of torte or class action suit
2) posting that stuff publicly makes them accountable and every BITOGer was calling them to ask about this, that and the other thing; overwhelming the obviously low-dollar "help" on the tech line


One of the few filters left that still has true Beta data on display is the 57151 filter for the Ford 6.7L PSD. 2/20/75 = 3/12/17. That's outstanding for a FF unit.

 

[Glenda W.]

I use two tests. The first is the ISO 4548-12 published efficiency that shows what a properly built filter is capable of. The second is a spot inspection(c&p) to determine if the filter is being properly built. Since the ISO data is normally published thats free and c&p’s are posted here for free or inexpensive to do at home.

 

[53' Stude]

Donaldson or OG ULTRA

 

[ZeeOSix]

I would imagine two things made it unreasonable to manage ...
1) anytime a filter was changed, they'd have to retest and certify; constantly managing changes and worry about someone "testing" their filters and then filing some sort of torte or class action suit

At this point I wonder how much re-testing is actually done. Big filter makers typically have in-house ISO lab equipment to test, so maybe those companies do re-test when a design is changed that will impact efficiency. The big filter companies do cross-check each other of their claims ... think MotorKing mentioned Fram would buy and test competitors filters now and then,

2) posting that stuff publicly makes them accountable and every BITOGer was calling them to ask about this, that and the other thing; overwhelming the obviously low-dollar "help" on the tech line

I still laugh when I recall the phone call I made to the Wix Tech Line years ago and asked what the efficiency of the Wix XP was after it was listed as "50% @ 20 microns" on their website, then that efficiency spec disappeared off the website. The guy at Wix said the Wix XP efficiency was "propitiatory". I asked him if it was now specified for some kind of Top Secret vehicle application.

 

[fantastic]

Even if the ISO data says it's 99% >46 microns? There's enough ISO data out there to know which ones are probably better than the others. Ascent's ISO testing proved that many of the claims made from the manufacturers were pretty close to reality.

If 99% at 46 microns is the only ISO testing data you have for an individual filter then yes. However, If there were specific ISO testing data for any individual filter you used you could compare them between one another. But until the rest of them start sharing it's pretty hard to discern the difference so go w/the one w/actual ISO tested data if one wants to know. If one is content with the marketing efficiencies, fine, many here follow the marketing claims and nothing catastrophic happens. Even the Ascent test you cited shows the Boss's better efficiency than claimed by 12%. Perhaps, M+H are showing their "Worst Performers" on their spec sheets but actually test better? Hard to compare other brands filters to Purolators since, unfortunately, when one asks for individual model efficiencies the companies like to say "proprietary" or never respond to emails.

Purolator Boss ISO 99% @46 Microns
vs.
XX Brand Filter Marketing claim of 99% @20 Microns w/o any ISO filter testing reference

In the above example which one I would pick if I wanted to know for certain about efficiency would be the Boss. OR do we gamble on "Probability" based on a marketing claim. Only each individual person can decide that but it should be presented as something to think about.

Purolator Boss ISO 99% @46 Microns
vs.
X Brand Filter ISO 99% @20 Microns

In the above example one could reasonably compare the two filters since they've both been tested to the same standard. Obviously X Brand Filter in the second example has higher efficiency, & has actually been tested, so one would naturally want to choose that one if it meets the mileage & or customers needs.

Then you have companies changing hands that may or may not keep things as they were good or bad. OP's question is one not easily answered, other than ISO testing data, & if you think you can answer the question to "Which One Filters Best" w/ISO data I'll start asking you which one I should use.

 

[ZeeOSix]

Purolator Boss ISO 99% @46 Microns
vs.
XX Brand Filter Marketing claim of 99% @20 Microns w/o any ISO filter testing reference

In the above example which one I would pick if I wanted to know for certain about efficiency would be the Boss. OR do we gamble on "Probability" based on a marketing claim. Only each individual person can decide that but it should be presented as something to think about.

I wouldn't use any filter advertised as 99%>46u (or 50% @ 20u) regardless if it was by an official ISO 4548-12 test or not. I'd rather take my changes on the one advertised as 99% @ 20u without the ISO reference if it was a well known brand filter - like the case of the Carquest Premium spec sheet that doesn't reference ISO 4548-12.

 

[fantastic]

I wouldn't use any filter advertised as 99%>46u (or 50% @ 20u) regardless if it was by an official ISO 4548-12 test or not. I'd rather take my changes on the one advertised as 99% @ 20u without the ISO reference if it was a well known brand filter - like the case of the Carquest Premium spec sheet that doesn't reference ISO 4548-12.

Okay, meanwhile OP's question is still not easily answered as a matter of fact b/c the information is still not available or extremely limited to certain models printed on the boxes. The OP wants to know which one filters the best. I would not tell them to go by a "Probability Meter" of sales marketing claims if they wanted to know which one was the "Best Filtering".

If enough ISO data between the filter makers was available we could make a recommendation based on actual ISO testing & not "probability". But you're not alone...some here do exactly that & go buy filters based on "it might filter better" & I don't fault them for it. Availability, pricing, quality etc are still important outside of efficiency that one must also take into account as well though too. One "high efficiency" claim could be wiped out by ripples.

 

[fantastic]

Now on their marketing page, they claim "up to 99% dirt trapping efficiency", but have no qualification as to what that test is based upon, model wise or particle size.

Speculating here... I do wonder if they have that claim based on "Course" test dust instead of "Fine" test dust.

One of the few filters left that still has true Beta data on display is the 57151 filter for the Ford 6.7L PSD. 2/20/75 = 3/12/17. That's outstanding for a FF unit.

Agree, Which I think might be the one recently migrated to the Purolator Boss media.

 

[dnewton3]

At this point I wonder how much re-testing is actually done. Big filter makers typically have in-house ISO lab equipment to test, so maybe those companies do re-test when a design is changed that will impact efficiency. The big filter companies do cross-check each other of their claims ... think MotorKing mentioned Fram would buy and test competitors filters now and then,

Generally, any ISO certified manufacturer has to validate any changes made to the product or the process; I had to deal with this frequently when I worked in the quality labs.
- design changes
- supplier changes
- new or refurbished production equipment changes
- production process changes
- component changes
- supplier location changes
- etc
Given the amount of change that has gone on in the filter industry since CVD supply chain issues, the resultant work required to "verify" these changes likely has been astronomical. The last thing any company wants to do is to post data and constantly have to update it; the logistics of keeping it accurate would be horrible. So instead of posting the "real" Beta data for each filter model, they just whitewash it now with the ol' 2/20=6/20 ... It's a target likely easy to hit for most models and takes very little effort to manage, as they have no intention of changing it.

I still laugh when I recall the phone call I made to the Wix Tech Line years ago and asked what the efficiency of the Wix XP was after it was listed as "50% @ 20 microns" on their website, then that efficiency spec disappeared off the website. The guy at Wix said the Wix XP efficiency was "propitiatory". I asked him if it was now specified for some kind of Top Secret vehicle application.

Yeah - I learned early on that Wix tech support line was a complete joke. Regardless if you called or emailed, all you got back was canned boiler-plate answers and no real substance.

 

[Fair enough]

I have a fairly thick skin so skipped over all the personal stuff. I have been around the block a lot more than that.
The idea was a way to find out what filter filters the best, what about comparing particle counts during use on your car? The technique is to use the oil a long enough time, maybe 5000 miles, or less, for the particles added and filter efficiency to stabilize, yet not have a filter filled to capacity. Then mount a new filter for a short time and compare. If the particles go down meaningfully, then the new filter filtered better. It’s as simple as that. What do we look for in filter efficiency? The lowest number of particles of various sizes.

 

[Lilliston]

I think you can follow those who say you can't possibly know anything no matter what you do, because you don't know anything and ISO labs are a lot better than you. Just forget the part about how ISO focuses exclusively on process and ignores outcome. There's good reason to do that, but it ignores and mocks what you're asking about trying to do.

I think you can also follow up on what you're asking about; do it and see what you get. Then you'll have actually done something. You'll feel better about your decisions, and you'll have some confidence in them. Whether a lab tech thinks it's a good way to run a lab is irrelevant since you're not trying to do that. Plus, if your results don't mean anything, but their position is (essentially, though they often don't quite come out and say this) that you can't feasibly know and anyway knowing has no impact on anything anyway, you're still in a better place than you were because you tried something and gained some confidence and experience. Kinda like the reason this forum was started in the first place.

Particle count data here has shown in fairly dramatic fashion that off-the-shelf oil filters can clean engine oil up from where it started new, out of the bottle. The worst outcome you can get from changing filters and doing a follow-up particle count a hundred miles later is the same ISO code on the PC.

 

[ZeeOSix]

^^^ The test would be better if you add some ISO test dust to your engine oil and see with filter gives the lower PC counts.

 

[ZeeOSix]

The OP wants to know which one filters the best. I would not tell them to go by a "Probability Meter" of sales marketing claims if they wanted to know which one was the "Best Filtering".

Anything that's better than one that has an official ISO 4548-12 spec sheet showing 99% >46 microns.

 

[dnewton3]

The idea was a way to find out what filter filters the best, what about comparing particle counts during use on your car? The technique is to use the oil a long enough time, maybe 5000 miles, or less, for the particles added and filter efficiency to stabilize, yet not have a filter filled to capacity. Then mount a new filter for a short time and compare. If the particles go down meaningfully, then the new filter filtered better. It’s as simple as that. What do we look for in filter efficiency? The lowest number of particles of various sizes.

Can you do the experiment as you describe above? Sure you can. What does it prove? Nothing appreciable in the real world. You're comparing apples and tangerines, with grapes and plums mixed in; a whole fruit salad of assumptions and lack of controls.

But the real problem, brushing aside the fact that your methodology is bizarre, is that you're going to try and run one filter and compare/contrast against another, assuming that a sample size of 1 is sufficient to declare something better or worse than another. That is total and utter poppycock garage science at its best. It is just as bad as running one UOA of lube X and saying it's better or worse than lube Y. You will have ZERO understanding of averages and variable deviations; none whatsoever.

This has nothing to do with the requirement of using lab equipment. This has everything to do with how to utilize correct methodology, properly control variables and make sound conclusions based on reliably accurate results; none of which exists in your plan.

 

[Fair enough]

Can you do the experiment as you describe above? Sure you can. What does it prove? Nothing appreciable in the real world. You're comparing apples and tangerines, with grapes and plums mixed in; a whole fruit salad of assumptions and lack of controls.

But the real problem, brushing aside the fact that your methodology is bizarre, is that you're going to try and run one filter and compare/contrast against another, assuming that a sample size of 1 is sufficient to declare something better or worse than another. That is total and utter poppycock garage science at its best. It is just as bad as running one UOA of lube X and saying it's better or worse than lube Y. You will have ZERO understanding of averages and variable deviations; none whatsoever.

This has nothing to do with the requirement of using lab equipment. This has everything to do with how to utilize correct methodology, properly control variables and make sound conclusions based on reliably accurate results; none of which exists in your plan.

The same oil is kept. The variable is almost entirely the oil filter. Its ok to have more kinds of tests to look at. What do you think people do here all day with pictures of this and that oil filter cut open? They look at them and form ideas, and even conclusions.
I never said one filter and done.