I wanted to share some comparisons of common oil filters' apparent cleaning ability.

The way I do this is to carefully cut the filter elements out of the cans without damaging them. Some elements are open at both ends, and I seal these tight at the end that goes toward the dome of the canister--away from the threaded end. I have a bread loaf pan of 5W-20 motor oil that I have mixed in a copious quantity of talcum powder (about 10 microns average as per The Filter Council); about half a cupful per quart. The mixture is thoroughly stirred, and the filter elements to be compared are immediately lowered in at the same time. When the the center tubes of the elements are full of filtered oil, A syringe is fitted with a 12 inch piec of 1/8 inch inside diameter tubing which is used to stir the inner contents of the filter and then draw a sample. The sample tubes are filled and plugged with pieces of nails at both ends. These sample tubes will be mounted vertically for more than a week to view the amount of filtrate that stacks up on the bottoms for comaprison. However, a lot can be seen just by the initial clarity of the diffeent samples.
I also note which filter elements fill faster/flow better.

A mention about the filters' advertized efficiencies:
The PureOne model that I used--a PL20195 states 99.9% efficient at 20 microns. (20 one-thousandths of a millimeter)

The Mobil 1 Extended Performance M1-209 states 99.2% efficiency, but without giving a particle size for which it can earn this rating. However, by telephone to the maker, Champion Labs, they told me they were "in the ballpark of a 10 micron nominal" filter. This means they catch about half the 10 micron sized particles--and more as the debris gets bigger of course.

The Toyota OEM made by Denso, part number 90915-YZZD1 stated nothing specific.

The Fram PH3593A (orange can) is shown as 95% efficient >20 microns. Presumably, they think you don't know what > means.

The WIX built NAPA Gold 1348 has a nominal rating of 21 microns and a Beta of 2/20=21/37. The beta means 21 micron at 50% and 37 microns at 95% catch, respectively.

The K&N HP-1004 I used, according to K&N via email:
Filter media efficiency (per ISO 16889):



100% at 40 microns

100% at 35 microns

100% at 30 microns

99.4% at 25 microns

98.9% at 20 microns

97.1% at 15 microns

92.3% at 10 microns

82.5% at 5 microns

But I am skeptical of the test conditions as this element was tested twice and never gave as clear of filtrate as it should have by comparison with know filter types.

Here are some pictures:

Left to right: Fram orange, WIX, K&N
These have settled for a few days. I am non-plussed by the NAPA Gold/WIX. But This does correspond to the given betas, efficiencies, in my opinion. It is a well built filter, but has never been an outstanding performer in my cleaning tests. I used Fram for the low standard.
The Wix NG still did a bit better than the K&N.



Left to right: PureOne, Toyota OEM (Denso), Mobil 1 E.P., WIX, K&N
It is hard to tell from the photo, but there is a clear visual difference in the first three, and the second two. The order of cleanliness from best to worst is also left to right. The PureOne, as expected, was outstanding. The Denso and M1 were almost the same but I have to give the nod to the Denso. The Denso also flowed best of all. The WIX and K&N were a bit on the mediocre side of cleanliness but both flowed well.
The PureOne flows twice as fast as the same models made several years ago. There must have been some concern about strat-up valve rattle--but I am speculating on that. The flow charts I have seen for the PureOnes show a nice, pleasing flow curve comparing favorably to other popular filters. The M1 flows about 2/3 again as fast as the PureOne in my test.
Of course, these are all at room temperature with gravity pushing the oil in.



So there you have it. Numbers are numbers, ads are ads, and dirt is dirt when held up to the light.
Take this for whatever it's worth.

Rob

Very good presentation, informative and interesting. Thank you for taking the time to do this.

Very interesting, thanks for the time and effort!

You're welcome guys.
Sorry about the dumbing down the terminology, but I posted this first on a non-oil forum and copied and pasted it here.

Would you mind testing some more exotic media like the Donaldson Synteq or Fleetguard Stratapore stuff?

Love to, but It's time and money. I'll get to it when I can though.

Wonderful! I'd be curious as to the results, since they have some very impressive specs (on paper of course).

Surprised to see Denso doing well in the test.

I was too the first time I tested one several months ago.
I expected the good performance this time--but still wanted to run it again for direct comparison with some new filter types...It's a really good filter. Especially for less than $6 at the dealer.

Thanks for the tests. This answers my concern with pureone having poor flow in tests done a few years ago but great flow on the flow graph circulated on the web.

Yes, I cut open a blue one and a new gold/yellow one (same part numbers) and posted here about the flow difference, etc. The media even looks a little different now than on the old style PureOnes.

Excellent home grown turbidity/opacity test.

I can't even pronounce those words, Gary! But thanks.
I'll post more pics when the grit settles on the last five tubes.

Cool, when i run out of wix filters, I wont hesitate to buy OEM filter's for my mom's camry

Interesting test.

But what that doesn't take into account is that oil is under pressure and forced thru the filter. This would force many particles into the media material and as the particles built up by constant circulation the efficiency would increase.

These filters also to not claim to be good at filtration that fine. Most use the 20-25 micron level to show good filtration.

But thanks for taking the time and coming up with the test.

You bring up an interesting point. Why are filters usually tested at 20-30 microns and the Filter Council is using 10 micron average talc? Haven't I read that most of the wear comes from 30 micron particles? Turbidity from particles too fine to cause much problem may not be important.

Thanks for the input, Z.
Yes, to a point they will collect more particles with recirculation.
There is a a point of diminishing returns, however, on the finer particles.
Most filters are already rated with the multi-pass test until the filter becomes plugged and, of course, even with that test they do not catch all the stuff.

The reason for this examination is to make a visual comparison between filters of known and unknow advertized efficiencies. e.g., if the PurOne is supposedly in the ballpark of 99.9% efficient at 20 microns, and produces a relatively very clear filtrate upon a single pass with the same contamination level in the oil as filter "B" which produces a milky, opaque filtrate, then we can probably assume that the filter B has a much lower efficiency than the filter with the advertized 99.9% rating (with either mulitple or single passes).

It helps me to decide which to put on my engine.
I don't care too much about % ratings, but if I have a filter that catches most very fine airborne-able particles such as talc in the very first pass, and flows easily, is constructed adequately well, then I consider it a superior filter choice for my engine. (I am not trying to figure a specific muti-pass beta.)

I believe the current consensus is 10-20 micron particles cause the most wear.
The filre council is not using talc as far as I know, but I got the average micron size of talc from their website.

Now that the oil is clear above the “grit stacks,” here are some more objective pictures:
Left to right; Fram orange, WIX, K&N
It appear the WIX does best, Fram second and then K&N.





Left to right; Mobil 1, PureOne, K&N, Denso, and WIX
This oil/talc concentration is stronger than in the above picture.
Note that I attempted to bracket the tops and bottoms of the grit stacks for visibility because some of these are resting on air bubbles rather than the bottom of the tube. The results are as expected considering the relative turbidity of the oil tubes before settling out with the exception of the K&N which must have developed a media leak. As in the picture above, it already has been tested. I don’t know what happened to the K&N as it was handled with the same care as the others. Must be a defect. (still unacceptable to leak like this)

Great work Rob got your PM been busy here not much puter time,anyway very nice man looks like the P1 is still a great performer i too would like to see more tested like the synthetic fleetgaurd and RP and AC UPF could PP funds if others would be willing too.

This is why I don't doubt the beta ratios claimed for the PureONE. This home test shows it's a high end performing filter. Thanks for the effort on this river rat!

I was curious if you sealed up any of the bypass valves mounted on the dome end of the element to ensure there was no leakage through the bypass valve during the test.

I don't doubt the beta ratios either!,id like to see some synthetic filters tested

You're welcome guys. I'll get to as many as I can eventually.
A buddy just gave me a Mann and a Mahle to test. I'll try to do those this week.


If I can't make it leak, I do not seal them.
Frams leak at the cardboard of course, so I siliconed that one. And the Denso I just removerd it and sealed a piece of sheet plastic over the end with silicon.

Numbers aside, there is no filter I've yet tested that clears up the oil as well in one pass. Period.

PM me your adress Rob i'll send you a AC UPF-52.

It might be more helpful in selecting a filter if leaky bypasses were allowed to leak. I never fully trust the metal to metal seals. Less so once they have let some unfiltered oil through.

True. I agee. But that is a separate test IMO for what I'm doing here.
Of these filters I cut open, only the Fram had a leaky bypass and I felt that it would let so much through that it would immediately be disgualified before I could look at the media alone. I wanted to look at the media because it has a published efficiency--even though it was vague by saying 95% at >20 microns. That way I could get a visual comparison of what that efficiency looks like.
(The reason I took the Denso bypass off and plugged it was that it was just easier. It didn't leak for me.)

Awesome!

So whats going on with K&N ?

Well, I would summarize it by saying that:

1. a lot of folks who have filters mounted in any but threaded-end-up have complained that the anti-drainback valves on the K&N and Mobil 1 filters have let oil leak out overnight and cause noisy start-ups. Although I have not experienced this myself.

2. They are a pricey filter similar to the M1, but don't seems to have the same media. This test does not show better than fair in the fine filtration results, whereas the M1 and some other premium filters were in the excellent category. (My opinions.)

3. They do have a very sturdy construction.

4. Since there was a flukey result on the second K&N talc filtration test, I would like to re-test with a new K&N filter.

Maybe I will show a little more care dissembling filters and test the bypass after use. The dome end ones, you could pour a little oil on top and see if it leaks through.

Rob that UPF filter is on it's way!!!!

Cool. Thanks, Dan


Ya. It would be good to test them after they been in use. That's a great idea. They could get crud stuck in the gap.

I'm in process of retesting the K&N with an HP-2004 this time. I don't fully trust the results with the one before. Might have got a defective one since it filtered fair at first, then practically no oil cleaning on the second try. Wierd.

I'd send you a Donaldson, but I think shipping to the states would kill me.........

Ya. Probably. I'd like to pick one up sometime and have a look though.

If you are handing out UPF filters I'll take one...lol I'd even cut it open and post pics after using it.

Well it's for testing purposes or i would not be giving one away lol, but i have two cases and access to more and want to know how they perform and if there worth the extra coin,i think there a great filter that will do well.

someone send RiverRat a RP to test!!!!!!

When the ACDelco UPF series filters came out, they were advertised (by ACDelco of course) as very efficient filters that also had very good flow characteristics.

Yup i remember,suppose to flow well,i run them on my Z71 with 10w30HM all winter and no problems with ticking or anything,great filter.

That's where this famous graph came from ... ACDelco was showing how well the UPF Ultragaurd filters flowed (with cold oil in this case) compared to others.

There was a big hupla by some on the Vette boards saying the UPF didn't flow well, but I had many emails with the ACDelco Tech Dept, and they said the people doing "garage flow testing" didn't know what they were doing and that the data in the graph was accurate based on controlled lab testing.

Thought you all might want to know that the second K&N oil filter, HP-2004, now I tested it three times cleaning the element with kerosene between and it's filtering as good or better than the M1, and almost as good as the PureOne every time. I think something was leaking in the first one I tested. Maybe a tear in the paper.
I'll post pictures when my wife gets back in town with the good camera.

Rob

I will gladly send you a Royal Purple filter and have an Amsoil EAO filter shipped to you.

Please do!!!!!!

you think the kero is aiding in filtration somehow? cant see the k&n doing as good as the M1?

Great! I'll PM you my address.


No, the relative filtration between the brands seems the same whether the element is brand new and bone dry, or cleaned with kerosene. The new, dry K&N was first tested against two other new dry, elements--all fresh from the canister. On the second and third tests with these same filter elements they were all cleaned simultaneously in a bucket of kerosene.
I try to keep all things exactly equal--especially since I learned early on that if an old element had been dipped in muck many more times than a newer element, the old mucky element shows a dirtier result. Apparently the grit packs in the paper to some point that it goes on through. So now I only use new elements or equally used and cleaned elements for side by side comparisons.
Once, I tested an unfiltered sample and it was cleaner than the sample from an old uncleaned element. So these pictures are after my attempts to prevent this phenomenon and should give some good idea of filtration.

Can you post pics of the bread pan, how you actually did the test? I'm having a hard time imagining the procedure.

From what I gathered from the description given in the first post by river_rat, he cuts the filters open to get the elements out, then simply lowers them into the pan containing oil that's been mixed with the talcum powder. The elements are lowered in with the open center core facing up to cover the media, but not enough go above the top end. The oil will then flow from the outside to the inside core of the element, just like it does on a car when in use except only gravity causes the oil to flow slowly through the media in this test. This filters the oil, so only cleaned oil fills the center core of the element. He then sucks the clean oil out of the center core into the clear tygon tubes, and positions them vertically so the talcum powder will settle to the bottom.

That's exactly right. Only now I've switched to a bigger better pan because my wife is out of town and didn't see me take it. Same procedure.
I can still post pics though. A video would be cool but I don't have a digital video camera.

I'll be shipping the RP and Amsoil filters toward the end of this week. If only we could get more people to donate some more high end filters like Donaldson, we could have the first high end filter shootout and see how the synthetic medias compare.

I've gotta step-up now that Buick has

PM me your address and I'll send you a Donaldson SYNTEQ.

Yes! You took the bait.

hahhaahah

Second K&N test with a new filter pic as promised.

L to R: PureOne, Mobil 1, K&N

Why can't you see the K&N filtering as well as the M1?

I must say, this thread is awesome. I race a Honda Challenge car, and have been using Castrol Syntec 10W-40 and K&N oil filters, replacing both every other event (~every 6 hours on the engine). After just a day on this site I already am feeling a need to change the oil I use, and it looks like a filter change may be in the works too.

Great info, great test, thanks for the wealth of information!

I thought the K&N and Mobil 1 was the same filter? I seem to remember someone saying they even had the same part numbers on the inside?

They are made by the same company, but may have different media specifications. I don't know.
The first test came out wonky (torn paper?). The second test with a new filter comes out pretty similar to the M1 every time.
I can test five times with the same two elements and one will be a little better one time, and the other a little better the next, and so on.
What I do see is a trend when one filter beats the another by an easily visible margin time after time.
In this case, laying the tubes down made the PureOne look a little worse because the grit flowed a bit.
But it looks to me like this K&N is consistently neck and neck with this M1 out of three tries. Seems to be one of the better filters, overall.

Yer welcome. It ain't perfect, but it's better than sales hype and vague specs.

So far my favs in order are PureOne, M1/K&N, Denso, Mann and the Wix 51348 did pretty good.
These are the ones that have had a good trend in repeated tests. All flowed well for me too.
I'm looking forward to the EaO, RP, and high-end Delco. I expect them to do well.

First...Excellent Job!!! I want to thank you for taking the time and effort to put together this report.

One request...if you get the opportunity...I would really appreciate if you could add a Motorcraft filter to your report.

I typically use the Motorcraft FL-400S in my '08 Liberty. I consider it my "bang for the buck" filter...$3.47 @ my local Pep Boys. I refuse to go cheaper, until someone clearly proves that they can make something better for less. (On the other hand, my next step up would be to the NAPA Gold / Wix...so seeing that comparison would be great.)

Not that I trust everything they publish...and I do not wish to start any bashing of them here...either, but it surprises me that Consumer Reports or some other organization hasn't tried to perform a thorough and independant comparison report on oil filters. The only really good information that I have ever found has either been by the guys here or on 1 or 2 other websites.

MY THANKS TO ALL OF YOU!!! Keep up the good work!!!

It's a conspiracy. They are paying CR off to NOT perform the tests

well from what i understand it's not suppose to,the M1 is a extended use filter the k&n is not,k&n is for better/higher flowing apps less filtration.

Rob my UPF show up yet?

That's what I thought--and maybe that was the idea and design when they started out?
The Betas the guy at K&N sent me were really good though.


Just today--thanks again for the contibution. Looks like an M1 baseplate. Seems really well built.
You want it tested up against a PureOne?

yea looks like it, there pretty good built filters,just test it up Rob then compaire to the others,keep records of your work.

Will do.

It's made by Champ Labs. That would be a good reason for it to appear to share something with the M1 or K&N for that matter.

Thats what i was thinking,it looks like it...

FYI, It has to be tested simultaneously with others since this is a side by side exam.
The oil talc mixture changes slightly in composition as each dipping retains some oil and talc in the filters. When I remix or add to the oil/talc, the comp changes a little again, so I need to run new elements at the same time, then clean all of them the same, then run them again at the same time, etc.
I try to keep all the variables the same so that when a pile of dirt keeps coming up higher or lower in one brand, I feel like I know what's going on.
I'll pick up a new P1 (the cheapest gold-standard), maybe a new standard Delco.

Rob

thats cool rob,whatever you need to do!!!

Dan, I went this morning and got the PureOne cross to a PF52 and a regular AC/Delco PF52.
However when I got home I found that the PF52 that I bought is an old Wix style rather than the Champ. Ecore. Since the old style is NLA except for old stock, do you want me to take it back and get the Ecore type PF52?

I've never seen an E-core PF52. I think you'll have to buy a ST3980 if you want the PF52 equivalent E-core.

I've never NOT seen one for two years--until today. Weird.

Don't you hate it when that happens? You're 100% confident ...seen it for as far back as you can remember ...someone says "what? I don't believe that this is so." ..you say "sure, I'll go get one (or whatever) ..and the universe just makes a liar out of you (so to speak).

Wait until you're 50+. It happens more and more all the time.

Not all AC's are made by Champ. Isn't the PF 52 too small in diameter to be an Ecore? I though that it was only the 3'' and up ones.

Ya. I do and it did. I just looked at the parts store on my break, He's right I was mistaken with the PF 53--I've used so many of both of these and still get the numbers crossed. All the PF52s are non-ecore up there... <---Me

No the PF 53 is the same ~2.98" diameter and is an ecore. It's just about an inch and a half shorter.

Testing the UPF52, PF52, and PureOne PL24011 side by side now. All new.
I'll post all the pictures and measurements at the same time when I get results back.
For now I'll say the UPF52 has a very different media--white and extremely thick like a blanket that's been all pleated up. Interesting.

yup there's Ecores pf52,have them in my wm.

the UPF has two layers of media,looks like a coarse then a fine underneath.

UPF

OK, thanks. I thought I was going senile. Whew!

Here's an intersting aside:
Just for giggles I put a K&N HP-1002 in the mix yesterday.
Remember I was a little dissapointed in the apparent K&N filtering? Then I got a new, larger K&N and it filtered really well? Well, the results aren't "final" enough to post a good comparison yet, but I can already see that the little HP-1002 did not filter anywhere close to as well as the large K&N.
I wonder if it has different media. Guess I won't be using that HP-1002 on my truck after all....Hmmmmm!

Should gladden the hearts of all the oversize filter people.

Indeed.
Myself, sometimes I use a "one up" size longer filter, and sometimes not--but I'm thikning about emailing K&N and asking what the deal is. Theres' abig difference in this case and I usually don't see any significant difference in this test with different sizes from the same type/brand.

You've got to figure a multipass test with the same stuff. The smaller filter may prove just as good over more evolutions. It would then break down to (basically) the life span of the filter.

When I had a PF 47 app, I often ran a PF 52. Not much choice for a cartridge filter. Now, about avoiding something smaller now I can't buy a ST 3950.

Thanks for doing this test Rob.

Perhaps so.
I'm surprised the little one has a different media. Maybe because of its size, they wanted more flow?
Anyway, the initial media porosity is evidently courser on that little guy.

Thanks to Daman, I got an AC-Delco UPF-52 to check out.
I decided to compare it to a PureOne PL24011 and a Regular Delco PF-52.
These are equivalent filters for the same application.


I was a little surprised the UPF52 was using a nitrile anti-drainback valve.
Not that that’s a problem but all the premium filters seem to use silicon nowadays.


A close up of the UPF52 media. That stuff is really thick—and packed in there. 34 pleats.


And the standard PF52--48 pleats:


Dip test--left to right and top to bottom: K&N, PF52, UPF52, and PureOne
Time to fill the center tubes to the level of the oil mix:
K&N and PF52 tied at about 60 seconds
PureOne = 3 minutes
UPF53 = more than 5 minutes


Left to right, PF52, UPF52, and PureOne
There is an area of hard to see settlement in the UPF 52 because the oil is still a little cloudy after more
than 48 hours. I tried to highlight the picture a little bit. This was the clearest of the photos.
I ran two identical dip tests with these three and the UPF52 filtered the best both times.


The UPF52 was a little bit ahead of the PureOne on filtering out my talc, but the flow rate was by far
the slowest of any filter I’ve tested so far (but also the best filtration I’ve ever seen…but I might just stick with the P1 all things considered.
The regular PF52 did a really good job too and seems to be an excellent choice.

I'd run any of these and feel good about it.
If you aren't having any low oil pressure indications with the UPF52, looks like hardly any dirt will get through it.
They aren't pulling you leg, apparently, with the efficiency quoted.
The PureOne is second place--not by much--but flows quite a bit better.
The PF52 flows great, and was third in this filtration test, but better by far than "average" as far as I can see.

rr, excellent quality pics, thanks for sharing, also to those who sent filters. Like you, I'm a tad surprised that the UPF52 uses a nitril adbv, but it's not that big a deal IMO. All look like quality filters.

fwiw, the pics of the procedure, make it easier to understand. Glad you interpret the results though. Adding in the 'cost factor' would also be a purchase consideration, at least for me.

Thanks. I was glad to get the chance to look at an AC Ultraguard as I've never seen one in person. I've heard they've been discontinued?

It's pretty simple.
1. The main thing is keeping everything even and the same between elements. e.g., you can't test a new dry element next to one that's been tested and is oily and talcy.

2. The mixture must be stirred immediately before dipping---all elements at the same time.

3. A 12 in. piece of clear vinyl tubing (1/8" I.D.) from Ace Hardware is used. The oily mix that fills the center tube must be stirred immediately again before drawing a sample as the talc settles out quickly. The sample is drawn from about the center of the center tube. Not sucking rignt off the bottom of the element where crud can form.

I use a no-needle medicine syringe from Walgreens. Free for the asking at the Pharm. counter.

I think it is a cool test but hard to judge flow as in an engine the oil is being pushed through and not just slowing filterting through the filter.

Right, all I can tell is one filter media resists flow much more or a little more than another...That it will do for sure.
It won't tell you the flow will be insufficient in an engine.

Since motor oil is relatively newtonion, the flow test shjould be fairly valid.

I haven't commented too much on this thread, but I am enjoying it. I muchly prefer numbers to guess work.

I'd give numbers if I could.
Filter manufacturers can but they have a vested interest in making those numbers look good.
Usually though they seem to be more or less in line with the comparisons I've done.

When I look at some brands, the website or box just says "excellent filtration" or something like that.
I can't use that info...but if their "excellent" filtration keeps coming up approximately the same compared to, say, a Wix with published info, then I can see that it is probably in the ballpark of the Wix Beta.

Otherwise what I can do is show that filter A always lets a bunch more junk through than filter B, and flows easier or harder in repeated tests.
I'm not selling anything, but that's good enough for me to choose a filter for my truck--along with general filter construction. I'll run BuickGN's filters when I get them and that will probably be all.
...My shop is a mess with oil and smells like baby powder and petroleum, I've picked my favorites and trust my results, and spent more than $150 on filters.

That's enough for application to The Fraternity of Curious BITOG Lubenuts.

Yes i knew from day one when i started using them UPF it had a nitrile anti-drainback valve but that doesn't matter as much to me yea would be nice to have the silicon but i love the better filtration.

thanks again Rob for doing these tests,and pics great job bud!. Very interesting for sure not surprised with the slower flow rate as you can see with my above link post there's two layers,but i still think that's relevant and the filters made to flow well,i get these filters for just over $8 for a case of 6 so i'm going to continue to use them for sure.

i have one on my pick up now and it's cold here in MI no start up noise with 10w30HM.

I resemble that remark.
To me, "guesswork" (if that's what labman meant)is looking at a filter's vaguely stated published efficiency--like XX% at >20 microns like Fram says...Not cutting it open and actually filtering with it.

I don't know what the filter's lifetime average % is, but there's no quessing relative media porosity when the crud settles--and they are definitely related.



I've run just about every possible type of filter I can get off the shelf, under all conditions, on my old van that has an OP gauge tapped after the filter. I've never seen any noticable pressure difference between any of them, so I'm not surprised.

FWIW, All these filter elements that were removed from the cans, including the UPF52, when new and dry--you can put your mouth over the center tube hole and breath all day through them. You could walk around as long as you want and not run out of air...If that makes any sense. When they are oily, sure, they seem really restrictive, but then so do the oily bearing gaps and oil galleries.
Seems to me like a pretty minor pressure drop across the element relative to the rest of the oil system in most engines.

Have to remember too when that oil gets hot 180-200 deg it thins out and flows waaay better.

so not an issue with any of these filters IMO.

I agree. I've washed some older elements in hot soapy water from the inside out to see if it would make any filtering difference after they dried. (it didn't) and the sink faucet cant even keep the center tubes full at full blast. It runs through too quick. And full temp 20-30 weight is about the same viscosity as whole milk...literally.

Main reasons I'd use the PureOne is cost and availability. But that tough built UPF52 really stops the dirt.

Guess work to me are those that look at the outside of a filter and rave about how good it looks, or even cut it open and do the same without comparing it to anything. Also speculating the results of what the end caps or center tube are made out of.

River Rat is producing some real data. Since motor oil's viscosity doesn't vary very much with shear, the gravity based test should be fairly valid. And room temperature is an interesting data point. Do we want a filter that bypasses at start up?

Ah. I see. Thank you for clarifying.
Yes construction is only one aspect.

I agree...

Real enough for me,it's at least something to go on....

Guys, the EAO and RP filters I promised are still coming. Got into a dispute with the bank over some checks. Long story short the funds are short but it will be resolved shortly. I've been wanting to see this sort of test for so long and I'm the one holding up progress now.

No worries. I haven't cleaned up the mess down there yet.

Man, there are some really great threads going on right now. Thanks for this one Rob! Can't wait to see the results of the RP filter.

Cheers!

Ken
(currently running M1 EP)

I can't wait for the Donaldson results

Can't wait for the EAO results! Thanks for all those involved!

I can not wait for the results of the Royal Purple filter either.

Ya me too. While we're waiting I'm testing an ecore. No pics yet but it's looking pretty good in the filtration.
I've gradually come to accept these ecores as a decent filter for cheap.

So with the high filtration rate of the pure ones, is there any chance they are filtering out the oils additive package?

Not a chance.

Not a chance in H E double hockey sticks!

I asked the same question but for bypass filters--which filter super duper micro teeny fine. Gary Allen set me straight by linking me to a site that compares the micron(s) range filter pore size vs. the nanometer range molecule size for the additive molecules.

No, like these guys said, it's impossible.

Johnny hasn't corrected me, but I suspect some of the solids in my YB Pennzoil may be ground limestone. They collect on the filter the first pass and lay there waiting for an acid to come by and nab it. Maybe other stuff too.

I wonder if shaking up the oil keeps it in suspension for a while?

Here is the Ecore test I promised, along with some notes on the Wix observations.
The picture below is representative of several comparison tests that kept coming of the same way.
Left to right: Ecore PF-53 IIRC, Mobil 1 Extended Performance M1-102, WIX 51348, and Denso (Toyota OEM 90915-YZZD1).
They are all for the same application.

The Ecore always does pretty well in the comparisons with it's published efficiency of 96% single pass at 15-20 microns and 94% mutipass at 20 microns.

The Mobil 1 always tests for me as one of the better filters for filtration, and is about the same as the K&Ns I've tested. This (the M1) is advertised at 99.2% mutipass efficiency but without a micron size given. Champion Labs however told me by phone that it was about a 10 micron nominal filter (50% at 10 microns), which is pretty respectable and the results show it. It does not test as good as the PureOne, but it isn't all that far behind it either.

The WIX efficiency for this particular filter is not spectacular, showing a 21 microm nominal rating, but in practice, always filters about as well as the M1.
Does this meant the M1 filters worse than expected, or the Wix filters better than expected in the real world?
Since the M1 consistently meets or beats any number of filters with good published efficiencies, and it not far behind the PureOne, I have to go with the weight of evidence that the Wix filters better than expected and that the Wix published ratings are conservative.
The are in fact a lot of variables in how the SAE tests are done, and so published ratings vary.
The Wix, may lose some of this efficiency after it has been in use for a while and may appear closer to the conservative published ratings--but the media is obviously stopping fine debris as well as higher rated filters. This is also evident in that the flow rate in the oil in the test with similar media surface area is about the same as the M1. This indicates to me that something--either median pore size, or depth of media is probably fairly similar, because they both use a blended media. The other possibility is that the Wix ratings that they published have not been updated in some time as media technology keeps improving.

The Denso flows very fast, and filters decently from past experience--maybe average. It's a good filter, I say, but probably not the best at capturing the smaller particles in the fewest passes.




Pictures of my messy shop. I've done a lot of tests!

Great job rob,your a valuable member here for sure,this stuff takes time and effort on your part.

Hey thanks man.
I've learned a lot in the process and think I've got the test procedures down for repeatability of results.
I have also reluctantly changed my opinion of the Ecores. I was kind of concerned about them at first, but at least the later versions, seem to stay together well and I consider them a viable choice for a fairly decent low cost filter now. I still don't like Frams except for the Extended Guard model.

Yes ... Rob is now a certified "oil filter testing mad scientist".

Sure makes me glad I installed a Pure One filter along with 6 qts of Redline 0w-30 last Thursday!!!!! Thanks for all your hard work testing the filters!!!!

...and just in time for Halloween. Think I'll dress like a big, scary Fram!
You can feel good about that!

Wow, fantastic thread! I'd like to add some comments, if I may.

I don't think this test is valid in testing flow rates. I think the rate in which oil flows from outside to inside of the element has just as much to do with media porosity and osmosis, rather than flow created from a pressure differential between the outside and inside of the filter. At 6" of head, pressure at the bottom of the pan would only be 0.2 PSI.

I'd also like to say some things about efficiencies and beta ratios. As most of you have already commented, you have to take efficiency ratings with a grain of salt. Not only is it absolutely mandatory to communicate what particle size is being measured, beta ratio is a more important number than efficiency.
Any beta ratio over 10 looks like a fantastic number, because it would be 90% efficient (Eff=[beta-1]/beta). Even 97% or 99% looks like a great number. Still, a beta ratio of 100 (which is good in the automotive world) is [censored] in the fluid power industry.

A beta ratio is derived from counting the particles upstream of the filter, and dividing it by the particles counted downstream of the filter, normally in a multi-pass test. So if you had 10,000 particles upstream, and 1000 particles downstream, your beta ratio would be 10 (which is [censored]). If you had 10,000 particles upstream, and 10 particles downstream, your beta ratio is 1000, which is the common standard for high quality hydraulic filters (beta ratio 99.9%).

So big deal, 99.9% isn't that much better than 98%, right? Well 99.9% is a beta ratio of 1000, and 98% is a beta ratio of 50. That means that if you have 10,000 particles upstream, only 10 particles got through the beta 1000 filter. And if you have 10,000 particles upstream of the beta 50 filter, you have 200 particles downstream.

So that's 10 particles downstream compared to 200 particles downstream. The beta 1000 element is 20 times better at removing dirt than the beta 50 element.

Also important is the particle size that's listed with the beta ratio (a particle size must be listed, or it's a useless number). Any old filter can get 99% efficiency at 20 micron. Show me a automotive filter than can do 99% at 3 microns.

Hydac makes hydraulic filter elements that have a beta ratio of 1000 at 3 microns (99.9%), and you could even use them in your car.

river rat, I might be able to send you one of those elements for testing...

edit: wow, I can't believe c r a p gets censored!

Most of us here are versed on the Beta ratio deciphering Gen1 ..what we don't have access to is the ISO protocols that are used in the determination. Former insiders to the industry have implied that anyone can manipulate their numbers due to loopholes in the protocol.

Sure, you can use hydraulic filters in a filthy automotive environment ..but probably at 1/3-1/2 the service life. It's very hard to trump the filtration triangle of size/efficiency/lifespan. Under no circumstances that I can figure, even with fancier depth medias, can you do so under the restraints of $$$/mile using industrial filtration of inordinately high efficiency.

Signed up just to comment...Great thread! A lot of work too! Picked up a free K&N HP-1004 filter with 5 qts of GC from Vatozone for $28 and didn't know anything about the quality of the KN...thank you. :)

Well, if most of you are versed in beta ratios, then I hope my post was new information to the few whom didn't. The thing about the higher-end hydraulic filter manufacturers, is that they publish a lot of engineering data that allows you to make competent choices, regardless of your knowledge of ISO protocol. Those high-end manufacturers publish the exact conditions in which they do their testing. Hydac for instance, even with their extremely high filter quality, stops the multi-pass test at 33% before bypass pressure. They also publish their flow charts, pressure drops of both elements and housings, dirt holding capacity, collapse pressure etc.

I know an engine is a very dirty place compared to hydraulics, and the extremely low oil pressure means particles do less damage (which is why engines last a ridiculously long time for a piece of machinery). But the example I use with hydraulic oil, is that the dirtier your oil is, the more it's like a lapping compound. Dirty oil causes and exponentially high rate of wear, and consequently, that wear causes the oil to be dirtier. If a 3 micron filter with a beta ratio of 1000 was used in either a hydraulic or automotive situation, the initial rate of clogging would be high. As the oil got cleaner from multiple passes through the filter, the rate of wear also goes down, and then the number of particles that are causing wear goes down too.

Running a remote oil filter may allow you to use a way larger filter element, with much higher dirt holding capacity over an automotive style spin-on element. Equipped with a clogging indicator or differential pressure gauge (not that expensive when the filter head is equipped for it. Princess Auto sells them for like $8), you'd never have to guess at how clogged your filter is, and it may even say you money over time by not replacing filters at regular intervals.

Well, I don't think I can agree with all that. There are surely studies that conclude that getting below the 10um level indeed reduces wear, the study was in a diesel where soot production and control is serious business. Oils have managed to keep the stuff from forming up into larger abrasive particles and bypass filtration filters out the rest.

It still, more or less, works out to how long you can keep the sump in place.


I'm not saying this well here. Yes, you can do as you prescribe. You won't typically get the sump to last long enough (for other reasons) to make it work in a gas engine. You'll be ditching the sump and resetting the particle counter, so to speak.

So, to make it work you would need someone like Bill in Utah ..that's done 60k/year from the day he got his license ..and have him manage one OCI/year to warrant the cost difference over cheap oil and filters.

..but that $8 filter minder isn't going to work on a full flow filter. It will probably work on a zero pressure return bypass filter if you have an intermediate orifice/restictor.

Beta Ratio IS the way the filter efficiency is expressed. They are tied at the hip.



Guess you better go to the forum (can't remember which one) where the board owner has posed a thread about profanity use on this board.

There are TONS of "any old filters" on the market that don't come close to 99% efficient at 20 microns. Your statement might be more true if it was more like "any old filter" can get 90~95% at 20 microns.

Thank you.

Please let me clarify a couple things:

1. No, It's not intended to be a test of flow rates. Too any variables.
At no point will I state that if filter A fills 3 x faster than filter B, then your engine will see 3 x more oil flow.
What I am trying to show here, for example, is that filter A's media will create a bigger pressure drop than B's, at the same flow rate and viscosity--which in turn is more than filter C's pressure drop (resistance).
It would be silly to not include this information, IMO.

2. Osmosis doesn't enter into the physics here. Osmosis is based on diffusion of solutes from high to low concentration through a membrane on which both sides are covered with solution.

Thanks for reading and for your input though.

What I'm saying, is that efficiency ratings are way too misleading, because all the numbers look high. 90% sounds like a high efficiency, but it's very poor. And there's a big difference between 99% and 99.9%.

And if automotive filters can't do Beta20 > 100, then that's pretty sad. I didn't realize how poor automotive filtration was.

With language: I understand it's sometime neccessary to censor, but odd that that word is considered foul.

Perhaps I will install an oversized Hydac filter on my work vehicle, and monitor differential pressure to see how long my chosen filter becomes clogged. It's definitely not a cost senstative application, but it could be a worthwhile experiment.

Don't get me wrong, Gen1. I'm all about fine filtration. I've just sliced and diced it from every angle I can think of. I've always slammed up against one of the many "triangles" that tend to form in many (for lack of an all encompassing term) "processes". These triangles tend to be immutable.

You can reduce wear with finer filtration. That's easily reasoned by just inverting the simple concept that no filtration will result in higher wear. But it's something like the Laughing Curve (Laffer Curve) in that if lower taxes yield higher revenue, then zero taxes should yield infinite revenue. Bringing that back into a plausible scenario, one would have to say that next to no tax should yield tremendous revenue, but the size of the tax base has to be nearly infinite ..which it isn't.

That is, in our scenario, while fine filtration will indeed reduce wear, the scope of how wear effects longevity would have to be measured beyond our (even exceptional) span of ownership. To see the benefit, one would have to own the thing over a couple of generations. Taxis can already get 400k with standard filtration. That's simply a product of the elimination of most of the unavoidable wear. That wear is Castrol's "90% of all wear occurs at startup", which is the first 20 minutes of operation. A pan warmer and block heater would eliminate most of that and would prove beneficial even for someone in Texas ..but that benefit would be hard to extract in added or prolonged utility.

We'll retire even cleaner and more OEM spec engines to the junkyard.

Excellent points, but those taxis probably run regular dino oil too. One could argue the requirement for synthetic fluids to be nearly as useless as ultra-fine filtration.

But just like synthetic oil which has higher performance in addition to better longevity, I'm prone to believe that using a large aftermarket filter will reduce filter change-out frequency.

On topic...how much dirt holding capacity does the average automotive filter have? Also, has anyone seen or heard of a remote filter location kit for a Mazda3? They have drop in elements, which makes it difficult to modify.

Board rules are made by the owner ... either abide by them or don't post.

http://www.bobistheoilguy.com/forums/ubb...969#Post1600969

http://www.bobistheoilguy.com/forums/ubbthreads.php?ubb=showflat&Number=1560876#Post1560876

what rob is doing is obviously not top of the line testing with state of the art equipment but it gives some sort of insight on how these filters perform,and until someone steps up to the plate and can do better this works for me...

go rob!!!!!

My hydraulic filter sitting here alleges 8gms holding capacity. It's a Beta10=2/20=20/22=75 filter. Purolator allegedly has a 12gm holding capacity ..at least one number.

Yes, bigger filters of like efficiency would tend to reduce change out frequency if one is exhausting your filtration before your oil fatigues ..or you're willing to swap out the sump without changing the filter. Cost is the ultimate governor there.

I think there's a spin-on conversion for your Mazda if it has a Ford counterpart.

Thank Dan. Well you know it's a comparison.
Some filters elements stop more dirt than others and its not hard to prove that in a basement. Buy the one that keeps stopping more dirt, even after it's been dunked many times and plenty loaded up with heavily fine particle contaminated oil.

Its pretty simple: What is on the market is what we can choose from it wouldn't matter how the factories rated them--A, B, C, or whatever.
Buy the one that keeps performing well in actual filtering and is solidly built. This is the missing element in those filter tests/examinations out there in internet land. None of these I've seen ever actually filtered anything with the filters.

Here is an interesting experiment I did today. There is a test for mean pore size called a bubble point test. The test, when accurately done with fluids of known characteristics, actually allows the tester to calculate average pore size in microns. This test is used bothe with membrane filter in biotech, and depth media for mechanical applications such as ours.
The theory is that when a filter media is thoroughly wetted with a fluid (usually alchohol or pure water) the air pressure it takes to break the surface tension of the fluid is proportional to the pore size.
In practice, a soaked filter element is submerged in the fluid with air pressure on one side. The air pressure is increased gradually until, at a fairly sudden point, a steady stream of bubbles is produced.
Interestingly, the pressure needed to reach this bubble point is relatively independent of the amount (square inches) of filter media used. This is completely different than measuring pressure drop when fluid is flowing--which does depend on amount of media surface area.

Well, I'm not setting up a laboratory at home when a comparison will do.
Submerged in a pan of WD40 (yes I buy it by the gallon) air pressure needed to burst into the bubble (more of a fizz) point was approximately:

PureOne = 5.5 to 6 inches water column

Mobil 1 = 4.5 " "

Denso = 2.5-3.0 " "

Wix = Disqualified because two in a row leaked air badly at the top junction of the end cap and media. Disturbing.

This surely seems to correlate with relative the filtration of these three.

These tests were as accurate as I could do with my Dryer manometer and pressure regulator. Not perfect but ballpark accurate. A little trick because I have to apply a bit of air pressure to the core of the filter to keep the inside empty of fluid for a valid test.
Slighty pressurized, I lower it into the pan and start running up the air pressure.

Great testing. Grease's filter study used a real bubble test machine at PALL Filtration. A fellow member, Schultz, is an engineer there. The data is long lost. It was parked on a site for a while, but then went away. They even did flow rates at 10psi with swatches of the media.

That would be sweet.
We had a Pall filter validator when I worked in the biotech field, but they'd-a killed me if I used it.

Hello river_rat,

Unless you have more test results coming up shortly, would you mind summimg up again for how you would rank the filters based on your tests and observations?

Thank you.

he's got more in the pipeline, a RP and EAO that i know of? stay tuned......

Noted. Thank you.

Sure.

There are some variations, but overall I'd rate them from memory here
Filtration:
1. Delco UPF 52 (discontinued)
2. PureOne
3. Mobil 1 Ext Perf. and K&N, Mann
4. Wix/NAPA Gold and Wix built Delco Duraguard PF53, Mahle
5. Ecore style AC/Delco, Supetech, STP, and Fram Extraguard
6. Denso

Construction:
1. Mobil 1 and K&N
2. Wix/NAPA Gold, Fram Extended Guard, Mann, Mahle
3. Purolator and PureOne
4. Denso
5. Ecore style AC/Delco, Supetech, STP
6. Fram Extraguard

Easy oil flow:
1. Denso
2. Mann, Mahle, Purolator Classic, Fram Extraguard
3. Mobil 1, K&N, Wix/NAPA Gold,
4. PureOne
5. Ecore style AC/Delco, Supetech, STP
6. Delco UPF52

Not all test were done on all filters as you can see.

You can still get the UPF-52 rob,from GM dealer or Rockauto.com