Best filter for preventing start up rattle???

[3311]

Originally Posted By: dnewton3


As a mass generalization, the brand/grade topics are moot, at least to me. Same goes for "normal" full flow filters.



dnewton,
I know you have stated in the past you have a large library of uoa's. Have you ever done any analysis of full flow filter brand performance? It would certainly be interesting.

 

[dnewton3]

No, but only because the data in UOAs is predicated to the lubes, and the filters are after thoughts.

I try my best to coordinate UOAs here and track not only lubes, but filters used.

As a generalization, filters don't seem to affect wear much at all. The tribochemical anti-wear layer is what seems to most affect wear.

The famous (or infamous) GM filter study did show correlation in filter efficiency and wear, but that was only AFTER then negated the effects of OCIs and heavily dosed the sump with fine particulate to purposely over-load the additive package. At the end of the study, they even acknowledged that one would never find correlation in real world analysis because the effects on wear (using all three control methods) cannot be delineated when all three methods are used concurrently. IOW - the effect of a normal filter in a normal system with normal oil is so small that the "normal" variation of wear metal data is greater than the effect of the filter by itself. Hence, in a typical UOA, we will never see statistical evidence that a PureOne or M1 is "better" than a Wix or a Purolator Classic; the wear data differences are just too small to attribute with clarity.

 

[dnewton3]

Originally Posted By: BlueOvalFitter
Originally Posted By: ZeeOSix
I think the discussion here between xW-20 and xW-30 grade was focused on the viscosity difference for cold start-up flow. Not really talking about the wear aspects between them. If an engine manufacturer specifically calls out 0W-20 there very well could be a reason related to good cold oil flow on start-up to prevent engine rattles.

It was 29* here this morning.Upon startup my engine just did its usual slight ticking,then faded away as it starts to warm up,as it has done since I bought it new.
I'm now leaning towards doing an oil change with a semi-synthetic or synthetic oil and/or a different brand filter;P1,M1,Bosch,or WIX. (I even thought about a FRAM XG

)



I would caution against this approach. Not that you cannot do it, but if you do have success, how would you know if it's the lube, or the filter??????

Change only one variable at a time, and experiement over a few OCIs. Don't take a "shotgun" approach to the experiment, or there is no rational conclusion to draw.

 

[BlueOvalFitter]

Originally Posted By: dnewton3
Originally Posted By: BlueOvalFitter
Originally Posted By: ZeeOSix
I think the discussion here between xW-20 and xW-30 grade was focused on the viscosity difference for cold start-up flow. Not really talking about the wear aspects between them. If an engine manufacturer specifically calls out 0W-20 there very well could be a reason related to good cold oil flow on start-up to prevent engine rattles.

It was 29* here this morning.Upon startup my engine just did its usual slight ticking,then faded away as it starts to warm up,as it has done since I bought it new.
I'm now leaning towards doing an oil change with a semi-synthetic or synthetic oil and/or a different brand filter;P1,M1,Bosch,or WIX. (I even thought about a FRAM XG

)



I would caution against this approach. Not that you cannot do it, but if you do have success, how would you know if it's the lube, or the filter??????

Change only one variable at a time, and experiement over a few OCIs. Don't take a "shotgun" approach to the experiment, or there is no rational conclusion to draw.

Very good advice.I'm going to start with the oil.I'm going to try an 0W20. Does this flavor come in the QSUD?

 

[dnewton3]

Excellent choice.

Allow me the latitude to be a bit pedantic and almost motherly here ...

My perception with some of the problems with this site (or more specifically its membership) is that we don't hold each other accountable for sane, rational decisions. Too often we just pat each other on the back, and reinforce any action without asking if it makes sense.

I've made mistakes before; I've been wrong. I don't mind when someone points that out, as long as I can get some decent solid advice as to how to change my thoughts or actions. I may or may not choose to act upon that advice, but I do appreciate when it's offered.

My concern with your initial offering was that if you did solve the "rattle" problem by changing oil/filter, you'd have no idea what fixed it. And that, in turn, really isn't helpful to the next guy in similar position who seeking similar advice.

My perception of the "best" application of BITOG is that we become a group that offers good detailed data on experiements we do (or have access to), and challenges those experiments when they go awry. Facts and data should outweigh opinion and rhetoric here.

In your quest, you should experiment with lube grades and filters. But just be thoughtful and thorough, and not hap-hazard, if you want it to be meaningful.

 

[KCJeep]

Originally Posted By: dnewton3

My concern with your initial offering was that if you did solve the "rattle" problem by changing oil/filter, you'd have no idea what fixed it.


Exactly, which is why I also suggested he just change his oil selection.

I've also struggled with start up death rattle recently, but extended my OCI and changed the filter ONLY in order to make an accurate assessment of any changes.

If you change more than one thing at a time you are just chasing your tail.

 

[3311]

Originally Posted By: dnewton3
No, but only because the data in UOAs is predicated to the lubes, and the filters are after thoughts.

I try my best to coordinate UOAs here and track not only lubes, but filters used.

As a generalization, filters don't seem to affect wear much at all. The tribochemical anti-wear layer is what seems to most affect wear.

The famous (or infamous) GM filter study did show correlation in filter efficiency and wear, but that was only AFTER then negated the effects of OCIs and heavily dosed the sump with fine particulate to purposely over-load the additive package. At the end of the study, they even acknowledged that one would never find correlation in real world analysis because the effects on wear (using all three control methods) cannot be delineated when all three methods are used concurrently. IOW - the effect of a normal filter in a normal system with normal oil is so small that the "normal" variation of wear metal data is greater than the effect of the filter by itself. Hence, in a typical UOA, we will never see statistical evidence that a PureOne or M1 is "better" than a Wix or a Purolator Classic; the wear data differences are just too small to attribute with clarity.
Thanks for the info.

 

[ZeeOSix]

Quote:
As a generalization, filters don't seem to affect wear much at all. The tribochemical anti-wear layer is what seems to most affect wear.


Study showing the correlation between wear and particle size, and the importance of a good oil filter.

http://www.amsoil.com/techservicesbullet...gine%20wear.pdf

 

[BlueOvalFitter]

I did some thinking about my situation today. My oil filter is positioned on the front of my engine at a 4-5 o'clock position. Being that it's in this position shouldn't the filter always have oil in it upon startup? And,the ADBV would sort of be redundant in this case?

While in WM tonight I was going to buy some MC 5W20 oil but they were out of the 5qt containers. Instead I bought a 5qt container ($17.27) and 1 qt ($4.97) of 5W20 Castrol Syn-Blend oil.

I'm going to get my cousin to help me change the oil and filter this weekend and will update Monday. It's supposed to be 28*-32* that morning.

 

[BlueOvalFitter]

My friend helped me change my oil earlier. The engine sounds somewhat quieter,but it could just be me wanting it to sound that way.The real test will be in the next few mornings.The temperatures are supposed to be 28*-35*.I kept the filter to cut open,one day this week.I will keep you posted.

 

[BlueOvalFitter]

31* at 6am this morning.Upon startup,no noises.

 

[Jim Allen]

Thanks for the update.

To be clear, you changed oil but not filter?

And you went to a 5W20 but what was the original oil viscosity?

 

[BlueOvalFitter]

Originally Posted By: Jim Allen
Thanks for the update.

To be clear, you changed oil but not filter?

And you went to a 5W20 but what was the original oil viscosity?


I originally had MS5K 5W20 in it.I drained that and installed Castrol SynBlend 5W20 and a new MC FL-400s filter.

The MS5K was the first time ever my engie had conventional oil used in it,since I had bought it new. (6 miles when new) Currently 93K miles

 

[Jim Allen]

Ok. For some reason I thought you had a heavier oil and were going lighter to test how much that aspect had to do with startup noise vs the oil filter... which would have been an interesting experiment.

 

[BlueOvalFitter]

The last couple of mornings it's only been 40*-45*.The engine is as quiet as a church mouse upon startup. This Sunday it's supposed to be in the 20*s so I will post more results then.

To be continued.

 

[BlueOvalFitter]

25* this morning.Started my truck,quiet as a church mouse!

To be continued.

 

[ZeeOSix]

Sounds like you got it licked.

 

[dnewton3]

To really see if the filter or oil was the solution, switch back at the next OCI and see if you can re-create the noisy condition. Then switch back yet again at the following OCI to make it go away. If you cannot recreate the condition, it was a mere coincidence. Being able to replicate the conditions assures control of the variable.

Don't presume that because it went away, you controlled the variable.

 

[BlueOvalFitter]

Originally Posted By: dnewton3
To really see if the filter or oil was the solution, switch back at the next OCI and see if you can re-create the noisy condition. Then switch back yet again at the following OCI to make it go away. If you cannot recreate the condition, it was a mere coincidence. Being able to replicate the conditions assures control of the variable.

Don't presume that because it went away, you controlled the variable.

So,you're saying to install the MS5K oil,which is what I had in the engine when it started the rattle?
If that's not it try another brand of oil filter?

 

[dnewton3]

To be honest, I've lost track of what position you were first in, so I cannot comment as to what you need to use.

But - I think you're on the right track. If I understand you correctly, you changed lubes, but kept the same type filter, right?

What I'm saying is that you had some combination of products (some oil and some filter) that was presenting a "rattle" at startup.

Now, you've changed one or more variables, right?

To confirm that you have "control" over the variable, you need to be able to make the condition come and go with switching the variable.

If you had MS5k in there and that is what rattled, then try it again on the next OCI and see if it returns. You need to keep the conditions as similar as possible. Then, yet again, swith out of that oil and see if the rattle leaves.

There is a HUGE difference between correlation and causation; very much misunderstood by most folks. Even though it's a subtle concept, it makes a big differnce in root cause.

If you can make the problem come and go, by swapping the fluid from one product to another a few times, then you can claim you have "solved" the issue. Otherwise, it's just coincidence.

Also, the filter plays into this. If you recall a few pages back, I suggested to make sure you don't "shotgun" the approach by changing both fluid brand/grade and the filter brand/type at the same time, because you would not be able to attribute any success to any one change; I presume you followed that advice? So, after experimenting with the oils, only then try the filters.

In a nutshell - you had a rattle and changed lubes and the rattle is gone, right?
First, run the lubes back and forth a few times and see if you have "control" over the issue.
Then, run the filter choices back and forth a few times to see if that has any effect.
If you can make the rattle come and go by controlling the variables, then you have driven down to the root cause and can claim victory. If you don't do this, it is merely correlation in that you made a change and a potential for causation has not been confirmed.

It could even be a combination of the two topics, but that GREATLY widens the experiment. So start with the easy stuff first.

That make sense?