AOS Diagram

[xmac]

I've had issues with oil ingestion issues on a track boxster with a 996 3.4 engine swap and an LN +2qt deep sump. The AOS was replaced with the real Porsche part but the white smoke problem continues. I contemplated AOS deletion over the motorsports version given there are many posts noting the super expensive motorsport version will fail as well and others that simply plugged the lines, noting it was purely emissions related. 

However, one post made me rethink when someone mentioned the AOS helps seal the low-tension piston rings used in M96 engines and that the mechanism is used to collect and reburn blow-by gases. I made a diagram to help me better understand why the engineers put in there in the first place and also figure out how oil actually gets sucked up. We know if it wasn't needed, Porsche would have removed it to save weight (like they do in every aspect of the car).

 

Looking for validation that the diagram is correct and and changes edits needed. Thanks in advance. 

 

 

 

 

Edited January 13, 2021 by xmac
added white background to image

[Silver_TT]

14 minutes ago, xmac said:

I've had issues with oil ingestion issues on a track boxster with a 996 3.4 engine swap and an LN +2qt deep sump. The AOS was replaced with the real Porsche part but the white smoke problem continues. I contemplated AOS deletion over the motorsports version given there are many posts noting the super expensive motorsport version will fail as well and others that simply plugged the lines, noting it was purely emissions related. 

However, one post made me rethink when someone mentioned the AOS helps seal the low-tension piston rings used in M96 engines and that the mechanism is used to collect and reburn blow-by gases. I made a diagram to help me better understand why the engineers put in there in the first place and also figure out how oil actually gets sucked up. We know if it wasn't needed, Porsche would have removed it to save weight (like they do in every aspect of the car).

 

Very interesting post, thanks for starting this discussion.  I just replaced the AOS on one of my cars last week so this topic is of interest to me at the moment.

 

It is definitely true that the AOS helps collect and reburn blow-by, which helps reduce costly emissions and increase efficiency.

 

When the AOS fails or starts to fail the vacuum pressure increases massively which can cause oil to actually get sucked into the intake and raises the possibility that you could destroy your engine by way of hydrolocking it with oil:

 

 

[JFP in PA]

58 minutes ago, xmac said:

I've had issues with oil ingestion issues on a track boxster with a 996 3.4 engine swap and an LN +2qt deep sump. The AOS was replaced with the real Porsche part but the white smoke problem continues. I contemplated AOS deletion over the motorsports version given there are many posts noting the super expensive motorsport version will fail as well and others that simply plugged the lines, noting it was purely emissions related. 

However, one post made me rethink when someone mentioned the AOS helps seal the low-tension piston rings used in M96 engines and that the mechanism is used to collect and reburn blow-by gases. I made a diagram to help me better understand why the engineers put in there in the first place and also figure out how oil actually gets sucked up. We know if it wasn't needed, Porsche would have removed it to save weight (like they do in every aspect of the car).

 

Looking for validation that the diagram is correct and and changes edits needed. Thanks in advance. 

 

 

 

 

 

 

On the track,  it is not uncommon to see one of these engines blow some oil smoke.  Sometimes this may only happen in certain corners, and not in others.  Problem stems from too much oil being retained in the upper engine area (read cam covers), where it cannot drain down fast enough through the M96's oil scavenging system, and the AOS becomes overwhelmed.  This is one of the reasons Porsche developed the X51 package, which uses a "northwest passage" extra scavenging system to get the oil back down where it belongs.

 

One trick a lot of track rats learn is to drop the oil level in the sump a couple of bars to help control the oil level in the heads.  On the street, we always filled customer's sumps to at least two bars below the max line for this exact reason.  We also tested everyone of the cars that passed through the shop for sump vacuum levels, as the AOS deteriorates and then fails a lot more often than people expect.  It should be considered a maintence consumable item that needs to be replaced before it begins to fail.

 

One comment on your diagram:  The actual intake vacuum signal at idle can be as high as 26-28 inches of mercury, and one inch of Hg is 13.6 inches of water.  The only time the intake would drop to anywhere near the levels in your diagram would be a wide open throttle.  So the AOS is not only separating oil and returning it to the sump, it is throttling the intake vacuum down to around 5 inches of water in the sump to help the low tension rings seal.

[Silver_TT]

 

3 minutes ago, JFP in PA said:

We also tested everyone of the cars that passed through the shop for sump vacuum levels, as the AOS deteriorates and then fails a lot more often than people expect.  It should be considered a maintence consumable item that needs to be replaced before it begins to fail.

 

I agree with this and never gave this much thought until I started to research it in depth a few weeks ago.

 

In the AOS I replaced last week in a different car I wasn't getting any CEL's and the car drove just fine.  But when I pulled the intake hose a bunch of oil streamed out everywhere.  When I inspected the AOS I could see it was no longer working 100% and allowing oil to bleed into the intake.  I will always replace my AOS every 40 or 50K miles going forward even if there is nothing wrong with it.

 

Also meant to say in my last post my 996 4S used to blow some oil smoke on the track, especially after the hardest laps.

 

  

3 minutes ago, JFP in PA said:

So the AOS is not only separating oil and returning it to the sump, it is throttling the intake vacuum down to around 5 inches of water in the sump.

 

As far as the throttling the intake vacuum down is this concept where the "breather" term comes from?

 

 

[JFP in PA]

A "breather" is simply and atmospheric vent, not unlike an open window.  The vacuum signal "throttling" done by the AOS is much like the thermostat in your home: It keeps the environment in the sump at a controlled level.  If there were no vacuum level, the low tension rings would not seal, pressure would build up, and engine oil seals would blow out.  Too high a vacuum signal and the intake starts inhaling oil as Jake showed in his video and the engine kills itself.  The AOS vacuum throttling function keeps the sump "just right". 😉

 

That part in the diagram marked "breather" is actually how the sump vents any pressure that might build up after the engine is shut down and there is no vacuum.  The pressure would wend its way around the system and back into the intake manifold, and back out into the world.  It all has to go somewhere; the engine cannot function as a sealed vessel. 

[Silver_TT]

Thanks.  Yes from what I saw taking apart the AOS the diaphragm with the tension from the spring is what throttles the vacuum to the specified level.  In the event of a torn/cracked diaphragm I have seen some people don't replace the part and only replace the diaphragm/spring.  While much cheaper, I personally think this is not a very good idea since I have seen problems caused by this where the diaphragm was a different material or the spring had a different size or tension, which can cause things not to work correctly.

 

There are several reasons actually but one of the reasons I replaced this part preemptively in my car is because when a AOS fails it can cause a lot of excess pressure to build up in the case which can damage the rear main seal (which is a PITA to get at).  I have never seen that cited on the M96 but in theory I would think a bad AOS on a M96 could blow out a RMS (or any other seal for that matter) too.

 

Nice diagram by xmac.  That's a good high-level visual of how it works.  Thanks for doing that.

[xmac]

Thanks for the info and inputs. I am working on a revised graphic. Posting soon.

[xmac]

Updated images:

Edited January 14, 2021 by xmac
re-ordered images

[Silver_TT]

One question I had in the revised drawings is on the hard turn I see a lot of the oil being forced over.  In my car there is a baffle in the sump which keeps the oil from doing this.  Does the M96 have this as well (I know xmac said he's running the LN deep sump)?

[xmac]

The baffle maintains oil near the pickup so that there is no interruption, but the oil outside will still slosh. I have the LN add-on with the windage tray that is supposed to further help. From what I gather from Jeff, I bought a +2qt oil pan to run less oil on the track by underfilling. LOL

Mac

[Silver_TT]

I just looked at what you have -- that is a nice setup.  Yes, you are correct it would still slosh outside the baffle.  I guess they are not mutually exclusive but I agree with you that is pretty funny about buying a +2qt only to underfill, it sounds funny at face value at least.

 

Bilt Racing Service BRS 2QT Deep Sump Oil Pan Kit MY97-08 M96/M97 inc. X51-Style Baffle

LNENGINEERING.COM

---

<p>The Bilt Racing 2 QT Deep Sump Kit provides similar oil pressure drop protection as an Accusump kit, but takes oil control to a whole new level. We've had Pro Racing teams use our kit to solve their oil starvation and aeration problems completely, when

 

[JFP in PA]

2 hours ago, xmac said:

The baffle maintains oil near the pickup so that there is no interruption, but the oil outside will still slosh. I have the LN add-on with the windage tray that is supposed to further help. From what I gather from Jeff, I bought a +2qt oil pan to run less oil on the track by underfilling. LOL

Mac

 

As each line on the oil level gauge is a small fraction of a liter (somewhere around  0.2 L), by adding a two quart (or liter) sump extension, and then lowering the level two bars from the full line, you have still increased your total oil capacity by around 1.6 L; and as any increase in oil capacity is generally a good thing, you are still ahead of the game. 😉

[Silver_TT]

On 1/13/2021 at 2:26 PM, JFP in PA said:

We also tested everyone of the cars that passed through the shop for sump vacuum levels, as the AOS deteriorates and then fails a lot more often than people expect.  It should be considered a maintence consumable item that needs to be replaced before it begins to fail.

 

Hey JPF, I wanted to know more about how you tested the cars that passed through your shop.

 

Did you use a digital manometer?  I have never used one before but wanted to try it.  It looks like you just need to hook it up to the vacuum line somewhere (can use oil tube for example if you can seal the connection) to have to be part of the pressure system for the crankcase.  When I look at images of the device it looks like it has two hose inputs as opposed to a traditional analog vacuum pressure gauge which only has one hose tube.  Would you put both of those hose lines into the sealed oil cap/tube?

 

Just curious how exactly you did this because I understand how it works in principle.   After connecting it I'm assuming just start the engine and let it reach operating temp.  I was going to try this on my car at some point just to test my brand new AOS for fun.

 

Edit:

I think I answered most of my own questions by finding a RennList post on this:

Air Oil Separator AOS Question - Page 3 - Rennlist - Porsche Discussion Forums

RENNLIST.COM

---

996 Forum - Air Oil Separator AOS Question - Hello all. I think I'm going to have to replace the AOS. The bummer is that it was replaced 3 years ago at around 70K miles. The car now has 103K miles on it. It is a 2002 C4 Cab. There is a lot of vacuum on the oil fill tube when the car is...

 

The only thing I can't tell from the picture is what the the second hose input (circled in red) connected to, if anything?  Upon further review it seems this manometer I showed specifically has dual pressure connection inputs -- but it isn't required, just a feature.  Some manometers just have one connection, for example, from what I can see.

 

 

Edit 2:  Did Jake ever release the improved AOS he talks about in the RennList link back in 2017?  Just curious as he indicates it could be any day now 3+ years ago.  I did not see it on his website, maybe it's on LN?  Curious what specifically he did other than bench test them out of the box -- shocking quality control to have them failing out of the box (Jake mentions, for example, glaring errors such as the factory forgot to glue the seams together).

Edited January 19, 2021 by Silver_TT

[JFP in PA]

Your photo happens to be the exact digital manometer I use, bought off of Amazon for about $35-40 at the time.  Only one tube needs to be connected the the vacuum source, the other functions as an atmospheric pressure reference that you are measuring the vacuum level against.  I also used a Porsche oil fill cap, but did it a bit more elegant than the one in your photo; I purchased brass barbed bulkhead fittings of Amazon and use O-rings on each side to seal it.  As every car out there has an oil fill cap, you can make up as many vehicle specific testers as you need.  To test, you need to first make sure the vehicle is thoroughly warmed up, testing it just after a 15 to 20 min. drive is optimal.  The reason for this is you will get some erratic data if the engine is not warm in in normal running condition.  The test sequence is simple and quick: Connect the manometer to the engine and turn the manometer on; set the units scale to inches of water, zero it if it has that capability, then start the car and let it run for about 1 min. to let the manometer stabilize, take your reading.

 

Yes, Jake did develop a "better mousetrap" AOS, but it never saw the light of day, and not for reasons of cost, or because it did not work.  If memory serves, Jake discovered that third part emission's hardware manufacturers are required to submit their devices for independent testing, at their own cost.  The testing takes a lot of time, and is expensive as Hell, and once you passed you had to get very expensive liability protection for it as you could be held responsible for damaging the environment, and fined millions of $ if it failed.  Net result was that it did not make any sense from an economic stand point as you would need to sell them for a lot of money in hope of recovering your sunk costs before you died of old age.

 

Sometimes, the laws designed to protect the environment totally suffocates innovation  …. 😬

[Silver_TT]

Awesome, thank you.  I have read several "reviews" that test this exact digital manometer you have against both a professional field unit as well as the U-type liquid unit and they all said it gives results that are both very accurate and precise.  I'm going to buy another oil cap for my car and do this for fun one day when I'm bored.  The idea would be to also do it every 6 months/5K miles/whatever going forward to see how the unit performs over time.  Obviously you want to be sure you don't have any vacuum leaks that would "pollute" the reading -- in my case, don't think I have any as I have looked the engine over obsessively to check for stuff like this.

 

That is really too bad about the improved AOS.  Seems like there's enough liability in just rebuilding an engine with OE parts -- he talks about it in that link, brand new installed factory AOS unit worst case is it goes out on a customer car that goes back home across the country and then dies resulting in Jake having to pay a local shop out of pocket to fix the faulty AOS and sometimes they try to take his lunch with it (to this end you can see why a Flat6 rebuild isn't cheap compared to some of the other people out there who frankly just don't stand by their product in the same kind of way).  Obviously the environment is incredibly important but it's just too bad in this particular case that the laws aren't written better in this regard for innovation.

 

Thanks for the good information, you have so much of it.

[JFP in PA]

If you had any pronounced vacuum leaks, your air/fuel ratios would go haywire, and eventually your computer would reach the end of its enrichment capability, both of which would trigger codes.

[Silver_TT]

Yes, I definitely agree.  If it's a serious vacuum leak on my vehicle it usually results in a CEL for oil pressure and/or the AFR being too lean.

 

You have mentioned before that for every car that came through your shop you did this vacuum test and I know you also load tested batteries (which is a common practice among good shops, at least the ones I know).  Is there anything else you had on your "list" each time a car rolled through aside from these two things?

[JFP in PA]

We would also visually inspect several things, like the brakes for thin pads, worn rotors, etc.; freeze point pH and clarity of the coolant; moisture level in the brake fluid,  wiper blade condition and washer fluid level, oil level and condition, any fluid leakage (oil, coolant, power steering, brakes), abnormal tire tread wear and tread depth, exterior light functions (do they all work), and signs of physical damage like dents or rust on the lines under the car or the exhaust system.  All this takes just a few moments to do, but can prevent the owners from getting stranded or stopped because a brake light is out.

[Silver_TT]

Thank you.  Didn't mean to drive this thread off topic but I just had to ask because some of your posts have caused me to think about things in a way that have subsequently allowed me to make important discoveries.

 

I am effectively mirroring what you your shop does.  Coolant freeze point pH is a good one, I'll add that to my list.  Thanks

[JFP in PA]

Get a coolant refractometer, uses one drop of fluid and is deadly accurate.  Amazon sells them for around $20 or so, and you can get pH test strips for a few bucks for a jar holding 20-30 that also only need one drop.  You don't always need zillion dollar tools to do the job correctly. 😉

[Silver_TT]

I just looked up the coolant refractometer.  I had never seen that before.  Looks like it can test maple syrup sugar content too, wild (don't worry, I won't use the same device 😛).

 

The following picture is what I believe you see.  What value range are you checking for?

 

Also should I think of measuring the pH level an alternative to using a refractometer?  What pH level would be equivalent to the specified ethylene or propylene glycol level?

Edited January 19, 2021 by Silver_TT

[JFP in PA]

On the bottom right side, you can see the zero C line (your unit is calibrated in C, you can also get them in F, which is more convenient):

 

 

  When a drop of pure water is placed on the prism, the blue line would be at the 32 F degree line, meaning the test fluid will freeze at that temp.  A 50/50 mix of distilled water and antifreeze would put the blue line around -50 to -60F. Note you have two antifreeze scales, one for ethylene glycol, the other for propylene glycol based coolants, so you need to know which type you are using to get the correct scale value.

 

This is the PDF manual for the Robinair 75240 model I own, which tells you how to calibrate and use the unit: https://www.robinair.com/sites/default/files/75240.pdf

I've had mine for literally decades, and it does the job every time. 

[Silver_TT]

Great explanation, I understand now.  One thing I wanted to ask you is that many coolants come premixed these days with the water so you just add it as-is, no need to mix with water in a specified ratio.  So you know when you put it in the ratio is spot on.  What is causing that ratio not to stay stable over time?

 

Also just to be sure I understand, the pH test is an alternative to the refractometer, correct?  In other words you did either or, not both.  Not sure if it's true but I read that using the pH strips is what you use for propylene glycol (check for pH under 7 indicating acidic) and for ethylene glycol coolants you test with a hydrometer.  If everything I stated is correct it seems the simplest and best to just use the refractometer only and just know what kind of coolant you have (it's listed on the back of the coolant).

 

Thanks for all this information.

[JFP in PA]

Antifreeze is a lot like oil, it has an additive package that controls pH, inhibits corrosion, and lowers the surface tension of the mixture.  And like oil, these additive packages break down over time and exposure to heat and metal surfaces, which is why the term "lifetime" antifreeze is a joke.

 

The pH of antifreeze solutions will buffer at a slightly basic pH, over time as the additive's break down, the pH range will shift to slightly acidic, which is not good, particularly in light alloy castings like most modern engines.

 

pH and refractometer freeze points are typically independent of each other, the pH comes from the buffer system in the additive package, the freeze point comes from the glycol concentration.  The only time they both move in one the same bad direction is when the pH becomes so acidic that it starts to break down the glycol, the pH drops, and the freeze point rises.  This is why we tested for both.

 

A hydrometer is testing for specific gravity, which is exactly what the refractometer is doing as changes in solution densities tends to bend the light in a prism more, which is how you get the reading.  Physical chemistry 101.......  Check the pH and use the refractometer 😉

 

And to comment on antifreeze premixes; I prefer to mix my own because I do not know the water quality of the pre mix, and that matters.

[xmac]

3 hours ago, JFP in PA said:

Check the pH and use the refractometer 😉

 

Thanks for the tip. I just ordered a Robinair 75240