grover

They will dim like that whenever the car is off; do they change to normal intensity while the car is running, or remain dim? Could be a bad contact in the headlight switch; it's a notorious failure item, and the low beam contacts get the highest workout of anything in the switch.

Directly above the yellow circle is a small black plastic vacuum line that connects to a black rubber junction fitting. These are the lines I'm referring to.

I inspected these pretty closely, and didn't see anything amiss, aside from not seeing the check valve you mentioned. Where should it be?

I think you might want to go back and read the codes again and write them down.

418 exists, 419, 491 and 492 don't in my factory 5.2 and 7.2 DME docs for 986s.

It is usually a good idea to post the model and year of your car when posing code interpretation requests as it can make a difference..

Mike, that's why I got confused and said 418/419. But 418/419 were NOT the codes, 491/492 were. They're real codes, the reader says "secondary air injection bank A" and "secondary air injection bank B" but they're not in the Bentley guide. The 996TT thread I linked to above lists the codes as legit Porsche codes. In their case, it was a vacuum leak, caused by various failures of parts, roughly half of which don't exist on the 986S, but some do. I double-checked and cleared the codes again today and they haven't come back (yet), but always have within a few days of driving. (I'll take photos next time if you still don't believe me!) 5 months ago, I thought it was just excess oil from the AOS failure working its way through, but that was 6 months ago so I'm having a hard time convincing myself it will just go away for good if I ignore it long enough...

There is a main vacuum line from the intake that connects to a junction fitting for the vacuum tank,secondary air and intake flap that can easily be broken or not connected. It is located very near the AOS so inspect it for damage.

Is it the fitting right above the yellow circle in this photo? I disconnected it when I changed the AOS; I just inspected it and saw no flap, just a quick disconnect fitting. Nothing on the intake manifold side but a hole, either. Should there be a flap here? There's no signs of any oil in the line at least, that's always a good sign. The pump came on for about 40 seconds after I turned the car on. Exhaust was clear while the pump was running, but a bit of smoke (normal exhaust smoke I'm used to seeing on any car at startup) after it shut down.

Edit: and it's actually CEL 491 & 492, not 418. And I found another thread about it on the 996 forums that sheds a bit of light. Man, I hate troubleshooting vacuum leaks.

Given that the pump is running and exhaust is clear while it's running and only started afterward, can I assume the vacuum valves are actuating properly and the secondary air system is, indeed, pumping air into the headers, and the problem is somewhere else? Would a missing flap or a vacuum leak allow the vacuum switch to work properly, but cause other issues?

Double post

Edit: 491 & 492, not 418/419. (Should have written it down instead of thinking I'd rememebr it!) 2004 986S.

I replaced my AOS in July (mostly squealing vacuum in the crank case as the symptom with very little oil in the intake) and the oil that did suck into the intake manifold has long ago worked itself out, but CELs 0491 & 0492 come back every time I clear them. The timing can't be a coincidence (well, I suppose it could...), but I can't figure out how the AOS could have caused any issues with secondary air injection. Nor do I really understand what these codes mean, besides "secondary air injection". The code reader says 0491 is Bank A, and 0492 is Bank B. I'm pretty sure I can hear the air pump's whine on startup. The 40A fuse in the trunk is OK. Concerned that I might have knocked something loose during the AOS repair, I've double-checked the connectors on the pump and they seem good. The secondary air injection valves are buried, though, and I haven't dug them out to check for residual oil; logic applies that the vaccum should have been pulling oil OUT of the system, though, not sucking it in.

Q1: What are Banks A and B? Left and right cats? Signifying an issue with a common point of failure between the two?

Q2: How is the computer determining there's a failure? Could it be something completely unrelated causing the issue that the computer is misdiagnosing as secondary air injection?

I think the beauty of that tool is that it works well in very tight spots because once that thing is on the clamp, it keeps it open. You can then maneuver the whole thing into the correct position and once you release the blue clip, the spring clamp tightens right at the spot you want it.

Problem there is that it's going to be just as much of a ***** to remove as it was before; the blue tool won't help you there. I like worm-gear pipe clamps better for something like this.

Is there any way to simply bypass/disable the AOS? I can't help but think one could simply disconnect & plug the intake manifold hose from the AOS, and leave the AOS output open to vent crankcase gasses... and it would be fine.

Impossible to say without reading the codes. Take it down to your local auto parts store; they'll read the codes for free.

Please excuse my stupid questions.. but How a head gasket can lead to increase in coolant pressure?

The only thing that should happen is coolant mixing with oil and possibly.. less pressure in coolant..

Thanks

Sasha

If the gasket fails between a cylinder and coolant chamber, it will draw coolant into the cylinder during the intake stroke, and blow hot gases into the coolant system during compression and power strokes. Symptoms are loss of coolant with no drips, and increase in coolant system pressure.

Maurice, good eye, thanks, I'll check that out; and thanks to everyone who replied, I really appreciate the help! Props to Bap Geon, too, who got me a replacement AOS within 4 hours, and didn't complain when I realized I couldn't use the bellows I'd special ordered.

:thankyou:

Since I couldn't find any write-ups on the 03/04 Boxster AOS online or in the Bentley manual, I went ahead and posted an article in the DIY forum. Hopefully this will help someone:

http://www.renntech....04-boxster-aos/

I know it's "required" for emissions, but can it be disabled, like by filling the top of it (where the vacuum line connects to the manifold) with epoxy, and drilling a hole near the top to still allow crankcase fumes to vent? Has to be better than spraying oil all over the place. I'd think the 3.6L AOS would work fine with the 3.6L, though. What are you doing to those poor AOSs??

This article describes how to replace the emissions system air-oil separator (AOS) on a 2003/2004 Boxster/Boxster S, which is slightly different than on earlier model Boxsters. The photos are of my 3.2L '04 Boxster S Tip.

There are two good writeups on replacing AOS (with procedures for up to '02 Boxsters) that give background and symptoms and I recommend you read prior to doing this repair:

http://sites.google....atorreplacement

http://www.pelicanpa...l_Separator.htm

REMOVAL:

Open the engine access cover; jacking up the car is not necessary on this model year.

There is a quick-disconnect tube readily accessible near the top of the AOS, connecting to the Y-pipe to the intake. [yellow circle in photos] Squeeze the top and bottom of the quick-disconnect (the textured portions) to disengage the plastic barbs. Pull this away from the AOS.

Next, disconnect the lower quick disconnect tube visible from the top [purple circle in photos]. It's disconnected in the same fashion- squeeze the top and bottom, and push away from the AOS. There are a lot of obstructions that force a bad angle; you may need to lean over from the left side of the car or lay on the trunk to find a position with both grip and leverage.

The AOS is held in by two star bolts [red circles in photos]. These bolts are accessed from the top, through the engine hatch. You should be able to remove them with a standard 8mm 6-point socket and 12" of extensions, but you might need a special socket if they're too tight.

A rubber tube comes off the bottom of the AOS, turns 90 degrees, and enters the block [cyan circle in photo]. There are two #^%$#-style hose clamps on this pipe. Use vice grips, screwdrivers, plyers, crowbars, whatever you can do to get this off. With the clamp clear, you can work on the hose. Grip the hose with a pair of pliars and twist it to break it loose; it should slide off. If you can't get leverage, securely grip your pliars, and (gently) persuade it off with a crowbar. The AOS is now free!

Tricks that worked for me on my 04 986S tip and might inspire you: I stood behind the car, and leaned over the trunk to work from above. I clipped a vise-grips to one prong of the hose clamp, and used a screwdriver wedged against the structure of the engine to brace the other end. While using the vice grips as a lever to compress the clamp open, I jammed a crowbar against the vise grips and worked at it until I slid the clamp 1" back from the block. You might get lucky or actually have the right tool and get it off easy, but It took me about an hour of trial & error. Note: I also removed a line off the upper intake manifold and moved it out of the way to get sufficient throw for the crowbar.

View from the right-rear tire well, with tire removed and car on jackstands:

INSTALLATION:

Old AOS I removed: 996.107.026.00

New AOS I installed: 996.107.026.01

Old on left; new on right.

Fortunately, installation is very easy, almost trivial. If your hose is in good condition, remove it and install it on the new AOS. (Take note before you remove it, so you can reinstall it in the same direction.) The hose should be pointing in the same direction as the top-most tube. Do yourself a favor, too- throw the old spring clamps away, and buy some stainless steel worm-gear hose clamps. Tighten the clamp that attaches it to the AOS- be sure to align it so that it won't foul anything when you install it, and so you or the next mechanic might be able to reach it with a screwdriver from the wheelwell. Fasten it tightly; this is fairly low pressure, but you don't want a leak.

Slide your 2nd hose clamp on loosely; align it so that it won't foul, and you can tighten it. Hold the end of the rubber hose as you feed it down, and slip it over the nipple. When you get it about halfway on, you should be able to set the AOS in place, but get it as far on as you can. Position and tighten the hose clamp, reattach the two quick disconnects, replace the two bolts, and you're done! A magnetic socket or mechanic's magnet-on-a-stick helps with getting the bolts in.

Start the engine, and admire your handiwork. You'll probably get a bunch of smoke at first (and the next few starts) as the oil from before burns off, but it should dissipate after a few minutes. Any auto parts store can read & clear your CEL for free, so don't pay the Porsche dealer. [Mine recorded a bunch of cylinder misfires.]

Curious what's in there? Grab a hammer and chisel, and go to town! The base is hollow, in two separate and isolated chambers. You can peep through the nipples and see. The interesting bit is the diaphragm on top:

Done and running! Reinstall sure was easy, too. Got the auto parts store to clear the CEL, and I'd swear she's a bit peppier than last week :D I'll see at the autocross this weekend, I suppose.

Curious what's in there, I grabbed a hammer and chisel and went to town! The base is hollow, in two separate (and isolated) chambers; there's nothing in there at all. You can peep through the nipples and see. The interesting bit is the diaphragm on top. Now, it looks like vacuum from the intake will pull the diaphragm shut upon start, but that high pressure from the crankcase will force it back open, allowing gasses to vent, but I'm unable to puzzle out how it excludes oil. Clearly, though, if it sticks open, the crankcase will be under vacuum, and oil will get sucked into the intake.

Was a royal pain in the ***, but I did get the old AOS out last night and am ready for the new. It's a significantly different part and a different procedure to attack, but still uses the same $^$%^#^ hose clamps as the older models' AOSs that quite literally took an hour of jimmying and two crowbars to finally persuade free. I at least didn't have to resort to the lead sledge!

kburgo- 2003/2004 has a slightly different style AOS that I've not seen a page on yet, either. Guess we're just on the cusp of late-model 986 work, huh?

I've found everything is accessed from the top on my 04S, including the bolts. They're not 10mm, either, they're tamper-resistant bolts whose name slips me at the moment, but I was able to remove them with an 8mm socket and a 12" extension. I'm taking a break right now from fighting with the bottom tube, but I at least got the clamp off. Just need to get the hose off, now... (yes, hose and not a bellows.) Since it doesn't match the bellows, I'm not sure if I can find a suitable replacement, and I'm thus hesitant to cut it off. When I get everything done, I'll post an article with photos.

Here you go, full docs on the air-oil separator replacement: http://www.pelicanpa...l_Separator.htm

If you remove the oil cap while the car is running and the noise goes away, then it's most likely the air-oil separator.

-Wayne

Well, joy. My AOS is not the model shown on the pelican parts page, or Mike Focke's, though it does look just like the photo Mike posts where he says 03/04 are different. On mine, the bolts are accessed from the top, and the bottom half looks significantly different. Not that it's going to be difficult to figure out, but did take me a few minutes to realize what was going on, heh. I just hope the auto parts store has the right part for me when I go to pick it up tomorrow...

you do not need to drain oil. you can take off your manifold, runners, and throttle body to clean if you want but I didn't have the patience and time. I just let it idle for ~30min and rove around a bit b4 I put it back on the road.

get a generic hose clamp to attach the tube to the engine-side when you replace your AOS. i'm 5'11" and I got scraped and bruise trying to use OEM clamp and was unsuccessful.

Thanks. There wasn't any smoke the last 20 minutes or so of my ride, just squealing, so I should be good with that respect. I'm buying a new boot and will just cut the old one in half if I need to :) I'll have to see if I have any pipe clamps; I'm pretty sure I've got some laying around, but have to figure out where I put them!

The cab top relay may also be sticking; I can't recall if the window drop signal goes through it or not, though. Know anyone with a known good relay you can swap out to try? It's probably the easiest thing to check, though likely also the most expensive component in the system.

And yeah, you could replace every single part of the mechanism for cheaper than a dealer diagnostic, I think!

Risk averse? Bah, I drove my old camaro with a bad head gasket until it hydrolocked... 4 years and a lot of coolant later! No additional damage from the hydrolock, fortunately. As luck would have it, I finally finished replacing the headgasket (and then tearing it back down to readjust the valves I'd overtightened, ugh) two weeks ago- I can drive it to work tomorrow & pick up the new AOS instead of having to steal my wife's car for the day.

I made it home safely, by the way. It smoked like mad, but I think it was only oil that got sucked in when I pulled the oil cap to test it this morning, but MAN what a smokescreen! Cleared up after about 5 miles at which point it started squealing. I'm pretty confident it's the AOS, now. Gonna go ahead and pull the old one tonight, I think. I don't have to drain the oil or anything for this, do I? The only oil to clean up is what's coating the inside of the AOS?

2004 Boxster S. My engine started squealing on the way into work this morning- a horrible squeal that sounded like the serpentine belt slipping; stopped and started for a bit; generally went away when I revved the engine. It *sounded* like the sound was coming from the mid/lower right side of the engine- power steering pump? The belt isn't slipping and none of the pulleys appear siezed. However, I don't have a 24mm socket here and am unable to loosen the serpentine belt to verify all the pumps and pulleys are spinning freely. There is no white smoke, and no CEL, though I did see one tiny puff of white smoke a few weeks ago on startup (I know it's normal and thought nothing of it) and one CEL a month or so ago that came on and went away and never came back.

Troubleshooting seems consistant with AOS- I couldn't remove the oil cap while the engine was running, and struggled to get it off 10 minutes later. I did not hear the sound while the cap was removed, but heard a LOT of air whooshing through it. After about two minutes, got billowing clouds of white smoke. Wasn't sure if that was normal to see with the oil cap removed, or if it's "a sign".

Q: Does this sounds like DEFINITELY AOS?

Q: Can I safely drive home on it? 20 miles, maybe 45 minutes in rush hour traffic, though I may cut off a little early today to lessen my time in traffic.

Consumer reports relies on user reporting. This works out well when you have lots and lots of owners reporting, but not so well when the sample size is very small. They don't even report some model year Porsches some years because they don't have enough data points; the years they do have don't seem statistically reliable.

I mean, when we see graphical reports like AWESOME AWESOME AWESOME CRAP AWESOME on a particular item that didn't change at all between model years, it's pretty telling. I think it tends towards bad because Porsche owners aren't really the target demographic for consumer reports, and those with a bad experience are more likely to take the trouble to report than those with good experience. If you look at other quality surveys, Porsche's marks are all over the board- the source numbers just aren't large enough to be statistically relevant. Most of them tend towards casting Porsche as excellent in reliability- right up there with the best of the Japanese car makers. Considering how badly Porsches are flogged on a daily basis, I'd say even par is good, but they seem to be significantly better than that.

I think what you're really looking for are underside photos with the underbody panels removed! You could probably exit the rear of the trunk and follow the brake lines, as they go straight back the length of the car and are easily accessible with the panels off.

04 Boxster S:

I had the T-fitting by the battery crack and break loose and do this, too. May be obvious already from where the wet spot is, but if you remove the plastic in the front trunk to expose the piping and have someone work the wipers, should be pretty easy to see where it's coming from.

Assuming 996 tips are programmed the same as 986, I've found it gives an extra margin of safety on downshifts, too. For instance, I can get my 04 986S to about 70mph in 2nd gear before it upshifts to 3rd, but I can't downshift from 3rd to 2nd unless I'm under 60. It will automatically upshift at the redline, even in manual mode, which is nice, and I use it all the time. The new PDK 997.2s do the same thing- latest C&D track test, they said they completely stopped manually upshifting, and just let the car do it.