When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
I was down at Buttonwillow and ran hard all day Saturday w/ issues. On Sunday I did my first session and everything seemed fine. I started the car for the second session and huge clouds of smoke started pouring out of both tailpipes. Everything sounded fine and I only noticed when I looked in my mirror to back out of the garage. I shut it down and began inspecting things. No leaks, or anything so I pulled the air intake apart and found oil in the the plenum.
The tell tale signs of an AOS failure. My car has 80k miles on it so I figured it's time was up and a friend trailer the car home as I was low on oil and I was five hours from home and didn't want to take any chances. Mechanic replaced the AOS and confirmed that the membrane was damaged. He checked everything out and said everything else looked good.
I get the car back and put it through it's paces as best as I can on some canyon runs to see if I can replicate the issue w/o incident.
A month later I'm at Laguna Seca. The oil was at the top "Recommended" level marker. First session was fairly mellow as there were a lot of cars and the grid pecking order wasn't ideal. Car ran great. I parked in my garage slot and when I started for the second session.... a cloud of smoke. Both tailpipes just like before. It quickly dissipated so I ran the next session towards the back of the pack and took it easy. A friend got behind me and said there wasn't anything while driving. I started up for the third session and another cloud of smoke. At this point I called it a day and my oil measured about One Segment lower.
I took the car back to the shop and had a compression & leakdown test done for piece of mind and had it looked over with a fine toothed comb. Everything checked out fine and I also had the Mantis installed.
I ran an event at T-Hill a few weeks ago and the oil level was showing up around the middle of the recommended range. Between each session I got a very small Puff of smoke that quickly dissipated. By the time I got out of the car with my phone to take a picture, it was done. I asked around and had several people tell me that it was fine and to just run a little less oil. A race shop that is very well respected was there and he told me that they track prep m96 & m97 engines a full quart low at the bottom of the recommended range. That given my run times I'm just sloshing too much oil and overwhelming the stock AOS.
After all the day's activities and puffs of smoke, the oil indicator didn't decrease at all and in fact is still reading the same amount of oil in the car a month later.
I replaced the failed AOS with the stock one because the motorsport version required some makeshift plumbing and my mechanic wasn't 100% confident doing it and I figured the factory one got me through 11 years / 80k miles and about 30 track days. A bad decision on my part.
The standard oil pressure sensor and gauge are not accurate or precise enough to draw conclusions from.
The sensor is very heavily damped. Its connection to the oil system is by a very small hole, around .020" inch. This makes the sensor unresponsive to fast changes in oil pressure. The sensor is also not purposefully grounded. That means that normal tiny voltage drops in the engine ground or elsewhere will change the sensor reading.
I think the changes you are seeing in your road tests indicate that the sensor is no longer working perfectly. The oil pressure should vary only with rpm (assuming the oil temperature is constant) and not with load.
The M96 pressure sensor location is unusual in that it is placed at the very end of the oil system. This means that the sensor reads lower than the pressure that is being delivered to main bearings for instance. My data shows about a 1 bar difference between the two locations. My data also shows that any fluctuations in oil pressure caused by the pickup sucking air are equal at both locations.
I am certain that under some conditions on track the oil system is indeed not doing its job 100%, but the gauge readings from your test drive are probably not showing this.
That's about what I figured. I've always monitored the gauges especially on the track but ever since the AOS failure, I've been even more diligent and noticing more questionable readings.
How would I go about installing more accurate gauges? i don't want to tap the case but perhaps in a sandwich adapter on the oil filter?
The stock location works well enough once you install a decent sensor and gauge.
There are many options for this: Motec, AIM, and lots of other standalone gauges. I use Motec in my endurance racing Spec Boxsters.
In my opinion, a gauge is nearly worthless in a track car. Datalogging is necessary if you want to see what is really happening.
Idiot lights and alarms along with datalogging work best. This situation is tough to cover with an idiot light since oil pressure varies so much even when everything is working perfectly.
I think our question is, why does the oil pressure only significantly drop on right turns for both Boxsters/Caymans and 911s (9x6 and 9x7.1 generation), when the engine is flipped 180 degrees, but the oil pressure sensor location is in the right valve cover in both cases. For my case, in my 2000 Boxster S, on a warmup session at Road Atlanta, the oil pressure is dipping like crazy on right hand turns, even at very low speeds/G forces, not even enough to cause my street tires to squeal. Oil temp in the low 200-210 range (captured at the other port on the other valve cover). I can't believe that the oil pickup is sucking air at these low speeds/g forces, and I have the LN 2 QT deep sump. I don't have any "data" for faster runs because the camera got repositioned and didn't pickup the gauges.
I sincerely hope that it is some localized condition in the right valve cover causing the dips and that the oil pickup isn't sucking air on every single low speed turn.
I plan to install an oil gauge on a sandwich plate and toggle between the two readings while I am on track, and then watch the video later. I haven't gotten to that, and I would be willing to loan out my sandwich plate with VDO 80 PSI oil pressure sensor and 300 F oil temp sensor as well to anyone that is interested.
If the pressure read in the stock location is about 1 bar lower than the pressure at the crank/bearings, I think that is ok, even with the dips. What I hope to see is that the pressure read in the sandwich plate doesn't drop on every right turn like it does for the sensor in the valve cover
Chris can you share any data traces that show the correlation between the two oil pressure measuring points?
I think that having a good oil that is resistant to heat degradation is key. I am running Millers Nan 5W40 but haven't yet gotten far enough along to do my first UOA. Hopefully the additive package lasts for a while because it is $$$. With the recent sale of Mobile One 5W50 I am thinking of changing to that when my current fill is used up.
Thanks,
Steve
Last edited by steved0x; 09-22-2016 at 05:54 PM.
Reason: Typo
I found this diagram of the oil system for the 997.1 (m97) engine.
Am I crazy or does it show a second Oil Pressure Sensor on the right side of the engine?
OK, I know this thread is old but I have some updated info.
I've spoked with an engine builder and he revealed some issues with the m97 oil system that I believe are the root of the issue explained in this thread. Essentially, the scavenger pumps can't keep up with the main oil pump. The scavenger pump on the left side is at the front of the car which results in the oil not being returned fast enough in high speed corners because the oil is pulled away from the pump towards the rear corner. Left hand turns don't have the same issue because the scavenger pump is at the rear and so on turns the oil is pushed into the pump. So, the oil pools in the heads and starts running the sump low. Pair that with the oil sloshing to the sides under turn loads and the main pump starts sucking air and foam.
It was suggested by the engine builder that the if one could keep the rpms up high enough to keep the oil pressure above the relief bypass, that it allows oil to bypass the galley system and recirculate back into the case.
The stock location works well enough once you install a decent sensor and gauge.
There are many options for this: Motec, AIM, and lots of other standalone gauges. I use Motec in my endurance racing Spec Boxsters.
In my opinion, a gauge is nearly worthless in a track car. Datalogging is necessary if you want to see what is really happening.
Idiot lights and alarms along with datalogging work best. This situation is tough to cover with an idiot light since oil pressure varies so much even when everything is working perfectly.
Chris Cervelli
Chris' word is money, as usual.
I will add that NO data acquisition sensor manufacturer recommends hard-mounting to the block ANY pressure sensor.
ALWAYS a flexible line from the block to the sensor. Best to use a short length of -3 AN to the sensor mounted to the chassis by an Adel clamp.
The VDO diaphragm types are crap and you should use solid-state Kavlico, KA, RaceGrade or Honeywell sensors.
__________________ -Peter Krause www.peterkrause.net www.gofasternow.com
"Combining the Art and Science of Driving Fast!"
Specializing in Professional, Private Driver Performance Evaluation and Optimization
Consultation Available Remotely and at VIRginia International Raceway
I will add that NO data acquisition sensor manufacturer recommends hard-mounting to the block ANY pressure sensor.
ALWAYS a flexible line from the block to the sensor. Best to use a short length of -3 AN to the sensor mounted to the chassis by an Adel clamp.
Yep.
Back when I was instrumenting my first E36 M3 I discovered several reports of broken oil pressure sensors - all of them had been hard mounted to the oil filter assembly bolted to the block. The relatively heavy sensor is just too massive to be cantilevered out from the relatively small fitting. Vibration takes its toll.
I did what Peter suggests with a flexible line and it has been trouble free.
A broken oil pressure sender could cause serious engine damage and even a fire, so it's best to treat it with care.
Well, you can see on the graph I just posted that the oil pressure changes only in response to changing rpm. If you look closely you can also see very small dips at 4900rpm. The Variocam solenoids steal a little oil pressure when they activate.
My previously posted graph shows the oil pressure going up and down wildly in response to G forces. This is what is happening inside most M96 or M97 engines when they are being driven hard.
It is not uncommon for the oil pressure to drop below 1 bar due to G forces. I accidentally and expensively proved to myself that the oil pump generates about .75 bar worth of air pressure even without any oil in the engine. So, if the pressure is below 1 bar there is essentially no oil being delivered to the bearings. The engine fails once it accumulates enough run time without oiled bearings.
These engines are remarkably tolerant of this situation. I cannot explain that.
09-21-2016, 06:45 PM
