Oil Supply Improvements

Oil Supply Improvements

Just read Crumb's response to a question someone had about improving an S4's oiling for the track. Kim suggests the GTS baffle and also a pair of oil line check valves in each head from a 944 (944 107 189 00). I have heard about the baffle, drilled crank, and Accusump, but no one had mentioned these check valves before. I'm planning on a valve cover painting this winter. Should I add these check valves while I'm there?

John Veninger
1990 928GT

This topic came up at the Frenzy and was discussed at length by Kim Crumb and Marc Thomas (DEVEK). As I recall, the 944 check valves help to limit the amount of oil that goes to the cam towers during high RPM conditions. With less oil in the cam towers, there should be more oil that remains in the crankcase. Hopefully enough that the oil pump does not suck air when the car is going through a high lateral G-force turn for any substantial time period. From Kim Crumb and D Lloyd, that time period could be about 12 seconds. Also, if the crank is not redrilled, rod bearings 2 & 6 will likely fail after about 12 hours of heavy track use. YMMV.

~ Merry motoring ~

Oiling system failure after 12hours of track use?? Really? My old engine with 170k miles had 47 driving events and that added up to over 60 hours of actual track beating time. My new engine has over 8 events (11hours) and sustained high revs, but never over 6krpm . It is one of the most used 928s in the country from what I can tell, and definitely one of the fastest road raced, street driven as well. (Mark Anderson and Joe Fan's are not street driven now, and Devek's is the open road racer).

The point is, sure there are ways to insure the oling problem wont bite you, but if you dont have those resources to fix it, try short shifting the car under 6K rpms, and maybe in sustained turns, keep it lower than
that. You don't get that much gain for the next 1k rpms, and it sure may infuence the failure of those bearings. (I'm knocking on wood right now, before I travel to Willow Springs to run with Mark Anderson on that high sustained "G" track.)

Also, in pulling my old motor apart, there was no visable wear in the bearings or cylinder walls. After 47 events!!!

Ive seen a few 928s at the track waste their motors, one thing they all had in common, was shift points over 6500. I think even if the motors are balanced, the bearings or rods may fail if not designed for the higher rpm compared to factory.

See you at the track,

Mark Kibort

I happen to think you are correct. Although I don't have anywhere near the amount of track time on my 928, I too do just what you said. The tracks that I frequent all have some sort of sweeper. I try to get through them in a higher gear than most recommend. Instead of being at 5500 rpm in 2nd, I'll do it at 4500 RPM in 3rd. If I'm a bit slower on the exit, I usually make it up by not having to up shift as soon. Another "trick" I employ is to top off the crankcase slightly above the upper fill line. YMMV.

When I replaced the pan gasket on the 86.5, I installed a GTS baffle and also replaced all the rod bearings. They were all in "like new" condition. Although I spent over $100 for parts I didn't need, I still believe the learning experience was worth it.

~ Merry motoring ~

Our testing has shown that the 2/6 rod bearing failure is caused by TWO issues, and the main reason for the failure is not the one most commonly thought.

We have spent years understanding the issue, 3 engine builds testing "theory" and teardowns and lots of engineering and here is the summary. How good is the summary, well, they are good enough for Porsche AG to purchase our solution for their racing 928 customers!

First, the 2/6 oiling path is the least contributing factor to the failure! This is the common misconception.

The main factor is simply oil starvation, and the 2/6 rod bearing feed off the crank is the first to get the gulp of oil/air.

Where does the air come from, well if you are cornering at 1g, the the oil in the pan is at a 45 degree angle, and depending on oil level, will uncover the pickup....hence the air!

The secondary factor is from off-angle oiling of the 2/6 journal off the crank. This is a minor issue and if oil pressure is sufficient it is a non issue at any rpm.

The oil starvation, or lack of oil in the pan, is caused by the fact that oil is pumped into the head in large quantities by high rpm and the head acts as a oil "pan" of its own. The crank also spins up a batch of oil and keeps it in suspension, especially at high rpm. This oil cloud circles the crank and is the reason most race engine have "scrapers" in the case. Up to two quarts of oil can be circling the crank at high rpm! In some cases it can be more as the S4 is known to "pump" oil up the oil filler tower and keep it there!

When there is not enough oil in the pan (caused by the high rpm pumping of oil to other areas, like the head, etc.), then under cornering at 1 g, the oil pickup is uncovered!

So, the worst thing you can do to your engine is this...a 6500 rpm corner entry, a 1 or more g corner force while maintaining are sure to pick up some air...and do it enough times, the bearing will fail.

It will happen more frequently and sooner on a S4 then on an early engine with the same cornering capability due to the fact that there are less places for the oil to stay in the head and the paths in the head are smaller. So more oil stays in the pan.

The solution as originally develop by DEVEK is as follows...use our Accusump systems! The DEVEK Accusump systems have two levels, street and race and each one offers a different level of protection, but only for so long. Approximately 10 - 17 seconds depending on oil type, temp, etc. These are the systems that are purchased and recommended by the factory.

An additional item is to decrease the amount of oil available to the head at low oil pressures is by increasing the spring pressure or my favorite, decreasing the feed orifice.

The other method, again as developed by DEVEK (by the late Bob DEVore) and currently used by other racers, is a simple dry sump system that uses the original oil pump to feed the engine from a remote tank, rather than the oil pan and a scavenger pump to fill the tank...

And of course, you can just dry sump the system externally and depend on no stock oils systems.

On OUR Scat stroker cranks, we simply pick up oil off the number three main bearing to make is a straight shot to feed the 2/6 rod as some one correctly pointed out, it is routed like a chevy. This is unlike our drilled cranks, which have a modified oil pathway.

I hope this clears up most of the misconceptions about a weakness of our engines and clarifies why some folks believe that the engine will last longer if the rev's are lower.....less oil pumped out of the pan.

Oh, by the way, can it happen to a non racer? YUP, if your oil level is low, and you are running at 6000+ rpm for a hundred miles of can happen. And this has happened.

More information is available in our catalog or give me a call or email....

Warmest Regards,

Marc M. Thomas

There appears to be some confusion on how the cranks fail, why they fail and the fixes.

DEVEK researched these issues which resulted in a factory visit to review the data. The factory then recommended the purchase of our accusump kits for their racing customers. In some cases, they purchased the kits for the customers. So far, no customer who has installed a DEVEK racing kit has lost an engine. One German racer is still racing on the same engine as he did last year, rather than pay for the usual 3 rebuilds per season!

The accusump should not be plumbed into the oil cooler lines as there are times when the cooler is not fully open due to temp conditions.

In under 6500/7000 rpm constant running or high rpm cornering, the failure is due to oil starvation.

It takes over 7000+ rpm to have a differential gap in the bearing that can support a altered flow amount necessary to service the 2/6 bearing. Oil pressure operates on bearing clearance. In order for it to flow in one direction rather than in another, then one bearing gap must be larger than others!

The fixes are simple............use an accusump, a scraper plate in the case or GTS baffle in the pan,and/or keep the revs low to prevent oil starvation. If the engine is out, it will not hurt to have the redrilling performed on the crank as well as using the above items.

And if you want to run 7000+ rpms, redrill the crank and revise the bearings, plus all the above.

And if you got the bucks, use a dry sump and make some more power.


The flow of oil in the 928 - the oil from the pump is presented with three paths. The first path is a pressure relief that opens at 8 BAR - which is why 928s peg the gauge when cold. The excess oil is routed back to the intake side of the pump. This one is a dead end and no longer is part of the discussion. The second path available is to a bypass valve which limits the amount of oil pressure flowing through the oil cooler so that it never exceeds the design limit of the cooler (This was deleted as of 1991 models) It opens at a pressure differential of.5 to 1 bar. The third path of oil flow for all but 1980-84 USA cars is through the heat exchanger in the radiator or the external air cooler standard as of 1990. There is a thermostat element mounted in the block just above the oil pressure sending unit (Porsche chose this location next to the ports for the oil cooler lines as the best place to monitor oil pressure). When the oil is cold both oil cooler line ports are exposed to full system pressure and being in a state of balance little oil would flow to the cooler but the oil lines and cooler are always completely FULL of pressurized oil. As engine oil temperature increases the thermostat expansion element closes the main flow path and the oil flows out the lower oil line port to the cooler returning to the upper return port. At this point the flow paths converge. From the thermostat / relief housing all the oil flows through the oil filter then into the oil galleys which are made as part of the block /lower cradle assembly at the point where they are bolted together. So what is the point of this description of oil flow ?? No matter where you add an outside source of pressurized oil the oil has the option to flow toward the bearings or to flow back toward the oil pump. The only time the reserve oil can flow is when there is low oil pressure in the system which would mean the pump was trying to pump air. It would make little difference where the oiling "extra oil pressure" is added to the system you must pressurize all the oil from the pump to the bearings including the oil cooler lines, oil filter, heat
exchanger. Unless you have added a one way check valve that all the oil must flow through all the time. (The oil filter mount is a reasonable alternative) And introduce the pressure above the one way valve the oil has the option to flow both ways. Additionally most pressure accumulators have a capacity of 2 or 3 quarts of oil. The accumulator is initially charged with 5-10 lbs. of air pressure on one side of a diaphragm and when the engine is started, oil under pressure displaces the diaphragm (be aware that there are some "accumulators" being sold that are empty cylinders and as such do not "store" any pressure). The discharge of the pressurized oil is a result of the compressed air ( liquids do not compress ). If the accumulator stores 3 quarts @ 5 bar as soon as 1 1/2 quarts have flowed back into the system the oil pressure remaining in the cylinder should be about 2.5 bar. As the last bit of oil flows it is at the 5-10 lbs. of the original air charge to the accumulator. In reality the pressure in the accumulator is what ever the engine oil pressure happens to be every time you get off the gas, brake, clutch and shift. For the accumulator to add oil and pressure there must be a lack of pressure in the oiling system. Those who have lost engines that I have talked with have NOT REPORTED any loss of pressure (Don Hansen as a recent example)( David Lloyd) no red warning, no central warning, no fluctuating oil gauge.... If no loss of
oil pressure the accumulator would not add oil pressure...... It would be a very interesting real world test to remotely mount the air pressure gauge from the accumulator and monitor the pressure or video tape its movements --- if oil is flowing back to the engine the fluctuating air pressure gauge will clearly indicate this..... if not , then it is another problem that kills 2 and 6 rod bearings. Which is the position that we have maintained for many years based upon Mark Anderson's experience using a dry sump oiling system and still having bearing failures of # 2 and 6...until he had the cranks redrilled.

928 Int'l

The Devek supplied Accusump provides the check valve to prevent the accumulator oil from back feeding into the oil pump in a unique way. They use a 944 oil filter which has a built in check valve. All the accumulator oil flows to the engine whenever the oil pump is supplying less pressure than the Accusump pressure.

Louis Ott

Sorry folks but the factory oil pressure gauge is VERY SLOW to react! I have two gauges mounted on the development is the standard factory oil pressure gauge and the other is at the accusump's output.

During testing at Laguna running at high rpms (6-6500) there were slight nudges in the factory oil gauge on the corkscrew with larger fluctuations in the accusumps activity!

I have built and run one engine on the dyno at 8000 rpm about a dozen times and then taken it apart.....there was no damage to the bearings due to lack of oiling. I performed the same test with a drilled difference in bearing wear. Obviously, at 70 or so psi, the oil does flow around the corner on the crank! Remember, no curve or off angle is going to stop the 70-90 or so psi going to the bearings and bleeding out. Remember the oil pressure is a measure of the BACKPRESSURE in the system!

This is what led me to the accusump development effort some years ago.

Adding the accusump to the oil cooler lines is undesirable, adding it to the oil filter with a sandwich plate (DEVEK standard accusump config) is ok, but allows oil to flow backwards and forwards. Adding the DEVEK racing config only allows oil to flow one direction....into the bearings.

The only thing better is the dry sump system.....the one developed by Bob DeVore, over 10 years ago, (the DEV in DEVEK) is my favorite and is used on Mark A racer. He developed the absolute best combo for using the stock pump as the pressure pump and an additional sump pump as the scavenger....expensive, but worth every penny!


928 Tips Home     Greg's Home