Sale ends todayGet 30% off any course (excluding packages)

Ends in --- --- ---

Dry sump massive oilpressure drop

Practical Engine Building

Forum Posts



Tech Articles

Discussion and questions related to the course Practical Engine Building

= Resolved threads


Hy everyone ,

I have here some Logs from a fully build Subaru EJ25 with a drysump system.

This car has some strange issues with oilpressure bouncing all over the place , and pushing the oiltank empty over the vent.

I first thought it is only a oiltemp related problem but on the track it´s at low and high oiltemps and then at simular oiltemps not?

we changed oilpress valve , complete tank to bigger one added baffle plates in the tank , air oil separator , changed the crankcase ventilation ,rebuild the engine checked all clearances nothing works. The droping oilpressure is coming back every time .

We running out of ideas what is the cause of this?

Regards Marco

Attached Files

What are the sizes of your oil lines? What is the length of line from the oil tank supply to the inlet of the pressure stage your oil pump? Is this hose vacuum rated, or does it have an anti-collapse spring installed internally? How is your scavenge setup - how many stages, hose/port sizes, do you have a crank scraper installed? Is there windage tray to keep oil below the rotating crank?)

Is there any possibility your belt is slipping? Is belt a power-transfer design (ie, had teeth in the pulley and belt so it can't slip).

If you install a manual oil pressure gauge in the same location and your pressure sensor, does it fluctuate just free revving the engine in the shop? If the manual gauge is steady, but the sensor does fluctuate, then check your wiring for an intermittent connection (broken ground, or signal shorted to power).

Are you running enough oil? How /when do you check the level? (Many dry-sumps require oil to be hot 180f+, and the engine run at an increased RPM (like 3000 or 4000) for say 15-30 seconds.

Hi Marco,

Where is the oil pressure sensor mounted? What sensor is it that you are using? Is the logging from the same device that the Oil Pressure sensor in read by?

Do I read it correctly that it is pressurising the oil tank and pushing the oil out of the tank through the breather? If so, it sounds like there is an error in the plumbing. With all of the dry sump systems that I have dealt with, the tank has sat around Ambient pressure internally.

Is this a Ej255 or Ej257 with Dual or Quad VVT? Can the CAM Aims be overlaid in the logging with the Oil Pressure channel?

First off, from your picture, I wouldn't drive the car until this is resolved, and I'd implement oil pressure safety in that ECU.

RCM, Dailey, Cosworth...which kit?

Who designed your dry sump implementation, meaning the lines, tank, breathers, valve covers, not the pump/pan portion?

Is this a new install or was it working fine then suddenly having issues?

As other mentioned, line size and routing are common issues.

Incorrect hose type (collapsing), or inadequate breathing for the main tank are other common issues.

Are you logging CC pressure?

GTG, but before that, I would suggest you post as much info as you can - ACTUAL useful info' on the installation, the pump make and model, the sump used, the plumbing, the pump dive, hoses used and even fittings, the engine and tank breathers, etc.

My 5c worth, maybe 10c?

OK, there seem to be two main issues, fluctuating oil pressure, and the venting of oil out the dry sump reservoir.

Looking at the latter first, because it may also be part of the first.

Depending on the specific plumping, there are normally going to be two vents to the sustem.

a/ is at the engine, primarily to allow the air into the engine that's required to carry the oil and droplets out of the crankcase (and head(s) if also scavenged, which I would expect on that engine), through the scavenge pumps and back to the the reservoir (oil tank).

b/ a vent at the tank to allow the air carried with the oil to vent to atmoshphere after separation.

If, as reported, there is oil being passed out of the tank vent it suggests at least one of four different potential issues - there may be more that don't come to mind at the mo'.

a/ the scavenge pump has too big a swept capacity and/or is being driven too fast, with the result that excess air is being drawn into the engine, passed back into the tank, and the breather there doesn't have the vent capacity - especially when c/ and d/, below, are considered.

b/ the tank breather lacks the capacity to vent the air being passed back from the engine

c/ the tank oil level is too high - as a general guide, IIRC, it should have the oil leval aroung mid-point in operation, but the tank manufacturer should be better able to advise

d/ the oil-air separator in the tank is unable to separate the air from the oil sufficiently - this may be due to a large amount of air mixed with the the oil/oil droplets, etc, and/or poor design and/or the incorrect oil level in the tank for correct operation.

Now, with the fluctuating oil pressure, when the above is considered, if the air isn't properly separated, there may be some drawn into the tank's pressure feed line and these 'bubbles' may be causing fluctuations. If the oil level is too low, which is possible but seems to be unlikely from the above, there may be some drawn in as the "whirlpool" affect.

Two other things that may be occuring is the pressure regulator being a little sticky, or as has been suggested by others, belt slippage - ALL oil pumps SHOULD be driven by Gilmer, or other brand, toothed belts - they can't slip and can be run slightly "loose".

Overall, from the descriptions, such as they are, I suspect there's too much scavenge for the tank and tank breather to manage.

If so, there are a couple of simple suggestions that may help.

Using a self-regulating engine breather (or breathers if required) set to hold some crank-case vacuum - 5-7 PSI or 12-15"Hg should be a good compromise and reduce the net mass of air in the oil that needs to be vented. And/or running a larger pump pulley to slow it down - keep an eye on the oil pressure to make sure it still has the pressure pump running fast enough to hold the set value AND monitor the tank to make sure the oil level stays stable in operation.

That said, you should also double check the tank has an oil separator, and that the oil level is correct.

Oh, some had also made an excellent point that you NEED to use oil lines on the tank>pump side that are larger diameter AND "vacuum rated"to prevent collapse, and from the engine to the scavenge pumps, for the same reason.

If you have a tight bend, even one that seems fine, they can partially collapse under a vacuum, causing problems.

I assume you have some fittings in the line(s), they may need to be looked at, too, because some can be rather restricted, especially if originally intended for heavy hydraulics

To cover all of your questions , and make it as short as possible here are the specs and the changes we made till now .

The Car is a Impreza GC8 from a friend of mine .

Ej25 IAG Billet Longblock ,CNC portet Intake AVCS Heads, Borgwarner EFR8374, ARE Drysump Kit for EJ (3Stage pump, Oilsump, Oilfilter relocation adapter

and Oiltank 7007-A with external Vent Can.) Runinng on E85.

that he build by his own and for the engine, ecu , wiring and maping by a tuner here from switzerland and got it back

about approx 5 years ago the car was working on the dyno.

The first testdrive on the track endet up with low oilpressure warnig and rpm limter cause of engine protrecton strategie of the ecu. At beginnig it only

appere on the higer boost setting, but later also on the low boost. It was not possible to get the log data from the ecu because of passwort protection

through the tuner. because of disputes between the tuner and the owner we did not get access to the passwort. Also the crankcase ventilation at this time was

a catchcan desing the one valvcover breather from left and right and rear block vent all in one can. The second valvcover breather blocked off. I am not sure if

it was vented to atmosphere or closed? But it overfilled with oil after feew corners.

So we check the hardware, Ventilation of the tank and the crankcase vent,oilfilter, checked the oilpressure with a analog gauge. The reusult was that the reading

of the oilpres.sensor was correct so there was a oilpres. drop but more a over all decrease of oilpres. with rising oiltemp and not a up and down like now.

Increasing the oilpres. was not possible the adjustscrew was already at max. A sugestion was that the oiltank was to small for the foamy oil returnig to the tank

and could not calm down? (it looked like cappuchino in the tank). We ordered a bigger tank 2.5 gal and plumped the crankcase breahter without catchcan togehter

and routed it with a AN12 hose to the oiltank PVC inlet. Did not realy help.

A look in the engine and a leakdown test showed lots of blowby in the crankcase and some scratches on the cylinderwalls.So we took the engine out and inspected it

for damage. There was only a little wear in the engine my friend decided to send the engine to a well-known german engineshop to get it rebuild.

As the oilsump was removed i saw there is no windage try installed and the oil pressure in the engine was through the (sump) in the location where the pikuptube is

in the engine, sealed with a oring between block and sumpplate.I made a windage tray and modyfied a dash fitting that reach in the hole of the pickup, so there s no

chance the oil could go between.Also i changed the oem oilpump and welded all holes that the oil has a shorther way to the gallery.( The fitted pump had still the

inner rotor inside, in my opinion this could cause also problems with the oilflow).

New engine back in and the a new pressure valve of the oilpump , and a air/oil separator. The only way to get access to the ecu was to erase all and load a new

firmware package to start new.The was running again but needed a mapping.Send the car to GP Performance in Germany, on the dyno it only pushed a half liter oil out

the oiltank ventcan no other problems.

We added a crankcase catchcan with a mesh to prevent oil getting to the return line in the oiltank.The Oiltank had also very poor baffle inside i added a tube and

a aditional grid to prevent it slosh around.

At the track the car did the strange oilpressure issuses from obove and the oil comming out of the oiltank vent . in the garage reving the engine it also had

oilpresure drops at 92C oiltemp.The 2,5 gal. oiltank is only filed with 6 liter even then its overflows.

The Hoses all are very stiff even in hot condition they do not colapse, i am not shure if the length of the hoses is to mutch for the pump.

if so we mount the oiltank to the passanger side.

Attached Files

First, I don't think your oil pressure sensor is in a good location. It should be measuring essentially a gallery pressure in the engine. So plumb that so it will be downstream of any oil filter or oil cooler. If you can tap off a gallery plug on the engine is is often ideal -- use a AN hose to connect the sensor and isolate if from direct engine vibration for best life of the sensor.

I would suggest that your oil filter would work best if you located it between the Spintec and the dry sump tank. Why not keep the tank as cool as possible? and have cooler oil for the pump pressure stage to operate with.

I think the long scavenge lines and large lift required by mounting the pump on top of the engine are contributing to your aeration problems. I bet that the sump has to accumulate a large amount of oil until it can fill the lines and make it's way from the scavenge pumps. Then it's sucking air (like a straw at the bottom of a milkshake), until enough oil exists to "prime" the pump again. So I think your oil is scavenged in "gulps". You might get this to work with properly sized lines (consider hard lines with a clear section to observe what is happening for testing in the shop).

Finally, with the 2.5 gallon oil tank in the trunk, And -12 lines running the length of the car, I can't see how 6 liters is enough oil. it might be enough to start, but after all the oil lines & coolers are full, you will likely need to add more.

Since you can run the engine at high RPMs in the shop -- what does the oil entering the tank look like, is it solid oil coming from the Spintric, or it that still aerated -- does it "flow" against the tank wall, or is it "splashing" into the oil tank or screen?

Just my opinion.

Looks like the chap(s) put a lot of work into the car. I'm not familiar with the kit, but from what I was able to confirm...

I'm missing where the oil pressure is being monitored - normally it would be at the rear of the engine, after all the other oil drillings, so it's the minimum being measured after the drops through the engine. I suggested checking the relief valve because if it's sticking a little it will be 'shuttling' between the pressure and the spring, as stcktion prevents a smooth transition and state - might be a bit of a long call, but it's not un-common and can then be eliminated as an issue.

If you use the installation guide here - https://spintric.com/v3/tanks/ - you can double check the pipe and hose routing, just in case something's been incorrectly plumbed, it's at the bottom. Also, note the comment about the breather hose - it's recommended so any condensation in the hose will drain back to the tank, rather than to the breather.

I suspect a big part of the oil from the breather is the tank is simply over-filled, you can check the correct process here - https://drysump.com/index.php/checking-oil-levels. There can be more oil in the system than one might realise, as it's possible for oil to be 'syphoned' from the tank back into the engine, especially if there's a prime height.

If the oil was SEVERELY airated, the bubbles could displace the oil and possibly be a problem, but the correct oil level, as per instructions above, and Spintric separator should minimise that.

As David suggested, having the oil cooler on the return from the scavenge pump would have some benefits - there won't be that pressure drop to counter on the engine feel line, the oil will be hotter and so the cooler will be more effective, you reduce the heat being radiated from the tank which will heat up where it is and if you have a bad prang, the cooler oil will be a bit harder to ignite and, if you get splashed, you won't be burned as badly.

About the only draw-back that comes to mind is the cooler oil will be a bit more viscous.

What happens to the oil level if held at a low rpm, medium rpm, and high rpm with the vehicle parked?

Forgot, the kit details state there is a "windage tray", are you sure it is the full kit?

Gord -- look at the diagram of dry sump picture. It shows the oil pressure sensor location near the oil filter on the left side.

I agree with you the oil pressure sensor should be on the engine block somewhere in between the main gallery.

The oil level of the tank is lower if the engine is off and rise about 0.5l-075l on startup. If the rpm is rised it slowly decrease the oil level and the return line spits the air Oli back in the tank it's hard to tel if the spintric separates the two . We disconnected the two return hoses air and Oli in to a bucket. At lower rpm the Oli is coming out on both lines and for more it got to messy holding them by hand. The spintric should work better on higher flow? Since I made the changes in the tank added a tube in the centre of the baffle discs and the grid and the return tube points now to the wall not straight down in the oil. The oil is more solid in the tank with a few bubbles on the top . If we fill the oil up and check the level like in the description it end up with spiting the oil out to the tank vent as t higher rpms with out driving. Approx 2 to 3 litres are gone .then we had approx the 6liters left +- the rest in the lines and cooler.

We gona change the pressure sensor location add some sensor s for CC pressure and tank pressure. And connect the old tank in front of the car to check what happens.We tried already but had no spare fitting for the return line. We mounted the top of the tank also because of the spiting return.but could not rev over 4k at 98C with the lose return line.

The windage they mean is to low . The relief valve is already new. But I wil check it again. What if the pump is spinning to fast? It looks like the oil pressure drop starts at higher rpms and get worse and worse. Could it be that through the over spinig the oil tends to cavitate?

Attached Files

I see it, thanks, David - was looking at the engine.

Marco, I, you, or both, are mis-understanding something.

Did you go back over ALL the lines, and how/where they're hooked up? It's easy to accidentally swap over similar lines, which is one reason I tend to use different diameters and/or fitting options (if two similar lines, have one a male end, the other female, etc).

There is a finite volume of oil in the system, and for the separator(s) in the tank to work properly they need to have the correct oil level to shed the air in it. If the return oil is highly air-rated, that air will also add volume to the oil level in the tank unless the air is removed/minimised first.

You asked about the pump speed, I mentioned this earlier as a possible issue - if the pump scavenge wafers are drawing excess air with the oil, that will tend to increase the load on the Spintric, and tank baffles, and if they don't do the job the level in the tank will rise.

I think you are concerned about a vortex (whirlpool) forming and allowing air to be drawn into the pressure suction line? This can be a problem with poorly designed tanks - on one hand diresting the tank return to the wall cabn significantly aid de-airing of the oil, but it can also help promote the vortex. There should be baffling towards the bottom of the tank to prevent this happening.

I'm not sure what the internals were, as that tank doesn't seem to match the tank data I have for that model (but that's not saying much), nor some of the details of what you've done.

It's important the tank level is checked immediately, because that best represents the running oil level, with the lines still full and no drain-back. The oil level may drop as the engine is run, and that's normal, or it may rise a little due to air content in the oil that hasn't been fully removed yet.

Oh, while i think of it, the tanks are in US gallons, so 2½ US gallons is approximately 9.5 litres, compared to the imperial which would be roughly 11.4 litres - could that, almost, 2 litre difference be part of your confusion?

The oil doesn't come from no-where - either the oil level isn't correctly set, there is a plumbing issue, the serarators aren't working, something else, or there're a combination of reasons.

At this point, I think it might be enlightening to install a visual check, such as a clear sight tube, to see exactly what the level is doing as rpm changes.

Might be a better representaion of the tank fittings - https://drysump.com/images/PDF/7007-A_TankDiagram.pdf

Oh, forgot - some fittings are MUCH more restrictive than others. Worst may be high pressure hydraulic, best are the low restriction/large bore motorsport ones - don't know if it'll be significant in your application, but may be something to bear in mind.

Thinking on this some more, and as it IS an already well established kit for the engine, the only SINGLE thing I can put the issues you have down to is a separator that isn't allowing the oil to pass through it, as it should be doing so. It 'should' be fine, but if it was marginal and/or there's some contamination blocking the mesh a little, it may be allowing the oil to build up above it. That could cause starvation at the oil feed to the pressure pump and the overflow that you've observed?

About the only thing I can be sure of is that it's going to be a "doh" moment when you do figure out what the problem is.

Did you ever get this sorted out?

We usually reply within 12hrs (often sooner)

Need Help?

Need help choosing a course?

Experiencing website difficulties?

Or need to contact us for any other reason?