S. Madman

Vacuum pump check valve diaphragm is almost completely gone. Turbo wastegates not holding pressure. Ordered one from China for $19.99 since the dealer wants the whole pump at $600+.

CaliBenzDriver

Nice to see ! What lead you to suspect these valve where were done?

I inspected mine a month ago at 50kMi: silicone looked perfectly new. I found it swamped in engine oil on both sides.
Same thing with vaccum booster check valve.

I wonder how much engine oil is normal if any, inside these vacuum pump???

Could it be crankcase pressure blowing through the vacuum pump passed the tiny silicone check valve ??

S. Madman

I think there should always be a little, as expansion caused by cold/heat will change the diameter of the diaphragm. It should seal up when there is vaccum/pressure.

These check valves should be changed every 5-7 years, depending on oil quality, engine mileage, and heat/cold cycles.

DFWdude

What scanner readout did you get that pointed you to this culprit?

CaliBenzDriver

that small vacuum line also feeds the plenum diaphragms. They are 3 of them moved around by ECU solenoids used to switch vacuum source.

The larger outlet goes to brake booster plus other accessories.

Both small/large check-valves looked like "silicone grade" not short lived EPDM soaked in petroleum.

Its good to have these vacuum membranes on regular inspection list. ✌️

This vacuum pump is a little misterious the way its soaked in oil like the plenum... where is the air outlet for it ? (not inside pressurized crankcase)

S. Madman

I was tuning the ECU when the logs showed low boost.

Seat of the pants pull was not getting better even with ambient colder temps.

S-Prihadi

S.M ,

Next time you can add a stand alone inline check valve. Yes, the hose barb of the new inline check valve need to match the MB hose ID in use.
The MB check valve, you must remove the silicone diagram, let it full bore open.
I would do this when and if my check valve goes bad.

Something like this :
https://shop.stoneyracing.com/univer...ter-catch-can/
https://www.vpw.com.au/proflow-inlin...-one-way-check
https://rceperformance.com.au/replac...-pump-kit.html

S. Madman

I thought about it, and I will swap it as soon as I find all the hardware.

S. Madman

As expected, power is back. If I get enough interest from it I will do a video. It seems easy enough, but some things should be avoided to prevent some parts being broken.

S-Prihadi

Guys,

Now that we are discussing VACUUM PUMP, let me brainstorm some :
The big diagram S.M replaced is for brake booster.
The turbo uses the smaller one.
I suggest to also replace the smaller one : https://www.ebay.com/itm/115045707705




I been messing with my HVAC vacuum pump, lost one RobinAir 6CFM unit to neglect and now I got a new good one a 8CFM unit.
Now I know what those diagram is actually for.
In a turbocharged engine, its intake pressure is above atmosphere due to forced induction by the turbo.
11 & 13 PSI, typical for M276 3.0 TT and M278 respectively.
Therefore piston blow-by may cause crankcase pressure to be higher than NA engine at time of turbo boost occurring.


The "sucked" air is dumped to engine crankcase.
Below is the animation




Vacuum pump does not work like a positive pressure air compressor.
Go to 3:11

and read here : https://www.processtechacademy.com/w...mp-not-a-pump/

So the diagrams used on the bigger and smaller outlets is to not allow positive or higher pressure from engine to ruin vacuum ( negative pressure ) level at the brake booster and
the turbo's wastegate vacuum actuators.

Now, if the reed valve is already un-healthy and its surface is scored, good vacuum won't happen and it can also introduce positive pressure from engine's crankcase.
The seal/vanes is the one doing pressure changes to get vacuum going, if the liner/wall or the seals/vane themselves are scored, deep vacuum won't happen.




Now, it is unfortunate for car engine's vacuum pump to be using dirty engine oil as its mechanical lubrication and vane's sealing.
In HVAC vacuum pump, a slight dirty oil and water moisture in oil will cause going deep vacuum to under 100 microns read immediately at input... NOT possible.
Also HVAC vacuum pump has 50% diameter of its pumping head fully immersed in oil, so better working conditions for higher vacuum reading.

Maximum vacuum at sea level is : -14.7PSI or 29.92 inches of mercury vacuum is equal to 760,000 microns.
In car vacuum pump we do not deal in microns, -14.50 PSI is already a very good value.
I am just explaining how common HVAC industry technician vacuum pump works.


I think you guys best test your vacuum system regularly simply by using vacuum gauge sensing at that test port of the brake booster.
01. Test what is maximum achieved negative pressure with engine idling , in 30 seconds, do not press brake pedal.

02. With engine OFF, how fast or slow does the negative pressure ( vacuum ) reading creeps up to atmospheric pressure.
If it looses negative pressure fast, that means the big and/or small diagram is not closing well.

03. With engine OFF , how many times brake pedal maximum push will deplete the negative pressure ( vacuum ) ?

Log these 3 values when your vacuum pump is still healthy and you have a baseline.


==================

Here is my engine's Log after 5 hours shut down in my Indy Worskhop and I head home, 34.8 minutes engine run time.
My Banks Gauge logger will start logging approx 5 seconds after engine started ( it will sense RPM as a wake up call and auto logging )






.
We can see brake booster "vacuum" consumption.






.
Highest vacuum achieved, -14.57 PSI




.

Whle under boost, we should not loose vacuum, if all is healthy.

S. Madman

Yeah I already checked them on this thread. https://mbworld.org/forums/e-class-w...76-3-0-tt.html

Thanks for the reminder though, hopefully others are doing the same.

CaliBenzDriver

Thank you Surya, now the inner working of our engine vacuum pump makes a lot more sense. It must have an oil inlet and outlet to keep moving parts sealed.

On turbo there's unstable intake vacuum due to boost pressure.

On NA engines we can expect some level of vacuum vs. blown-by crankcase pressure mostly relieved through PCV conduit into plenum.


> Brake sensor:
I think I've seen our W212 brake booster does feature a vacuum sensor. Bosch ESP does not expect reliable vaccum and wants to modulate brakes accordingly.

We can read sensor live to assess what the pump is up to dynamically. Pump has two separate check valves in each outlet ie. one can be bad and not the other ✌️


> Reverse pressure:
The enemy of our vacuum lines is positive pressure. This brings an interesting topic about OP silicone diaphragms:
-- Why are valve parts disintegrated?
-- How is engine oil forced beyond good valves in one way circuit?


The discharge reed valve is what normally keeps crankcase pressure from entering the pump through the discharge path, right?
The reed is the true hardcore check valve, like inside two-strokes engines.

Diaphragm usually has little limited pressure differential with similar vacuum on eother sides. When diaphragm sees boost pressure + vacuum... the differential is too great: it's history blows through.

When reed leaks, the pump vacuum is quickly spoiled, pump spreads crankcase pressure blowing through the litle diaphragms: (thread OP).

In that sense the tiny vacuum line outlet may act as the weakest link to relief pressure and preserve the brake vacuum path.

When reed begins to fail, the brake booster sensor should register a poor vacuum number, right?

SMad: do you want to lookup some vaccum numbers around stationary idle?


> Maintenance ?
I came here thinking silicone part was it but now we now it's just a weak sacrificial scapegoat.

-- Is it worth replacing only diaphragms + reed + seal kit and not the moving parts sweeping around aluminum chamber ??

-- Near what mileage pump goes (TT/NA) ?

> Root cause:
I guess the real fix here is to deal with the blow-by before it jams the flexible reed blade.

++++ Maybe we should summarize counter measures againt abnormally high crankcase pressure.... Rings stuck by carbon from burned oil over not-sprayed cooled pistons.
​​​​​​​

S-Prihadi

This is so good. Every time we/buddies have issues, we go into R & D mode and learn some more.

Here is how the oiling system in our vacuum pump works : https://www.ms-motorservice.com/int/...oil-supply-561
Let's use middle value of 45 liters per hour or 450cc per minute or 7.5cc per second.
So our engine vacuum pump is a constant circulating oil, while common portable HVAC vacuum pump is like our Rear Differential oiling system, simple oil bath 1/2 level full.



I wonder, what kind of oil pressure is the oil injected into our vacuum pump worth ?


.
Now, while Silicone is one of the greatest flexible diaphragm material for mechanical longevity for repeated flexing, it is not truly engine oil resistant actually.
https://jehbco.com.au/products/chemi...ibility-chart/






.




Now I realized Master Tasos question answer is here: How does oil get into his customers turbo wastegate vacuum actuators ( and a ruined actuators they become ) ? He often asked.

It will take a few issues for oil to go as far as turbo wastegate vacuum actuators.
01. The small silicone diaphragm has to be first ruined and hence open up permanently or at least bad sealing.
02. The metal reed valve has to be in an un-healthy state of being.

This could be another cause of vacuum capability degradation when and if too much worn out :
The dynamic seal of the rotating part of the vacuum pump if leaking, it could divert back oil to engine instead of being used by vacuum vanes.
Or this is simple bushing bearing like some turbocharger and no polymer seal ? Good machining tolerance can produce good enough sealing with acceptable drip like leak.




.
M276 vacuum pump : https://www.ebay.com/itm/266263084747
M278 vacuum pump : https://www.ebay.com/itm/385886841403




So the oiling method of M278 & M276 vacuum pump would be no 3




============

OK, I hope my vacuum pump will last 20 years from my super clean engine oil.
Reason being, I now since 2022 do a 6 months oil change, while mileage still stay the same 5,000KM or 3,105 miles. No more maximum 1 year.
This is because my car now does not run often, say once per 2 days like it used to when my kid went to school and I get my driver to rotate use among 3 cars for twice per day send-to and pick-up-from school.


If the Chinese/3rd party diaphragm is silicone is that cheap, might as well keep buying that once per 3-5 years. No need to modify anything else like the inline vacuum valve I suggested.
However, do note that it is the only visible worn out parts, the others I mentioned need vacuum negative pressure measurement to verify their state of health.

CaliBenzDriver

Now we know more than Master Tasos, at least on a couple of our pet topics !!
I remember seeing his video where he showed wastegates all filled up and tolerating oil.

1-- We now why the wastegate fills with oil
2-- We also know about pump solenoid



> Back-pressure protection :
Do you think this can prevent any pressure and oil from filling up westgate...

-1- Add check-valve to let any positive pressure blow out of vaccum line

-2- Add drip-leg pipe to catch liquid oil just like in building gas plumbing

-- Insert this to existing rubber vacuum line with a plastic "T" hose adapter.

-- The drip leg must be facing down.

-- Drip leg is shut closed at the end by a one way check-valve that blows out pressure but seals in vacuum.


When the silicone diaphragm let any pressure in the vacuum line it blows out through the added check valve with any oil stored in catch pipe - Good deal besides, Me no turbo

I think that can help prevent blowing your westgate vacuum diaphragm, right?


Oil inlet located near 8 O'clock position...

Based on the outlet hose position we can guess the metal reed facing 6 O'clock position is used to expel oil circulation back into crankcase with any air flow vacuumed out.

When pressure prevents that from happening, the incoming oil has no place to go but fill up the outlet vaccum lines - Brake booster outlet is all the way high up near the top, less exposed to a filled up condition.


+++ incoming oil has pressure! +++


looks like inlet has a mesh screen

The oil entering the vacuum pump chamber is under pressure... there must be some kind of system to let in the lube oil without loosing too much vaccum to oil port.
Vacuum inside and pressure on the other side
A tiny hole like in HVAC "orifice tube" to limit the flow rate?

S-Prihadi

Here is my rough estimate of the oil flow vs suction airflow.
Since oil is not air aka non compressible, even it is in positive pressure during its delivery from engine oil pump to the vacuum pump, it should not do any vacuum negative pressure harm.

Let's use middle size vacuum pump, at 210cc per revolution.





We then also use middle oil flow size, 45 liters per hour.


So its very little oil actually, only 2.5 cc compared to the vacuum pump capability to evacuate 210cc of air per 1 revolution of the pump.


That 210 CC pump is 65 liters per minute or 60/28.32 = 2.12 CFM vacuum pump
Which is a small pump. Like this one, battery powered one : https://navacglobal.com/product/smar...-pumps-np2dlm/
or AC voltage powered one : https://www.asada.co.jp/english/prod...ps/040401-008/
Our car uses single stage vacuum pump, those I linked are double stage vacuum pumps because they want <20 microns capability.

S-Prihadi

Now....if the turbo wastegate vacuum actuator can be oil filled/clogged from leaky vacuum pump ............ won't the brake booster have the same faith ?


YEP..........




.

CaliBenzDriver

Even FCP knows about .

Sounds like this is caused by crankcase pressure forcing oil through the pump towards outlets. Something to keep in mind in case of hard brake pedal specially when plagued with popping dipsticks.

​​​​​​

S-Prihadi

Well.... we learn more each day. As long as we want to think about X, Y, and Z, we may see A, B and C

S. Madman

Wanted to add this short video from Tasos, about the vacuum pump. Talks about it around 3:38.

Then on this he talks about the check valve.

S-Prihadi

Wow, S.M it seems we were connected last nite by telephaty
I spent 4+ hours watching the same Tasos videos you linked and many more videos to figure out our vacuum pump.
M276 and M278 uses exact same vacuum pump.

Interesting facts I notice , because Tasos video has shown the internals of our vacuum pump.

01. Our vacuum pump is a two unit in 1. I never realized it till I saw the false or 2nd liner/wall for the big check valve.
Let me explain :
The two check valves in our vacuum pump works independently one from the other, in respect to 1 full rotation of the vane, that brown polymer material non-metal.



In red are the holes. Hole 1 serve small check valve #2 ( turbo wastegate actuator ) and hole 2 serve check valve #1 the big one for brake booster.
In pink is the 2nd wall/liner for hole 2 or brake booster.
The animation from Pierburg does not show hole 2 per se, because it is a simplified version of our vacuum pump. It is also a twin hole or two check valves vacuum pump.

The pink 2nd wall/liner for hole 2 or brake booster is on purpose relocated as such, so that oil does not clog it.
Pierburg vacuum pump engineer surely would want oil be at the bottom and air suction at the very top, hence that special 2nd wall/liner for check valve #1 brake booster to be at the very top.
CV = Check Valve





The animation below is showing CV #1, which is the big check valve for brake booster has its diagram open up to allow suction hence creating vacuum.
Bear with me, this animation has opposite rotation to our vacuum pump, but principle of operation remain the same.

Anytime a small space becomes bigger space while no new air is added, pressure drop and vacuum occurs.
Example, 1 liter or 1 quart of air space in a vessel, tightly seal after being introduced to 1 atmospehere, we get 1 atmospheric pressure in this air vessel.
Expand the volume of the air vessel to 2 liters or 2 quarts while no new air is allowed in, the pressure inside will become 1 divided by 2 = 0.5 ATM, thus lower pressure and this is vacuum happening.



Again I am reminding, this animation has opposite rotation to our vacuum pump, but principle of operation remain the same.


As the vane approaches the small check-valve #2 (turbo wastegate one ), the above pink region air pocket is of higher pressure and the small check valve #2 has to stay close using its diagram, it has to allow the reed valve at the back of the vacuum pump to dump out this air into the engine.

There are two reed valves actually at the rear of the vacuum pump, one serves Check-Valve #1 ( brake booster ) and one serves Check-Valve #2 (turbo wastergate ).


.



Damn, I am in Bali at the moment and is using single screen 15" laptop, very difficult for me to work with images compared to my dual screen 27" PC at home.

I am still trying to figure out, at what degrees of vane stroke does the reed valve for the big CV1 ( brake booster ) will vent out.
This is a single vane but twin seals vacuum pump and twin action/hole, so me is still rather confused ....LOL.
One seal end can do X and the other seal end can do Y at the same time.


This is the total 360 degrees stroke of the dual seals vane.
3 different events occur at stroke A, B and C.
The zero degree I use the mid center between the 2 reed valves vent holes.




Rotation of vane is Counter Clock wise for our real vacuum pump.
I believe A and B is combo event.
I hope my explanation make sense, here goes.
Below video animation is Clockwise Rotation, remember that.
Take note that stroke B in blue is when CV2 ( turbo wastegate ) check valve will open, also indicated by animation...small opening.





So, during stroke A in red , the vane rear section creates vacuum or negative pressure. Its forward section being positive pressure, pushes out air out the reed valve, as explained earlier



By the time the vane end seal passes the small hole of check-valve #2 ( turbo wastegate ) which is stroke B in blue, each degree the vane goes clockwise, the space volume becomes bigger and pressure will drop,
the check-valve #2 stay open during this stroke.





Stay open, below :





CLOSED, the check-valve #2 diaghpram. Below.




As I said, I am still figuring out at what degree of vane stroke will the reed valve for check-valve #1 brake booster will open ?
Animation does not show that reed valve opening, but I think it is at the cyan color zone I inidcated below, where the Check-Valve #1 itself get CLOSE SHUT, that is when
the reed valve would probably open, as reed valve works opposite of its respective check valve.







==================

Here is another fun fact.
The spinning shaft which holds the vane, uses no seal whatsoever. It is precision machined and expected to use the engine oil as "sealing" mechanism.





This video shows a single hole or single check-valve of BMW-MINI engine vacuum pump, which its vane is metal and not polymer, when the vacuum pump worned out,
it can freeze and brake the camshaft end and timing chain chaos will happen !!!!

The same chaos is expected by this techy doing Chevy , albeit the vane is polymer/composite. Single hose single check valve. Oiling method is different to our vacuum pump.



This is BMW N63 composite blade ruined, but lucky the engine I think survived. Twin hole, twin check valve.



Will continue..............

S-Prihadi

So, the check valve #2 for turbo wastegate get very low stroke volume of the vacuum pump, since its air space is small, only the half apple size x 2 of twin turbos wastegate actuators.
The check valve #1 brake booster has like 10 times more volumetric stroke, since brake booster air space is so large and it is a big vacuum pressure consumer each time brake pedal get depressed.

I am still wondering :

01. Our engine crankcase is called a positive pressure ventilated unit. I wonder at its peak and zero leak from turbo boost into engine crankcase, assumed all healthy with proper piston blow-by value yada yada.... what is the atmospheric pressure inside a crankcase ? It is a tiny bit negative for sure, because our intake manifold is connected to it, but is it as low as actual manifold vacuum level of
a normally aspirated engine ? I never measure it yet. But I know older cars dump this gas/pressure to atmospheric air, that means it is above atmospheric air ...when not connected to intake manifold.

02. In order for reed valve of vacuum pump able to dump air into engine crankcase, it has to achieve higher pressure than crankcase internal pressure.... what is the positive pressure the vacuum pump
reed valves are pumping out ? In my HVAC vacuum pump, the positive pressure output is very very small, perhaps under 0.5 psi above atmosphere.


What negative effect can a vacuum pump leaky co-workers do to your engine ?
The co-workers are : Brake Booster and turbo wastegate actuator or non turbo cars using vacuum to operate its intake flap like M272. Does M276 NA uses such vacuum flap for its intake manifold ?

Brake booster play with atmospheric pressure and vacuum negative pressure using the rubber membrane to assist brake pressure boosting to master cylinders.
If the valve or o-ring which controls the shut-open of atmospheric air is leaky, the vacuum pump sucks fresh atmospheric air and dump it into engine crankcase and
if engine is using MAF sensor, it will run lean from too much un-metered fresh air. If my engine, M276 3.0 turbo, it uses MAP sensor instead of MAF, by right it will be able to
experiece much less disturbance as the MAP sensor reads that phantom incoming fresh air from brake booster leak as overall pressure in manifold.

The same effect can be said for torn vacuum hose to turbos vacuum actuators or if a leaky vacuum management solenoid M276 3.0 TT and M278 uses or M272 vacuum actuator for
intake manifold.


MEASUREMENT OF VACUUM FOR VACUUM PUMP
Now that I know better how our vacuum pump works, I mean 2 separate "channels" it has, I now then when measuring vacuum..... must do it at 2 locations.
1 is small check valve #2 for turbo-wastegate and the other one is the current location my sensor is permanently hosed to, which is the brake booster prepared hose barb/nipple
for vacuum sensing. The test can detect if any of the silicone diaghpram of the check-valves is leaking too.
I think using cotton bud and swipe the check-vaves inner pipe for oil trace and also its hoses, would be a good inspection method too.

============= ADD===========

Since all vacuum pumps are rated by air flow, albeit negative air flow in CFM or Liters per minute, I can use my scuba tank as a measuring tool/container.
The tank I will use is a 10.94 liters of internal volume scuba tank called AL80 in USA or the diving world.
I will need to plumb both the small and big hose of the vacuum pump to a single manifold and hose.
1 ATM is 14.7 psi absolute at sea level or very near 1 BAR and that is 10.94 liters (11.56 quarts ) worth of air in the AL80.
The lowest I seen on my vacuum pump is -14.52 PSI . I can use 14.2PSI as target.
I then clock the time needed to get -14.2PSI as baseline. Every once per year I can repeat the test and see the declining performance.
Engine Idle must be at always the same speed, so it has to be at engine operating temperature like 90C or higher.

For peak vacuum capability, my current Banks Gauge is always logging the negative pressure at the brake booster, that is bonus to also see
trend in declining performance.

CaliBenzDriver

these are excellent questions! Even Master Tasos recommends dumping oil out of vacuum pumps, duh!

The touble with this pump is liquid can not be compressed. Once filled with oil, any additional incoming pressure has few places to go. High pressure then goes towards low pressure...

I think the answers towards a solution are centered around pressure differences:

• Crankcase internal pressure
• vaccum pump chamber

-- What happens to vaccum pump capacity once its chamber is half filled by oil up to outlet reed level ?

-- How does incoming oil pressure counter act inner chamber vacuum? The check-valve witnesses the pump pressure getting very positive.

Even my NA does have some oil found past the small vacuum check valve and brake line as well.... swampy near valve but not filled up.
The vacuumed in the line sucks up what's near check-valve: ok!
We can guess the vaccum stored inside the actuated diaphragms of turbo/plenum acts as... a vacuum cleaner helping to suck oil inside.

I see blown check-valve (OP; Tasos) as signs of abnormally high pressure being present inside the vacuum pump. Imagine trying to pump down vacuum under these conditions. Turbo gate likely to have zero vacuum available for actuation!

This two chambers vaccum pump has timings during 360° rotation. Thus it needs working check-valves to hold vacuum until rotor sweeps by.


+++++++ Move check valve away !!
Because the enemy of our vaccum pump is pressure, how about we move the check-valve function away passed a pressure relief....?

Leave the nipple with missing diaphragm in place.

• Add a T on a pipe with two check-valves.
• One check-valve holds vacuum towards wastegate
• One check-valve dumps pressure out.

The outcome is pump is now pressure relieved. It has no pressure, only vacuum. Wastegate is protected.

this maybe a case of "new things, new problems": oil is going to drain out through relief valve, more oil is going to come into pump - We really need to limit incoming oil!!

+++ Booosted booster....
For the past 2 weeks, I've noticed that my brake pedal has becomed stronger and more touchy proportional.
I am going to attribute that to
oil valve trick sealing cylinders better with wet rings... less crankcase pressure...
vacuum pump less counter acted
30% more vaccuum available to booster
✌️

S-Prihadi

Yep...it is confusing to imagine in accuracy to a 100%.... the working of this vacuum pump due to its unique OIL IN and OIL OUT system, where such liquid will
make pump air space becoming a variable one. Where manipulation of pump internal air space ( without worrying of oil eating up that space ) is the one dictating the vacuum pumping
efficiency/volume per 1 revolution.

CaliBenzDriver

Essentially the vacuum pump and diaphragms (plenum; wastegates) are affected by crankcase issues. Fix thoses issues before anything.

-- We know what upsets pump check-valve:
high crankcase pressure

-- We know what helps vacuum output:
lower crankcase pressure.

> Then question becomes:
how to reduce crankcase blow-by and boost leak?

> better wet piston ring seals (pump solenoid!)
> better PCV exhaust diaphragm/clean pipe.
....??

✌️