How does the Brake Booster Solenoid work?

[eacmen]

We have been struggling for a year or so with brake pedal going rock hard while on track. Originally hypothesized that this was ice mode. This past weekend at VIR i’m starting to think it might be brake booster related.

The problem is intermittent. It seems to happen deeper in the session when everything is hot and/or when pushing.

My question for the brain trust is does anyone know the purpose for the solenoid that seems to drive brake booster vacuum? I would think that with an intermittent problem like this its unlikely that the booster itself is failing. It seems to hold vacuum even after the engine is off.

On normal E46 chassis it looks like the brake booster vacuum is driven directly off the intake vacuum. But on our car there seems to be a solenoid between engine vacuum and the booster.

Does anyone know the purpose of this? Under what conditions does it open/close? Is it normally open or normally closed?

[sapote]

It seems to happen deeper in the session when everything is hot and/or when pushing.

I'm not surprised at all.

Racing and pushing throttle all the time will lead to empty booster vacuum, and hard pedal symptom.

wide opened throttle has low vacuum and so not much help from the booster. Must close the throttle sometime to build up brake vac.

The solenoid is for close-off the suction path from the intake under some condition.
You can bypass the valve -- let it open by unplug the connector to it and the suction jet pump is enabled all the time, but still runs out of vac if wide open throttle most the time.

[eacmen]

Thanks sapote! How does the suction jet pump work? Is it powered via venturi?

Odd that other E46 M3 owners at the track don't encounter this. Maybe I am the only one left foot braking?

Maybe the booster is actually bad and not holding enough vacuum in the reservoir.

I should probably inspect check valve as well.

[sapote]

How does the suction jet pump work? Is it powered via venturi?

.

It works using the same principle that fast moving air above the wing lifts the airplane up.
The intake pulls air out at I port of the valve, and the same air mass is coming in at the A port connected to near the air filter. The air channel is narrowed down in the mid section of the valve which is also where the brake booster port B connected to. Air must be moving much faster through the narrow section ( I wish the same happens on the road when FWY closed some lanes but stupid human do the opposite compared to air or liquid and caused traffic jam) , and fast moving air creates lower pressure than at I port, and this lower pressure pull air out of the brake booster can to create vac needed.

Our car has a solenoid valve connected at A to be able to close this port when needed, but I am not sure under what condtion.My guess is that at wide open throttle, most of the air flow thru the opened throttle and not much flow thru this valve and so it's useless and the DME closes the valve to minimize booster leaking from B to A as the delta pressure is high.

[heinzboehmer]

I think sapote is spot on. Sucking jet pump helps generate additional vacuum for the booster and DME closes the inlet (labeled "A" in sapote's diagram), when additional vacuum is not needed. I also don't know what specific conditions drive the DME to close it though...

Have you tried logging the brake pressure sensors on track? I can't really think of any other telemetry that might help you find the problem, but at least that would be a start.

Edit: Not M3 specific, but this is what TIS has to say about the system:



A simple disassembly of the DME FW should tell you what conditions need to be met for the valve to be closed

[eacmen]

I think sapote is spot on. Sucking jet pump helps generate additional vacuum for the booster and DME closes the inlet (labeled "A" in sapote's diagram), when additional vacuum is not needed. I also don't know what specific conditions drive the DME to close it though...

Have you tried logging the brake pressure sensors on track? I can't really think of any other telemetry that might help you find the problem, but at least that would be a start.

Edit: Not M3 specific, but this is what TIS has to say about the system:



A simple disassembly of the DME FW should tell you what conditions need to be met for the valve to be closed

Brake pressure sensors dont yield any meaningful conclusions as when it happens I seem to make up the difference with additional pressure. So peak pressures don't really change but the brake rate seems to be more gradual build up to the same pressure.

I don't think the DME can know how much vacuum is in the booster so it must simply open or close the solenoid based on other parameters.

As much as I would like to log values this condition can only be replicated on track so opportunity to test it is limited. I might just need to fire the parts cannon and hope for the best!

[sapote]

I think your booster can is leaking either at the front hose connection or the rear seal.

To do: park it overnight and test the brake with engine off; good sealed can should still have vac and brakes fine but leaking can will cause hard pedal.

[eacmen]

I think your booster can is leaking either at the front hose connection or the rear seal.

To do: park it overnight and test the brake with engine off; good sealed can should still have vac and brakes fine but leaking can will cause hard pedal.

Car has been parked for about a week. I just went out and there was still vacuum assist in the booster. So I think the booster is at least holding some vacuum.

However what I did discover is when trying to command the "Sucking jet pump" on and off with the Shwaben scan tool the voltage to the solenoid did not change. It was constant 12v - meaning that solenoid is always open.

Can someone else with the schwaben tool confirm that commanding "Sucking jet pump" on and off in the Active Test menu commands the brake booster solenoid on and off? If so, then either a defect with the ECU or its been coded out with a tune.

[eacmen]

More info:

I removed the hose feeding the brake booster that has the sucking jet pump and the one way valve. In an attempt to test each component:

- Cleaning it real good and using my mouth verified that the one way valve seems to be working (though my lungs likely cannot generate the same amount of vacuum as the booster holds). However, if the one way valve was defective I would assume the booster would leak vacuum soon after shutting engine off. Vacuum stayed in the booster for several days.
- I verified the sucking jet pump was working by sending compressed air down the A inlet (diagram above) and holding my thumb on the valve that goes into the booster. I could feel vacuum on my thumb.
- I verified that at least under normal scenario the brake booster solenoid is closed with no voltage applied and open when voltage is applied. Even though I could not test the actuation of the solenoid itself. Maybe there is a tool32 entry I could find to do this?

Really this all just left me with more questions than answers. Was hoping one of these components would be more obviously faulty. Though I am not surprised as this condition only happens with a hot engine deep in a track session while pushing hard. Steady state on th the bench in the garage hard to replicate. Maybe I can put some of these components in the oven and replicate heat soak and test again?

Any ideas are welcome

[George Hill]

What is the failure mode?

Is it hard like no assist all the time or intermittently?

If intermittent what is the pattern?

[eacmen]

What is the failure mode?

Is it hard like no assist all the time or intermittently?

If intermittent what is the pattern?

Rock hard pedal, pressing harder does yield more braking force.

Happens intermittently.

Typically starts 10+ minutes into a session when things are nice and hot. Sometimes the following braking zone it gets better, sometimes it takes 3 braking zones.

[George Hill]

Does it happen after a long WOT scenario?

Does it happen after low throttle scenarios?

My car did this intermittently, the booster would bleed down when WOT for extended period of times and lose assist. Once I got into situations where the engine was in a state building vacuum the assist would come back. The booster would hold vacuum with the car off as well. It also made a little bit of a swoosh sound intermittently.

[bigjae46]

So far in this thread...it describes how the solenoid valve works incorrectly. The solenoid is open or closed so it controls flow from the air rail to the booster back to the intake manifold.

The solenoid is connected to the ECU. Then it controlled by the hall effect switch on the brake pedal, LCU, and ABS unit. When the hall effect switch is de-energized (you step on the pedal) it opens the circuit to the ECU. Forget what it is, its a blue wire with red dots, might be a bus loop, IIRC it feeds into the ABS, ECU, and LCM. It is a blur but the ECU is reading the hall effect switch position and then the ECU opens and closes the solenoid based on the hall effect switch.

Awhile back, I was trying to delete the LCM so I replaced the hall effect switch with a momentary push button to activate the brake lights. After driving about 100 feet the brakes would start to stick on and lock. If I shut the car off and restarted, the car would go another 100 feet and then the brakes locked. I believe the solenoid is closed when it is deenergized - that where I fixed the issue and stopped investigating.

The T is a 2 way valve, I looked inside of it. It is not pumping anything.

[heinzboehmer]

It's not an active pump, but rather works on the venturi principle, just like the stock suction jet pump in the gas tank.

See the "Air Ejector" section here: https://en.wikipedia.org/wiki/Vacuum_ejector

[sapote]

1. When the hall effect switch is de-energized (you step on the pedal) it opens the circuit to the ECU. Forget what it is, its a blue wire with red dots, might be a bus loop, IIRC it feeds into the ABS, ECU, and LCM. It is a blur but the ECU is reading the hall effect switch position and then the ECU opens and closes the solenoid based on the hall effect switch.

2. Awhile back, I was trying to delete the LCM so I replaced the hall effect switch with a momentary push button to activate the brake lights. After driving about 100 feet the brakes would start to stick on and lock. If I shut the car off and restarted, the car would go another 100 feet and then the brakes locked. I believe the solenoid is closed when it is deenergized - that where I fixed the issue and stopped investigating.

3. The T is a 2 way valve, I looked inside of it. It is not pumping anything.

1. I don't see the reason for open or close the booster solenoid based on the brake pedal actives or not active. The booster should always be pumped unless there is not enough vac in the intake such as with WOT. Why would the DME disables the pump during braking, or during not braking (how would it get vac boosted if pump not enabled during not braking)?
2. Solenoid is opened when no power applied to it. TIS says so. I don't think the state of the brake booster has any effect on why your brake was locking up.

3. It's not just a 2 way valve, but it amplifies the vac of the intake to create higher vac for the brake booster. The intake vac level is not strong enough for the booster to work.