CDV Hydraulics Explained

[wsmeyer]

*** Edited 1-27 after moved to new thread ***

To understand why I'm writing all of this I'll start with an analogy and my experience removing my CDV.

__________________________________________


Let's say that "Steve" on another forum states that he has figured out how to remove the top speed governor on our cars and writes up a DIY so that others can do the same. Then "Bob" posts that he did the mod, and while he hasn't been able to test the top speed, he posts that the engine has more power and the car runs much smoother at all driving speeds.

Since I haven't driven Bob's car either before or after the mod, I can't say that his observation is wrong, but there's really only two possibilities:

A) His observations are wrong and it's all in his head.

OR

B) Something else is responsible for the observed changes because whichever method BMW uses to limit the top speed, it's not going to have any effect on the car at normal driving speeds.


__________________________________________


That's where I'm at with the CDV. I had read about removing it on countless forums and the consensus, from those that posted anyway, was that it was a simple and virtually free mod that resulted in a much better "feel" to the clutch. All sorts of adjectives were used to describe the difference, and since the CDV is a restictor valve, they certainly all sounded plausible so why not remove it?

It sat on my mod list until about a year ago when I needed new brakes. I figured since I was going to be completely fushing out the fluid it would be the perfect time to replace the brake and clutch lines with stainless steel ones and remove that pesky CDV. It was really easy, I just took the old line out with the CDV and replaced it with the new stainless steel line. It was when I got to the flushing and bleeding part that I realized that the whole clutch system wasn't at all how I had imagined it in my head. I could see that it was just two rigid mechanical lever systems, joined together by a simple hydraulic line. That meant that virtually NONE of the claimed benefits of removing the CDV were even possible.

After all of that, I DID feel a difference in my clutch, and it definitely felt better, but I had also changed three, possibly four things at once:

1) Rremoved the CDV.
2) Swapped out the rubber line for a stainless steel one.
3) Completely flushed the fluid.
4 - possibly) If there was air in the system, it was bled out.


So now I'm not sure what caused what, but I am sure of what is and isn't possible and is the basis for my opinion:

IF YOUR NOT SIDE STEPPING THE CLUTCH OR RELEASING IT SO FAST THAT YOUR FOOT IS COMING OFF THE PEDAL, REMOVING THE CDV WILL NOT HAVE ANY EFFECT


To show why I don't believe the CDV can explain the changes that people are observing, I'll tackle four things:

1 - Hydraulic systems and their characteristics
2 - The layout of the clutch system in our E 46's
3 - The CDV valve itself and what it can and cannot do
4 - Pneumnatic systems and how their characteristics can explain a lot of the observations



***End edit ***



Part 1 - Hydraulic Systems and their characteristics

A hydraulic system is nothing more than a tube filled with an incompressible fluid and a piston at both ends. The idea is to use the incompressible property of the fluid to transmit movement of one piston into movement of the other piston. In the case of a clutch or brake line, where the forces are pushing liquid into the hydraulic line, ANY movement on one side, MUST have a corresponding movement on the other side. More precisely, as Bernoulli figured out a few hundred years ago, a volume of movement on one side will result in the exact same volume of movement on the other side.

7052

Diagram(A) is the simplest hydraulic system and is the type used for the clutch in our cars. Because the fluid is incompressible you can see that if you moved Piston(1) one inch, Piston(2) would also have to move one inch. If the clutch line was perfectly straight as in the diagram, you could just replace it with a metal rod and it would behave exactly the same. What's cool about hydraulics though is that it doesn't matter how long the line is or how many bends there are in it, all that matters is that the fluid is incompressible.

Diagram(B) is a little more complicated and is a simple diagram of what is used in the braking system. Let's assume:

- Piston(3) has an area of 10 square inches and represents the COMBINED surface area of all of the caliper pistons.
- Piston(4) has an area of 1 square inch
- you push the brake pedal attached to Piston(4) with 100 lbs of force and it travels one inch


When you exert 100lbs of force on the pedal, because I made Piston(4) one square inch in area, the pressure in the fluid is 100PSI - Pounds Per Square Inch.

On the other side of the system, the 100PSI is exerted over 10 square inches so the force on Piston(3) = 1000lbs

The idea of turning 100lbs into 1000lbs might seem like black magic but it's really not and there is another ways to show it.

Using Bernoulli's Principle, the one inch square Piston(4) moved one inch which means one cubic inch of fluid was moved. On the other side, Piston(3) has an area of 10 square inches, so one cubic inch of fluid would move it one tenth of an inch:

Movement(4) = 1 inch --> We'll call this Distance(4) or just D4
Movement(3) = .1 inch --> We'll call this Distance(3) or just D3

Newton figured out a few hundred years ago that all energy within a system is conserved:

Work is defined as Force x Distance --> W = F x D

To conserve the energy in our system, the work done on one side of our system must equal work done on the other side.

Above I set the force on Piston(4) = 100lbs -- We'll call this F3

Now we have the equation:

F3 x D3 = F4 x D4

And solving for the unknown F3 gives us:

1000lbs x .1in = 100lbs x 1in

Thats the principle behind the hydraulic brakes in our cars, and the hydraulic jack you use to lift it; you're converting a small force over a large distance, into a large force over a small distance.

[wsmeyer]

Part 2: The clutch system in our cars.

(reserved)

[wsmeyer]

Part 3: The CDV. What it is, and what it can and cannot do.

(reserved)

[wsmeyer]

Part 4: Pneumatic systems and how it can explain many of the observations.

(reserved)

[kayger12]

Moved to its own thread.

PM me what you'd like the title to be and we can change it.

Looking forward to reading the rest.

[Mtnman]

Subscribed for the knowledge drop. So far so good.

[Oli77]

Cool stuff, combine that with a compressible gas in a closed system and you have the Citroen 1960's hydraulic suspension.

[echo46]

I'm all in.

[Crickett]

Ooh, science!! I'm all ears!
:idea

[derbo]

I am very curious about the physics behind it :)

[wsmeyer]

I haven't forgotten about this but I've been too busy the last few days to add much to it. I did add something of an intro after it was moved to it's own thread, an in preparation for explaining the CDV I carefully dismantle mine the other day and took pictures of it. Just like the whole clutch system, it wasn't at all how I had imagined it in my head and I spent a lot of time wondering in my head why they made it the way they did. Then, all of a sudden, it was like a light bulb going off in my head and I am now completely convinced that I know why they put it there and what it is supposed to accomplish. It's not a safety feature at all, it's there to reduce driver fatigue.

[derbo]

I haven't forgotten about this but I've been too busy the last few days to add much to it. I did add something of an intro after it was moved to it's own thread, an in preparation for explaining the CDV I carefully dismantle mine the other day and took pictures of it. Just like the whole clutch system, it wasn't at all how I had imagined it in my head and I spent a lot of time wondering in my head why they made it the way they did. Then, all of a sudden, it was like a light bulb going off in my head and I am now completely convinced that I know why they put it there and what it is supposed to accomplish. It's not a safety feature at all, it's there to reduce driver fatigue.

I appreciate you taking the time to even write it to this point! Driver fatigue eh? That is definitely an interesting point and I'd love to hear the reasons you came to that conclusion.



After all of that, I DID feel a difference in my clutch, and it definitely felt better, but I had also changed three, possibly four things at once:

1) Rremoved the CDV.
2) Swapped out the rubber line for a stainless steel one.
3) Completely flushed the fluid.
4 - possibly) If there was air in the system, it was bled out.


From my own experience:

I replaced my clutch in June 2012. During clutch job, I also bled my clutch with a full brake flush. My latest brake/clutch flush was in Aug 2012. In Sept 2012, I removed my CDV and rebled the clutch and brakes again. I am still using the rubber clutch line with no "dummy" CDV in place. I certainly feel a difference in my clutch pedal even with use the CDV deleted.

While this doesn't exactly eliminate all the variables, it does help gauge which variables are more of a factor to the feel of the clutch. My "feel" of the clutch may and very well be completely different from the "feel" your car has. I just think the CDV alone has a significant contribution to the change in the feel compared to the rest of the variables.

[RITmusic2k]

I'm going to copy a couple of my posts from another thread here - I, too, tried to dispel some rumors and misinformation surrounding what the CDV delete does, and I think they'll serve well enough to add to the discussion:


The biggest benefit you get from the CDV delete is consistency and feedback.

The thing to remember about the CDV is that it is a limiting factor, not a scaling factor. In other words, the CDV doesn't make the clutch close at some set fraction of the speed that your foot moves on the pedal - it lets the clutch close at the same speed your foot moves on the pedal *up to a certain point*, after which any faster motion from your foot doesn't increase the speed at which the clutch closes.

The effect from the driver's standpoint is that while at slow pedal-release speeds the clutch engagement point correlates nice and directly with clutch pedal position, but when you try to release quickly, the engagement point moves artificially (and inconsistently) higher up the pedal's travel. This throws your muscle memory out the window, and most often results in you applying throttle too soon, which translates into excessive and unnecessary clutch slippage.

For that reason,


I never did this mod but heard that it made the biggest primary difference in paralell parking where precise modulation is key. Is this true for those that have done it?


the thing you heard is exactly backwards - at low engagement speeds you will not notice a difference, but when driving aggressively and shifting quickly, there's a significant difference between cars with and without the CDV installed.

At the end of the day, rev-matching your upshifts and downshifts does much more to prolong the life of the entire driveline, and it's only possible to learn how to do this properly on a car without a CDV installed.





for me the biggest improvement was in low speed parking situations where you're on & off the clutch pedal in 1st gear & reverse. Should be mod #1 on any model that has it.
Hmm, that brings up a bit of detail that I may have left off my reply - the CDV's effects aren't based on *vehicle* speed, only on *clutch pedal travel* speed. So during parallel parking or other low vehicle speed scenarios, I can see how it could still come into play - especially if you quickly 'pulse' the clutch pedal, or really do anything at all that would exceed its restricted flow rate.

[midlandtech]

Well written guys

[BavarianZHP]

Hmmm. Maybe a blind test is in order? Have someone remove, install, or do nothing to another person's car (or many cars/drivers to increase sample size) and not inform the driver(s) for a month... The CDV may or may not be deleted every weekend... Record the driver's observations on the feel of the clutch on a per drive basis. If it's fairly evident that the CDV does have a significant impact on the clutch, the driver's observations will improve with the removal of the clutch. Seems simple enough. Of course you can choose to trick the driver by stating a mix of true and misinformation by saying sometimes, "yeah, I removed the CDV" (when in fact you didn't and when in fact you did, etc....).

[derbo]

I think the biggest difference is felt during downshifts. The clutch pedal can be released faster during a downshift without a CDV. With a CDV I feel like I have to leave my foot near the engagement point until it finally engages before fully removing my foot off the pedal.



Sent from my iPad using Tapatalk

[RITmusic2k]

I'm absolutely in agreement with you, Derek. It's a scenario that generates repeatable results highlighting the difference between CDV and no CDV.

When I first got my car, I tried dropping two gears and stomping on it to see how quickly the car could get out of the way. Coming from a car where pedal position corresponded exactly to clutch disc position, I got hard on the gas as soon as my foot was off the clutch pedal, and the engine revs surged and settled like a failing clutch would, because, well... the clutch was slipping exactly the same way due to that restricted flow.


Actually, the above is a little simplified... the truth is that the clutch disc makes contact with the flywheel much sooner than the perceived "engagement point"... what we typically think of as clutch disc travel is actually more like "clutch disc travel" over the first 30% of clutch pedal travel followed by "progressively increasing clamping force" over the remaining 70% of clutch pedal travel; the engagement point we feel is somewhere in that 70% range where you hit a threshold of sufficient clamping force to hold the clutch disc and flywheel together without slippage.

That threshold amount of force rises slightly with increasing disparity between the rotational speeds of the clutch disc and flywheel (e.g., when revs are poorly matched), but for the most part it's in the same place. More accurately, the relatively-fixed amount of clamping force required to engage the clutch corresponds with a relatively-fixed position in the clutch pedal's travel. Or it should, if there's no flow restriction.

As quickly as you can get to that position, and thus that amount of clamping force, is how quickly you can get on the throttle after a shift. But the rate of application of that clamping force is what the CDV limits. So if your foot makes it to the point where that amount of clamping force is normally applied sooner than that force is actually applied, you'll get slippage until the CDV allows enough fluid through. In my experience, my foot could get there an entire second before the CDV caught up. That's enough time for some very annoying over-revving and slippage.

[ryankokesh]

Great thread!

http://i.qkme.me/3qavhx.jpg

[terraphantm]

When I did the CDV on my ZHP, I kept the stock line and did not change the fluid. If anything I introduced air into the system (but I doubt I did since I didn't lose a single drop). I did not feel a night and day difference as some report. However, I did find 1-2 shifts to become considerably smoother and found myself with an improved ability to feel the engagement point.

I want to point out that this explanation is inadequate. In real life, most systems do not perfectly follow the ideal cases we're taught in high school physics. The biggest thing to note is that the restriction in the CDV is not a smooth reduction; it suddenly becomes a very small diameter. This does make a difference in hydraulics and probably accounts for the difference most people feel. Look in any introductory physics textbook -- every hydraulics problem with diameter changes will specifically note that the change is achieved smoothly. A sudden decrease in diameter will cause resistance and energy loss.

More info: http://books.google.com/books?id=rbcTeQ4UqsoC&pg=PA140&lpg=PA140&dq=hydraulic+losses+sudden+contraction&source=bl&ots=a-yfAZU0uF&sig=IWhlCUMWLaT5A-EPmfws-knYX3g&hl=en&sa=X&ei=c3EPUe3HNYyp0AHT34CYDg&ved=0CEMQ6AEwAg#v=onepage&q=hydraulic%20losses%20sudden%20contraction&f=false

[llll1l1ll]

I'm looking forward to the next installments!

[terraphantm]

For anyone who has doubts about the CDV, try this:

Remove your CDV and clean it really well
Blow into both sides

You'll notice that the return direction has a ton more resistance than the release direction. If you can feel that much resistance with your breath, there will definitely be a very large pressure drop in the actual system too. The reason the pressure drop is only in one direction is because the CDV has a mechanism that allows fluid to go around the restriction in that one direction.

[BavarianZHP]

Man, I REALLY need to get rid of mine. Every time I have a bad shift I blame the CDV. LOL

[RITmusic2k]

Haha, nothing better than having nothing but yourself to blame :)

[wsmeyer]

For anyone who has doubts about the CDV, try this:

Remove your CDV and clean it really well
Blow into both sides

You'll notice that the return direction has a ton more resistance than the release direction. If you can feel that much resistance with your breath, there will definitely be a very large pressure drop in the actual system too. The reason the pressure drop is only in one direction is because the CDV has a mechanism that allows fluid to go around the restriction in that one direction.


Not exactly, there is no cut off mechanism in the CDV therefore it cannot limit pressure. It simply restricts the flow of the fluid to reduce the power of the return stroke.

[terraphantm]

Not exactly, there is no cut off mechanism in the CDV therefore it cannot limit pressure. It simply restricts the flow of the fluid to reduce the power of the return stroke.

http://en.m.wikipedia.org/wiki/Orifice_plate

Pressure (and velocity) will be greater before the restriction than after. It doesn't limit pressure in the sense that the pressure cannot exceed X psi. But there is certainly a pressure drop across such a restriction.

[wsmeyer]

Pressure (and velocity) will be greater before the restriction than after. It doesn't limit pressure in the sense that the pressure cannot exceed X psi. But there is certainly a pressure drop across such a restriction.

Nope.

See the section titled "Incompressible flow through an orifice" in the link you gave.

[danewilson77]

Does there not have to be a pressure drop for the equation to hold true?

http://i1015.photobucket.com/albums/af278/nicee46/Maintenance%20and%20Misc/Pressuredrop.jpg (http://s1015.photobucket.com/user/nicee46/media/Maintenance%20and%20Misc/Pressuredrop.jpg.html)

[Mtnman]

:scratchinghead So carry the one, then divide by......oh hell. gonna have to wait till little man goes to bed to digest that equation.

[danewilson77]

Additionally, this chart shows expected pressure drop through a cylindrical sharp edged orifice, which I think the CDV essentially is.

http://i1015.photobucket.com/albums/af278/nicee46/Maintenance%20and%20Misc/Pressuredrop2.jpg (http://s1015.photobucket.com/user/nicee46/media/Maintenance%20and%20Misc/Pressuredrop2.jpg.html)

http://www.womackmachine.com/engineering-toolbox/design-data-sheets/flow-through-orifices.aspx

The only place I can think of where an "orifice" is encountered with no corresponding pressure drop is in a stage (Moving blade-fixed blade-moving blade) through a reaction turbine.....but then again....steam isn't an incompressible fluid either.

http://i1015.photobucket.com/albums/af278/nicee46/Maintenance%20and%20Misc/Pressuredropsteamturbine.jpg (http://s1015.photobucket.com/user/nicee46/media/Maintenance%20and%20Misc/Pressuredropsteamturbine.jpg.html)

[Avetiso]

After going over hydraulics in my physics class just last week, this all makes sense.

[terraphantm]

Nope.

See the section titled "Incompressible flow through an orifice" in the link you gave.

Read that section a little more carefully. You forgot to take into account losses due to friction and turbulence; that's why that dimensionless coefficient is there.

I also refer you to the section titled "Permanent pressure drop for incompressible fluids"

[Dave_B]

Brain hurts. Will just pull the damn thing next time I'm under the car.

[wsmeyer]

I think the biggest difference is felt during downshifts. The clutch pedal can be released faster during a downshift without a CDV. With a CDV I feel like I have to leave my foot near the engagement point until it finally engages before fully removing my foot off the pedal.

Because fluids are incompressible, hydraulic systems cannot store potential energy.

Gases are compressible so pneumatic systems can store potential energy.

If there is movement in your system after you stop moving the pedal you have air in the lines = pneumatic system.

[danewilson77]

Because fluids are incompressible, hydraulic systems cannot store potential energy.

Gases are compressible so pneumatic systems can store potential energy.

If there is movement in your system after you stop moving the pedal you have air in the lines = pneumatic system.

Yep.....but pressure drop will still occur through an orifice with flow, correct?

[wsmeyer]

I'm absolutely in agreement with you, Derek. It's a scenario that generates repeatable results highlighting the difference between CDV and no CDV.

When I first got my car, I tried dropping two gears and stomping on it to see how quickly the car could get out of the way. Coming from a car where pedal position corresponded exactly to clutch disc position, I got hard on the gas as soon as my foot was off the clutch pedal, and the engine revs surged and settled like a failing clutch would, because, well... the clutch was slipping exactly the same way due to that restricted flow.


Actually, the above is a little simplified... the truth is that the clutch disc makes contact with the flywheel much sooner than the perceived "engagement point"... what we typically think of as clutch disc travel is actually more like "clutch disc travel" over the first 30% of clutch pedal travel followed by "progressively increasing clamping force" over the remaining 70% of clutch pedal travel; the engagement point we feel is somewhere in that 70% range where you hit a threshold of sufficient clamping force to hold the clutch disc and flywheel together without slippage.

That threshold amount of force rises slightly with increasing disparity between the rotational speeds of the clutch disc and flywheel (e.g., when revs are poorly matched), but for the most part it's in the same place. More accurately, the relatively-fixed amount of clamping force required to engage the clutch corresponds with a relatively-fixed position in the clutch pedal's travel. Or it should, if there's no flow restriction.

As quickly as you can get to that position, and thus that amount of clamping force, is how quickly you can get on the throttle after a shift. But the rate of application of that clamping force is what the CDV limits. So if your foot makes it to the point where that amount of clamping force is normally applied sooner than that force is actually applied, you'll get slippage until the CDV allows enough fluid through. In my experience, my foot could get there an entire second before the CDV caught up. That's enough time for some very annoying over-revving and slippage.



The movement of the diaphragm spring inside the pressure plate is a bit complex but hopefully you can see that there is a direct correlation between the clamping force and the linear travel of the throwout bearing. The clamping force is variable from 0-100% just as the linear movement of the throwout bearing is variable from 0-100% Whatever the correlation is it is fixed. if 20% linear movement of the throwout bearing yield a 50% reduction in clamping force, that will always be the same.

Now realize that the amount of linear travel of the throwout bearing is fixed to the volume of fluid that is pushed into the slave cylinder.

Since the CDV has no ability to store fluid, any fluid pushed out of the master cylinder, must go into the slave cylinder.

It doesn't matter which direction you are analyzing:

volume in = volume out

Therefore:

THE MOVEMENT OF THE THROWOUT BEARING RELATES LINEARLY TO THE MOVEMENT OF THE CLUTCH PEDAL

There is no exception to this, unless of course, you have compressible air in the lines.

[terraphantm]

Of course volume in = volume out. But that doesn't mean the CDV can't slowdown the whole system. And that's exactly what it does. Everything eventually goes where it needs to go, but it takes a little (and I do mean little -- the delay is probably small fraction of a second) longer. The clutch return spring is still there, and the fluid does move with the pedal. But there is enough resistance to prevent hard launches and make it difficult to perform shifts which require precise timing.

Think of drawing fluid through a syringe with a needle attached vs. one without a needle. If you want to draw fluid at the same rate through both, you'll need to exert a lot more force on the former (and there's generally a small empty space, meaning that you need a pretty large pressure gradient to move the fluid that quickly). Why is this? It's because the resistance through the tiny needle opening is quite small. You can think of the CDV in a similar manner -- the CDV is more restrictive than a direct connection between the lines. The diaphragm spring and clutch return spring exert the same force regardless of whether or not there's a CDV. As such it is reasonable to conclude that the increased resistance delays everything. As for why the difference is noticeable in lower gears than in higher gears, that's because you don't really need full clamping force to prevent the clutch from slipping in higher gears.


At the end of the day, all of the empirical evidence disagrees with your conclusion. Almost every single person who has done this mod noticed a difference, whether or not they bled the clutch (I've personally never bothered bleeding the clutch both times... mostly because I noticed my clutch line had blue fluid in it even though I never actually bled my ATE superblue through it). So either everyone is wrong and you're more well versed in hydraulics than every BMW owner, or your understanding of the mechanisms are flawed. Take your pick.

[wsmeyer]

I've personally never bothered bleeding the clutch both times

I'm not the least bit concerned about what other people think or want to believe and quite frankly, your statement above demonstrates that you really don't understand how hydraulic systems operate and negates all of you observations.

[wsmeyer]

Does there not have to be a pressure drop for the equation to hold true?

That equation would be used to calculate the movement of fluid from a high pressure area to a low pressure area.

If you had a hole in your line, P1 would be the pressure in the line and P2 would ~ 15psi (1atm) and the amount of fluid spraying out would depend on the size of the hole.

If your head gasket is leaking P1 would be the oil pressure ~ 60psi and P2 would be whatever the relief valve in the ET cap is ~ 20psi and again the flow of oil into the coolant would depend on the size of the hole.


Our clutch system is a bit different though. The pressure plate is providing the resistance to movement. You apply a force to the clutch pedal that increases the pressure in the hydraulic line. From Pascal's law: "increasing the pressure at any point in a confined fluid, there is an equal increase at every other point in the container" so according to that equation P1-P2 would always = 0 and there would never be any flow.

That doesn't mean the equation is incorrect, it's just not applicable to our system.

Think about this. If there is a bucket of water on the floor and I push it a distance with my foot obviously the water moved but could you calculate it with that equation? No.

[wsmeyer]

Read that section a little more carefully. You forgot to take into account losses due to friction and turbulence; that's why that dimensionless coefficient is there.

I also refer you to the section titled "Permanent pressure drop for incompressible fluids"

I didn't forget about anything. If this was an assignment in a Physics class we would have to account for a whole bunch of stuff that I neglected because I believe they are insignificant compared to the larger forces in our system:

compressibility of the fluid
amount of gasses in the fluid and it's effect on density
compressibility of the gasses in the fluid
friction losses through the CDV

There's probably more but if you tink they are significant you should start with friction as it would be the largest.

The orifice in the CDV is .060 inches in diameter and .0165 inches in length.

If we call the friction loss through the CDV 'f' and the friction loss through the rest of the system 'F' then f+F / F would give the error factor of neglecting it.

[Mtnman]

I really enjoy reading all of this, and enjoy the debate we can have on our site. Unlike other sites, we don't lob grenades from the back of the fight or make snide comments to others of differing opinions. Im glad this thread is here. Lets continue this discussion the same way it started. Facts, observations, and principle of physics. I look forward to more info Wsmeyer, terra, and dane. I have a hard time shifting 1-2 compared to other cars, and want to find out why before i spend the time to remove something bmw thought should be there.

[danewilson77]

Great discussion in here.

Let's ensure we all come into this thread with open minds, and all in the best interest of learning the system better.

Let's not resort to the dialogue that a thread like this would produce at other forums.

HTC DNA, Williamsburg, VA

[RITmusic2k]

The movement of the diaphragm spring inside the pressure plate is a bit complex but hopefully you can see that there is a direct correlation between the clamping force and the linear travel of the throwout bearing. The clamping force is variable from 0-100% just as the linear movement of the throwout bearing is variable from 0-100% Whatever the correlation is it is fixed. if 20% linear movement of the throwout bearing yield a 50% reduction in clamping force, that will always be the same.

Now realize that the amount of linear travel of the throwout bearing is fixed to the volume of fluid that is pushed into the slave cylinder.

Since the CDV has no ability to store fluid, any fluid pushed out of the master cylinder, must go into the slave cylinder.

It doesn't matter which direction you are analyzing:

volume in = volume out

I concur with everything here; this is the stuff nobody gets to argue about ;)

However -


Therefore:

THE MOVEMENT OF THE THROWOUT BEARING RELATES LINEARLY TO THE MOVEMENT OF THE CLUTCH PEDAL

Here's where I suspect there's something else at play. I concur that the above would be true if the pedal itself were hard-connected to the rest of the system. But from what I've observed it really seems like the pedal was springing up as quickly as my foot allows it to, though faster than the diaphragm spring would be pushing it, based on the delay in full engagement.

This would only be possible if the pedal wasn't hard-coupled to the system, and there would have to be a secondary "pedal return spring", and therefore should be easy to verify by inspecting the mechanism in the footwell. If that's not the case, I'll have to conclude that my foot was actually coming off the pedal sooner as the pedal rose more slowly than it would have unrestricted, but I just didn't notice. I almost want to reinstall my CDV to do some A/B testing and really clarify my observations...

[wsmeyer]

I concur with everything here; this is the stuff nobody gets to argue about ;)

However -

Here's where I suspect there's something else at play. I concur that the above would be true if the pedal itself were hard-connected to the rest of the system. But from what I've observed it really seems like the pedal was springing up as quickly as my foot allows it to, though faster than the diaphragm spring would be pushing it, based on the delay in full engagement.

This would only be possible if the pedal wasn't hard-coupled to the system, and there would have to be a secondary "pedal return spring", and therefore should be easy to verify by inspecting the mechanism in the footwell. If that's not the case, I'll have to conclude that my foot was actually coming off the pedal sooner as the pedal rose more slowly than it would have unrestricted, but I just didn't notice. I almost want to reinstall my CDV to do some A/B testing and really clarify my observations...

Excellent!

You are now one step away from where the light bulb went off in my head.

The pedal is physically attached to the clutch master cylinder, it looks like this:

http://c1552172.r72.cf0.rackcdn.com/329296_x600.jpg

And there is a spring attached to it, it's part #8 in this diagram:

http://www.realoem.com/bmw/showparts.do?model=BW53&mospid=47707&btnr=35_0186&hg=35&fg=05

It's 13mm in diameter and looks like this:

http://www.ecstuning.com/ES63485/

There are also springs inside the master cylinder. They aren't there to bring the pedal back up though, they are there to ensure that the piston in the master cylinder is retracted back past the inlet passages after the pressure has been released. If these weren't there, as the clutch disc wore out the pedal would come to a stop farther and farther towards the floor. More about that can be read here:

http://www.tegger.com/hondafaq/mastercylinderreplace/howworks.html

Remember, when I first realized all of this was when I was bleeding my clutch line. If you hold the pedal to the floor and have someone open the bleed valve, the pressure of the pressure plate spring releases and pushes fluid out of the system. The fluid stops when all of the pressure is released. At that point, when you close the valve and take your foot off the pedal, it's no longer the pressure plate pushing the pedal back up, it's those small springs, which on my car took about 10 seconds.

[LivesNearCostco]

Question for WSMeyer... since I didn't understand all the details of how this thread get started. Are you saying that the CDV doesn't do what the BMW engineers intended, or it doesn't do what most of us enthusiasts and fanatics think it does, specifically that removing it doesn't provide the commonly ascribed benefit? I'm assuming that BMW engineers are good enough that the CDV does what they think it does, and removing it must do something, even if BMW might have mis-judged what drivers wanted or we drivers might not really know what removing it does.

My car came with it pre-removed so I don't know what it's like to have one. Sometimes my car is balky going from 1st to 2nd but my previous indie (whose sometimes drove his wife's ZHP 6MT) said it was due to worn 2nd gear synchros.

[Avetiso]

Question for WSMeyer... since I didn't understand all the details of how this thread get started. Are you saying that the CDV doesn't do what the BMW engineers intended, or it doesn't do what most of us enthusiasts and fanatics think it does, specifically that removing it doesn't provide the commonly ascribed benefit? I'm assuming that BMW engineers are good enough that the CDV does what they think it does, and removing it must do something, even if BMW might have mis-judged what drivers wanted or we drivers might not really know what removing it does.

My car came with it pre-removed so I don't know what it's like to have one. Sometimes my car is balky going from 1st to 2nd but my previous indie (whose sometimes drove his wife's ZHP 6MT) said it was due to worn 2nd gear synchros.
Mine is balky from first to second as well. 64k miles, and I think my synchro is in good shape, but it is a rough feeling sometimes when shifting.

[wsmeyer]

Question for WSMeyer... since I didn't understand all the details of how this thread get started. Are you saying that the CDV doesn't do what the BMW engineers intended, or it doesn't do what most of us enthusiasts and fanatics think it does, specifically that removing it doesn't provide the commonly ascribed benefit? I'm assuming that BMW engineers are good enough that the CDV does what they think it does, and removing it must do something, even if BMW might have mis-judged what drivers wanted or we drivers might not really know what removing it does.

I'm certain that the CDV performs exactly as the BMW engineers designed it to. It's designed to reduce driver fatigue by limiting the power of the release stroke. Of course in some situations this may result in negative side effects, but look at the myriad of perceived benefits of removing the CDV;

Smoother shifting / no jerking / lurching
If the CDV were delaying the clutch disengagement too much you'd end up with the opposite, the slipping clutch would let the engine rev and sound like a worn out clutch. Ever heard anyone say "My clutch slipping turned out to just be my CDV..."


More predictable / non varying engagement point
As I've tried to explain above, with or without the CDV the relationship between the pedal position and the pressure plate is fixed and the engagement point will always be exactly the same. A varying engagement point is a very common problem, it's not the CDV though, it's caused by the pressurizing and equalizing of an air bubble in the line.


Easier low speed driving (parking)
I don't even know what to make of this one but I'd guess they are trying to describe this one ^^.

All of this is just more anecdotal evidence that the effects of the CDV are misunderstood because BMW has been putting these in almost all of there manual transmission cars for roughly 15 years. If in fact the effects of it were negative for 100% of the people than why is it still there? At the very least why haven't they fine tuned it so that fewer people perceive the effects as negative? Have you ever heard anyone say:

"I like the way the car shifted better with the CDV"
"I liked it better without but the wife found it more difficult to shift so I put it back in"

I've done a of of searching and I have not found anyone claiming anything other than "much better" or "no change". How is that possible?

And another thing, the people removing the CDV are all enthusiasts, the group that drives there cars the hardest, yet you rarely hear anyone breaking or prematurely wearing out their transmissions. If it's there to protect the transmission, why don't you ever hear about someone damaging theirs?






My car came with it pre-removed so I don't know what it's like to have one. Sometimes my car is balky going from 1st to 2nd but my previous indie (whose sometimes drove his wife's ZHP 6MT) said it was due to worn 2nd gear synchros.

Second gear synchros almost always wear out first. This is because if you look at the ratios of manual transmissions, the biggest jump in ratio is almost always from 1st to 2nd and therefore the 2nd gear synchro does the most work.

If they are all balky, just 2nd seems worse, try bleeding your clutch line. If you have an air bubble in the line, when you press the pedal down the air pressure compresses the air bubble first, then it expands as it equalizes and engages the clutch. This means there will be a slight delay in pushing the pedal down and clutch engagement, because of the massive difference in mass of the air and pressure plate, the faster you push the pedal down the more the air bubble will be compressed and the greater the delay will be.

[wsmeyer]

As for why the difference is noticeable in lower gears than in higher gears, that's because you don't really need full clamping force to prevent the clutch from slipping in higher gears.

What is the universal test to see if your clutch is slipping?

The virtually universal answer is to shift 2 gears higher than normal and floor it, the car will accelerate as the torque output of the engine increases and then if the engine revs "run away" you know your clutch is worn out and slipping.

Why shift to a higher gear? Because the lower the gear the fewer mph the car needs to accelerate per engine rpm of slippage before the clutch locks again. When a clutch first starts slipping due to wear this will happen so fast that it won't even be noticeable in the lower gears.


To further explain this:

The maximum amount of force that can be transferred through the clutch disc / pressure plate interface is fixed and depends on whether or not it is slipping.

The MAXIMUM would be:
FL (force, clutch locked) = Coefficient of static friction x force pressure plate
FS (force, clutch slipping) = Coefficient of kinetic friction x force pressure plate

For all materials known to man:
Both the Coefficient of static friction and the Coefficient of kinetic friction are greater than 0 and less than 1
Coefficient of static friction > Coefficient of kinetic friction


It doesn't make any difference what gear you are in, if the output of the engine is greater than FL the clutch begins to slip and the maximum force transferred becomes FS. Since FS is always less than FL the engine revs spike, but the car will continue to accelerate because some force is still being transferred. For a cars clutch disc, the coefficient of kinetic friction will be pretty high, probably .9 or higher. Meaning more than 90% of the force of the engine is still being transferred through the slipping clutch.



So either everyone is wrong and you're more well versed in hydraulics than every BMW owner, or your understanding of the mechanisms are flawed. Take your pick.

I stated in the very first paragraph of this thread that I was not looking to disprove anyone's observations, I was merely trying to explain why many of the observations cannot be attributed to the CDV and that they must be the result of something else.

[kayger12]

If the CDV were delaying the clutch disengagement too much you'd end up with the opposite, the slipping clutch would let the engine rev and sound like a worn out clutch. Ever heard anyone say "My clutch slipping turned out to just be my CDV..."


I experienced exactly this with quick shifting with the cdv installed, and have never experienced since the cdv was removed.

Sent from my SCH-I535 using Tapatalk 4 Beta

[wsmeyer]

I experienced exactly this with quick shifting with the cdv installed, and have never experienced since the cdv was removed.

I probably shouldn't have put that one in there as I can see how it would cause confusion.

I now believe I know WHY removing the CDV eliminated the clutch slip for you but it wasn't the CDV that was causing the problem per se. That's probably even more confusing but I will explain more later when I have more time.

[LivesNearCostco]

Will, thanks. I could see the CDV being designed to limit power of the release stroke.

I agree removing CDV would not affect actual clutch engagement point but couldn't it affect the perceived engagement point by slowing down how quickly the slave cylinder can release the pressure plate?

I don't see it affecting low-speed driving unless one is constantly working the clutch, which can happen in parallel parking, or starting on a hill.

I have bled my clutch line twice in 3 years and don't recall it making any difference in 1st-2nd shifting, but IIRC there weren't many air bubbles coming out. also having bled it with slave in place, it's possible I didn't get all the bubbles out. Plus my transmission has 173,000 miles.


I'm certain that the CDV performs exactly as the BMW engineers designed it to. It's designed to reduce driver fatigue by limiting the power of the release stroke. Of course in some situations this may result in negative side effects, but look at the myriad of perceived benefits of removing the CDV;

More predictable / non varying engagement point
As I've tried to explain above, with or without the CDV the relationship between the pedal position and the pressure plate is fixed and the engagement point will always be exactly the same. A varying engagement point is a very common problem, it's not the CDV though, it's caused by the pressurizing and equalizing of an air bubble in the line.

Easier low speed driving (parking)
I don't even know what to make of this one but I'd guess they are trying to describe this one ^^.

Second gear synchros almost always wear out first. This is because if you look at the ratios of manual transmissions, the biggest jump in ratio is almost always from 1st to 2nd and therefore the 2nd gear synchro does the most work.

If they are all balky, just 2nd seems worse, try bleeding your clutch line. If you have an air bubble in the line, when you press the pedal down the air pressure compresses the air bubble first, then it expands as it equalizes and engages the clutch. This means there will be a slight delay in pushing the pedal down and clutch engagement, because of the massive difference in mass of the air and pressure plate, the faster you push the pedal down the more the air bubble will be compressed and the greater the delay will be.

[wsmeyer]

I hadn't updated this thread in a long time because I really wasn't sure how to move forward. When I started this thread I had hoped that by examining the characteristics of hydraulic systems other people would also realize that most if not all of the claimed "benefits" of removing the CDV simple couldn't be attributed to the CDV. I didn't set out to prove anyone wrong, there are so many people with similar observations that there must be something going on. I just didn't believe that it was the CDV. Unfortunately the thread didn't progress because I didn't seem to convince anyone to even think of other possibilities, and most were so hung up on the CDV that they were just adamant that I was wrong because what I was proposing didn't jive with their observations.

Well, as it turns out, I was wrong. The CDV is definitely having a negative effect on the way the clutch operates. BUT, it's not because it's poorly designed or a "nanny" device there to help overcome poor technique. It's because the vast majority of us have inadvertently altered the system in a way that prevents the CDV from operating as intended. The overbearing CDV is merely one symptom of the problem, not the problem itself.

I stumbled across the key to all of this while thumbing through the Bentley service manual I purchased a little while ago. In the front of the manual, there is a chart with all the fluids listed, it all looks as expected:

9196

But notice at the bottom where it says: "NOTE - See 340 Brakes for more brake fluid information"

In section 340 under brake bleeding procedures, there's this chart:

9197

Which indicates either the use of DOT4 or DOT4 LV depending on the type of traction control system. That chart looks like this:

9198

Our cars use DSC, the one that requires DOT4 Low Viscosity.

I myself had never heard this before and I had never even heard of "DOT4 Low Viscosity".

With a little googling I found a picture of the BMW brake fluid bottle with the part number on it: http://tinyurl.com/k6nrwkx

Depending on if you buy it in a bottle or jug the part numbers are:

12 ounces - 81220142156
gallon? - 81220142155

If you search for those part numbers you'll come across this over at Turner:

http://www.turnermotorsport.com/p-11331-81220142156-genuine-bmw-part.aspx

Which clearly indicates this is the genuine stuff and is DOT4 LV. It also has the Pentosin label on it. That's not too surprising to me as I had read other places that the OEM fluid is Pentosin.

And this over at bmwpartsweb: http://www.bmwpartsweb.com/Part-Number/Product-Detail__81220142156_PEN_9BA48850.aspx

That lists those part numbers and Pentosin as the manufacturer, again, DOT4 LV.

Now we go over to Pentosin and things really get interesting. http://pentosin.net/f_brakefluid.asp

They list DOT4 for BMW up to 6/2002 and DOT4 LV for BMW's after 7/2002

So what's the difference between DOT4 and DOT4 LV? There are spec sheets for both on Pentosin's website:

DOT4 - http://pentosin.net/specsheets/Pentosin_SuperDot_4.pdf

DOT4 LV - http://pentosin.net/specsheets/Pentosin_Dot4_LV.pdf

The difference that is important here is the viscosity, they list it at different temperatures but I think they chose 40C as one of the temperatures as that would probably be a normal operating temperature ~ 104F

DOT4 = 7.9
DOT4 LV = 6.4

That means that DOT4 is 24% higher in viscosity than DOT4 LV.

If you look at the equations for fluid flow vs pressure: http://en.wikipedia.org/wiki/Orifice_plate there is a one-to-one relationship between viscosity and flow. That means that if you increase the viscosity by 24% YOU REDUCE FLUID FLOW BY 24%. As far as our cars go, we have reduced the maximum fluid flow through the CDV by 24% and that is causing the negative effects.

This also explains why I didn't feel any difference when removing my CDV. Up until that point, the dealer had done all work on my brakes and presumably had been flushing / refilling the system with the correct OEM DOT4 LV fluid.

So now the big question is WHY DO WE NEED DOT4 LV?

That I have no idea, the fact that it depends on the type of traction control system leads me to conclude that the intricate valving used requires a lower viscosity. That's just a guess and I don't know what negative effects on the ABS / DSC system there will be from using DOT4 but I'm just going to flush mine and put the DOT4 LV back in and call it a day. When in doubt, go OEM for me.

[danewilson77]

Wow.....mind blown. Great research. In effect, you're saying most diyer's don't use the correct fluid, and by removing the CDV gain the desired effect of a lower viscosity fluid.

[Crickett]

Well sheet . . . looks like it's out w/the Super Blue & back in w/the OEM!

[wsmeyer]

Wow.....mind blown. Great research. In effect, you're saying most diyer's don't use the correct fluid, and by removing the CDV gain the desired effect of a lower viscosity fluid.

Kind of, the problem is that DOT4 is a higher viscosity than what the system was designed for. The system being the brake and clutch hydraulic systems.

ONE SYMPTOM of the higher viscosity would be the overbearing CDV. Removing it will alleviate that symptom.

Other symptoms would be in the effectiveness / operation of the ABS / DSC systems. I don't really know what those would be but BMW certainly had their reasons for specifying DOT4 LV and the Bentley manual eludes to those reasons having to do with the ABS / DSC.

For me personally, I was using Motul RB600. Not because I felt that I really needed a "racing" fluid but because for a few extra $$ I would rather just purchase something I know is of the highest quality. Knowing what I know now I'm going to flush all of that out and use Pentosin DOT4 LV which I know is the correct viscosity and is also a well respected brand.

[danewilson77]

Got it. Thanks Will.

[johnrando]

Thanks William. I bought a bunch of DOT 4 LV for my brake job, so I have extra unopened liters for sale if anyone wants them. I'll sell for what I paid.

[terraphantm]

I was under the impression that the cars without the precharge pump didn't need DOT4 LV. Let me see if I could find the corresponding info in TIS.

Edit: Other way around according to TIS. So I guess we all need DOT4 LV. I wonder if Castrol SRF is considered LV. (despite what the Bentley says, 2003+ M3s have the MK60). Ideally I'd like to flush the system completely before putting in the SRF as it's quite expensive

Also I want to mention, I was always on OE fluid with my old ZHP. I did notice a difference with the CDV, but I will admit it wasn't night and day like a lot of folks claim. It was noticeable however -- and that makes sense. Just because it is "low" viscosity doesn't mean the effect will be eliminated. I also don't know if you can blame the issue entirely on brake fluid, since the CDV was installed in cars that don't specify DOT4 LV


http://i.imgur.com/SdyMY44.png



I do think you might be right about the effectiveness of the ABS system. I noticed last winter that my ABS didn't seem very effective. It felt like it was getting triggered way too early, and more importantly, it felt ineffective at actually allowing me to stop. I didn't make the connection with SuperBlue until now.

[quikryptonite]

Thanks William. I bought a bunch of DOT 4 LV for my brake job, so I have extra unopened liters for sale if anyone wants them. I'll sell for what I paid.

How many liters did you end up using to flush and fill?

[M0nk3y]

I've never noticed a difference in ABS ability at high speeds with super blue, high speeds being 130.

The only time abs was kicking in early was when I started to overheat pads

Sent from my XT907 using Tapatalk 2

[Crickett]

I've never noticed a difference in ABS ability at high speeds with super blue, high speeds being 130.

The only time abs was kicking in early was when I started to overheat pads

Sent from my XT907 using Tapatalk 2

Is the Z4MC's DSC the MK 60 type (no precharge pump) like us E46s or the MK 20 (w/precharge pump) like the M3s?

[wsmeyer]

Thanks for the added info, I didn't realize you had edited your post until today. This is quite interesting:


I do think you might be right about the effectiveness of the ABS system. I noticed last winter that my ABS didn't seem very effective. It felt like it was getting triggered way too early, and more importantly, it felt ineffective at actually allowing me to stop. I didn't make the connection with SuperBlue until now.

Because you observed this before I wrote about DOT4 LV possibly effecting the ABS system there isn't any way that you were influenced by it.


despite what the Bentley says, 2003+ M3s have the MK60

I know very little about the differences between E46's and E46 M3's but realoem lists a precharge pump for all years M3:

http://www.realoem.com/bmw/showparts.do?model=BL93&mospid=47714&btnr=34_1131&hg=34&fg=20

But only up to 09/2000 for non M3:

http://www.realoem.com/bmw/showparts.do?model=BW53&mospid=47707&btnr=34_0993&hg=34&fg=20

Also coincides with this document on E46 Traction Control systems:

http://images.chaleur.com/BMW%20Library/E46%20Traction%20Control.pdf

[wsmeyer]

...the CDV was installed in cars that don't specify DOT4 LV

Another light bulb just went off in my head. If you do a google image search for BMW CDV you'll notice that the CDV's don't all have the same size hole in them:

https://www.google.com/search?q=bmw+cdv&safe=off&source=lnms&tbm=isch&sa=X&ei=eLxFUvP3PKLMiQLkr4HoCw&ved=0CAcQ_AUoAQ&biw=1654&bih=769&dpr=1

And it is known that despite having the same part number, the CDV's on M3's look different than the ones on non M3's.

Check out figure 4 here:

http://zeckhausen.com/cdv.htm

The hole in mine is TINY, a #53 drill bit. I'll bet that the M3 CDV has a larger hole for use with DOT4.

[PirateZHP]

So... in conclusion... one should or shouldn't remove the CDV?

[danewilson77]

So... in conclusion... one should or shouldn't remove the CDV?

Remove it.

Sent from my Samsung Galaxy S4

[johnrando]

How many liters did you end up using to flush and fill?

Sorry, I don't know. Maybe 1 1/2?

[terraphantm]

Thanks for the added info, I didn't realize you had edited your post until today. This is quite interesting:



Because you observed this before I wrote about DOT4 LV possibly effecting the ABS system there isn't any way that you were influenced by it.



I know very little about the differences between E46's and E46 M3's but realoem lists a precharge pump for all years M3:

http://www.realoem.com/bmw/showparts.do?model=BL93&mospid=47714&btnr=34_1131&hg=34&fg=20

But only up to 09/2000 for non M3:

http://www.realoem.com/bmw/showparts.do?model=BW53&mospid=47707&btnr=34_0993&hg=34&fg=20

Also coincides with this document on E46 Traction Control systems:

http://images.chaleur.com/BMW%20Library/E46%20Traction%20Control.pdf

I think the ETK has that info wrong. The 2003+ M3s use the same system as the M3 CSL, and the CSL doesn't have a precharge pump listed. Wouldn't be the first time that there's a mistake in there

On a different note, the Z4M has the same CDV (same part number) as the E46 M3, and those cars specify DOT4 LV

[terraphantm]

I've never noticed a difference in ABS ability at high speeds with super blue, high speeds being 130.

The only time abs was kicking in early was when I started to overheat pads

Sent from my XT907 using Tapatalk 2

Do you drive it in snow/ice? Viscosity in cold weather will be more than at high temperatures. I suspect that cold viscosity is most important for ABS in snow/ice (especially if you haven't been driving all day)


Is the Z4MC's DSC the MK 60 type (no precharge pump) like us E46s or the MK 20 (w/precharge pump) like the M3s?

MK60e5. Don't know if they added the precharge pump back to those.



What kinda sucks is that there aren't really any low viscosity fluids with competitive dry/wet boiling points. The various DOT5.1 fluids are close, so that's probably the route I'll go.

[Yankeefan]

Reviving this super old thread, but can someone please post the link for where to order the LV fluid that is recommended here? Thanks!

[anandoc]

Reviving this super old thread, but can someone please post the link for where to order the LV fluid that is recommended here? Thanks!

Extremely old thread but somebody might benefit from this info. BMW now only carries LV fluid at their dealerships so you no longer need to pony up extra cash for Pentosin DOT4 LV etc. Here are the part numbers you are looking for (I could not find these part numbers on FCP Euro, ECS tuning etc):

83132405977 - 1L
83132405976 - 500ML

Price was $17.30 for the 1L bottle at Budd's BMW in Oakville, ON (Canada).

37181