Stock BPV Leak? I may have a cost effective semi-solution...

[TRex]

This was posted on m3f....thread is here Stock BPV Leak? I may have a cost effective semi-solution...

THIS IS NOT MY WORK....it was done by Phate from the other forum...im just moving it here because it i s good info

.................................................. .................................................. ........................
Sooo, recently I have been having some issues with an intermittent fluttering of power when I get into boost, especially after I shift (quick shifts, in particular). I have read time and time again how the stock bypass valves are known to leak, and I suspected it anyway, just because of how the power was affected. Rather than simply ordering a new valve, I did some investigating because I like to know what is really wrong with something before I replace it (and I'm cheap ). I found some very interesting things about the stock valve when I went and did pressure testing.

Let's review first: In layman's terms, the BPV/BOV is there to rid excess pressure from the intake system. It is there so that when you let off of the gas, an excess amount of pressure doesn't build, because after you let off of the gas pedal, your turbo is still spinning pretty damn fast. This excess pressure, if not vented, can cause the turbocharger to run backwards. Yes, that is a bad thing. Think 100k rpm to backwards in a split second. I've said it before, and I'll say it again, flying chunks of exploded turbo are a bad thing...

Let's review how a normal bypass/blowoff valve works. Here's a pic:



Looking at the picture (thanks GReddy), we can see that the mechanism is quite simple. The large valve at the bottom (#3 in the pic), needs to open when we want to vent pressure. Simply pushing against a spring wouldn't do, because it wouldn't react quickly enough and may not allow the valve to open a sufficient amount to vent enough air. So the spring needs a little bit of help: in comes manifold pressure.

Some bright mind knew that if we could help the spring stay closed with a little more pressure when under boost, and help the valve open when we close the throttle plate, we would have a happy medium. If we look at manifold pressure, that is to say after the throttle plate, we can easily see where the assistance comes in. When under WOT, the pressure in the manifold is the same as everywhere else after the turbo (yeah, yeah, not really; but for this it doesn't matter). When the throttle is closed, it forces the cylinders to suck air past a tiny slit created between the throttle plate and throttle bore. This huge restriction causes a partial vacuum. So why not seal that chamber where the spring is holding down the valving mechanism and utilize these two pressure scenarios? That is exactly what we do! We simply hook up (reference) the vacuum nipple (#1 in the pic) to a manifold vacuum fitting, and let the spring and pressure differential work together as a team. So under boost, we have spring pressure + boost pressure holding the valve closed. On the other side of the valve is only boost pressure, so we have a winning combination there. When we suddenly close the throttle plate, our Mazda's create somewhere around 10lbs of negative pressure (vacuum). So spring pressure + (-boost) gives us with spring pressure - 10lbs or so. If we utilize the correct spring in the happy medium, then we have a perfect setup. [This is why running the proper spring or having the spring tension set correctly is so important. If too tight, you won't vent pressure. If too loose, it could open too early and not close very quickly afterwards.]

So I know that most of us know how a BPV works, but it never hurts to reiterate...Especially after what I've learned from the OE valve in question.

Mazda, in all of their wisdom and money spending for creating the 2.3L turbocharged DISI, decided to reinvent the wheel (that wheel of course being the bypass valve). After repeatedly having the issue of fluttering boost, I suspected that the bypass valve was not closing correctly for some reason. After thinking about it, I figured that the BPV was not closing properly after shifting and venting air; and it would also open on occasion after staying in boost for a second or two. Why this could be: Looking at the workings of the valve above, I thought that it was either an extremely weak spring, or the boost pressure was not getting to the top of the spring due to a leak or restriction. I took the valve off, played with it, pressure tested it, and found some new and interesting things about the valve.

My first test was to check spring tension with the precise method of pushing on it. I found that it has quite a bit of tension (no idea how much, don't care because it was quite a bit) so I dismissed the notion that it might be a weak spring. That thing was surprisingly strong, I was impressed.

My second test was to pressure test the chamber which the spring resides in. I set up my compressor to put out only ~6.5lbs of pressure and I used my rubber cone tip to seal the against the BPV's vacuum nipple. As soon as I put pressure to the BPV, it leaked. The flange that connects the top half of the valve to the bottom half of the valve leaks pretty much all the way around. Not a huge amount, but definitely a leak that could cause a problem. Issue identified! BUT! I thought that the spring tension was quite high (higher than I expected anyway), so leaking a little bit wouldn’t cause the huge problems, would it? I suspect that the chamber is only losing a pound or two of pressure, at most. Which, even at 5lbs lost, I would have spring tension +10lbs of boost, which should be more than enough to close the valve and overcoming the boost pressure of 15lbs! So my investigation continued.

Here’s the interesting part…As I tried to figure out how to take this thing apart (more about that later), I found a little hole on the bottom of the flange next to the valve which vents manifold pressure (#3 in the pic below depicts this)…I looked at it, looked in it, and didn’t think it went anywhere. But I was frustrated so I went ahead and pressure tested it. Lo and behold, the valve opened. The valve opened and there was not a leak anywhere. After consideration, I postulated that there must be a second chamber, below the spring chamber. Yes, a bi-chamber bypass valve. I searched google for quite a while and couldn't find anything that looked like this or any articles that referenced a valve that works like this...I presume that it looks something like this, but I don't have a valve to open [break] so that I can look at the actual construction.



So what does this mean for us? Well if you consider the different pressure scenarios that we looked at earlier, you will see that it is a very impressive design. Under boost, the pressure at this orifice would be the same as the pressure in the intake manifold, making the pressure equal in both chambers of the valve. The net effect being zero, allowing only spring pressure to keep the valve closed against the pressure in the manifold pushing against the actual valve face (#2 in this pic). As soon as we close the throttle, we have vacuum at the top of the valve pulling it open, AND we have pressure on the bottom of the valve (because this is just attached to the intake manifold before the throttle body) and the pressure pushing against the valve face trying to escape!. This makes the valve react lightning fast and when there is only a small pressure differential. Well, we already know that the top chamber leaks! Now we have created a pressure differential at all times under boost!

Now we have (spring pressure) + (top chamber boost pressure) - (bottom boost pressure chamber) - (valve face pressure). Which, if nothing leaked, would leave us with only spring pressure against the actual boost pressure on the face of the valve. When we are leaking from the top though, the bottom chamber pressure will be higher than the top chamber, which then allows the pressure differential to push the valve open! I believe this fits perfectly with what I, and others, have experienced!

So how do we fix this? Well, I don't have a complete solution (yet). Simply sealing the flange where the top and bottom connect would be the simplest solution, but I don't think the valve can be opened (I didn't want to break mine). I considered removing the top chambers vacuum line, but then we wouldn't be helping the valve stay closed and the bottom chamber would just be pushing it open, anyway (actually, I did try this and it definitely does not work). We need the pressure in the top chamber to help keep the valve closed.

SO, as a temporary, semi-solution, I decided to try and block the bottom chamber's boost reference. It WORKS. You are essentially turning this into a standard BPV when you remove the second chamber's boost reference. I used the top of a vacuum hose nipple cover cut long enough to be compressed into the hole when you bolt the BPV back to the manifold. See my caveat below, but I will be experimenting with different sized holes in the stopper to allow some airflow to the second chamber. After doing this, I went out and did some datalogs. Before, I was only making about 14lbs of boost on the top end near 6k rpms and my wastegate duty cycle was ridiculously high for most of the power band (like 80%+ from 4k-6.5k rpm). This told me that the turbo was doing all that it could to produce pressure, but it still wasn't building enough. It only makes sense that if the BPV is leaking (and thus, reducing pressure), that the turbo would have to work extra hard. After plugging the bottom chamber's orifice, I am now making about 20lbs of boost at 4k rpm and it tapers to about 16.5lbs on the very top end. All of this with duty cycles at their highest of 60% at 4krpm, and declining to about 40% at 6krpm. Boost went up a little bit, and wastegate duty cycle went down, by only taking the second chamber out of the equation. The power doesn't flutter and the car feels SOLID.

CAVEAT: As a side effect, though, this does cause the turbo to flutter a little bit (yes, like the supra’s…). I must warn, though, so you don’t just run out and do this, that this IS potentially harmful for your turbo and that it may reduce the overall lifespan of the turbo. It is the turbo cavitating, and the pressure waves bouncing between the turbo and throttle plate. I’ve seen and heard of many cars that run like this for years. Numerous high power drag cars (think outlaw 10.5) and many street cars run bov’s with the spring tension just a little too high, which is what causes this. Many of these cars will never have a problem. [Actually, I don’t really know of anyone that has had an issue (i.e. breaking or bearing failure) when their turbo’s flutter a little bit.] But, I think that you should fully understand what is going on. Do this at your own risk. I am

Now, we just need to figure out how to keep the top flange from leaking.

-Phate
.................................................. .................................................. ........................

discuss

[darth vader]

Funny, I figured this out myself shortly after getting this car. I didn't find it actually did much in my case, as the issue for me wasn't the holding of pressure, but an annoying light throttle sogginess in throttle response. Forge + Red fixed that up, sicko-style.

[TRex]

yea im not sure how it will affect everyone...like he said he is still playing with it

i personally havent experience the leaking valve yet (11k miles)...but i may try this if im experiencing the same issue

[usmc_adams]

stupid question how do you kn ow if it is leaking? i disconnected the hose to recirulate back to airbox and as i drove around i did a WOT and i heard or what seemed like air rushing out like someone blowing air threw their mouth so what could i do to be a cheapo and fix this

[TRex]

well he pressure tested his....most of the time you can hear it...another way is to look at your boost gauge and see if you are holding boost

[usmc_adams]

Originally Posted by TRex well he pressure tested his....most of the time you can hear it...another way is to look at your boost gauge and see if you are holding boost

i would if i can find the boost gauge i want and beable to place where i want lol

[TRex]

^just get a dashhawk...best gauge around imo

[prenedo]

I wonder what this guys mods are to be holding 20, tapering down to 16.5 at the end of the powerband.

[usmc_adams]

what excatly does a dashhawk do for you/car?????

[TRex]

It is an electronic gauge that you plug into you OBDII port that basically lets you read every single sensor on your car....lets you know everything the car knows

DashHawk Vehicle Information Display

installed (it is the orange thing on the left)

[darth vader]

Plus it's got a shift light, staged alerts for anything, 0-60 and 1/4 mile timer, etc. etc. etc. Best 200-ish bucks you can spend on this car, bar none and, you can take it with you to your next car.

[Fagwagon]

i didnt read the OP because i really dont care, but why wouldnt you just get you a new bov quit being cheap asses.

[TRex]

^because problems can arise if you dont set up the aftermarket BOV/BPV correctly...besides i feel most would rather take that money and apply it to other more performance enhancing mods











on another note your avy mad me jizz my pants

[Chinsel]

epoxy that top flange!

[TRex]

that was considered in the other thread...so far it hasnt been figured out how to take apart the valve without damaging it....and if you just epoxy around the edges...when you go into boost it will probably just blow it out anyway

[ScarMS3]

Sub

[trdsw]

the previous post was over 10 months ago. You might get a better result if you pm phate directly on m3f. The best result is to get an aftermarket BPV.

[phate]

I'm here, and still running the tweaked bpv. I've put about 15k miles on the car since I did this. I never did get a second stock BPV to test, and I don't run the plug with a hole in it (still completely sealed). No problems and the car runs amazingly well. The stock turbo is completely maxed for airflow. I see no reason to get an aftermarket piece, unless you want to spend $200+ on something shiny for your engine bay.

Edit: I did update the post on M3F to give some more detail about the workings of the stocker BPV. I also posted a couple of pics of the piece i used to plug the hole.

[ScarMS3]

Oh iwasnt subbing for results I just do it to remember where it is, but thanks

[himurax13]

Subbed for later.

[GrumpyMPS]

@phate did you ever picked up any issues with doing this?

[phate]

Nope. I even run this on my MS6. Probably had 60k miles like this on my MS3, and I have 30k miles on my MS6 like this.

[GrumpyMPS]

Thanks mate, I have done the mod and it seems to be holding 21psi real good. Awesome find, well done!

[cliffordrayfagan]

Originally Posted by TRex This was posted on m3f....thread is here Stock BPV Leak? I may have a cost effective semi-solution... THIS IS NOT MY WORK....it was done by Phate from the other forum...im just moving it here because it i s good info .................................................. .................................................. ........................ Sooo, recently I have been having some issues with an intermittent fluttering of power when I get into boost, especially after I shift (quick shifts, in particular). I have read time and time again how the stock bypass valves are known to leak, and I suspected it anyway, just because of how the power was affected. Rather than simply ordering a new valve, I did some investigating because I like to know what is really wrong with something before I replace it (and I'm cheap ). I found some very interesting things about the stock valve when I went and did pressure testing. Let's review first: In layman's terms, the BPV/BOV is there to rid excess pressure from the intake system. It is there so that when you let off of the gas, an excess amount of pressure doesn't build, because after you let off of the gas pedal, your turbo is still spinning pretty damn fast. This excess pressure, if not vented, can cause the turbocharger to run backwards. Yes, that is a bad thing. Think 100k rpm to backwards in a split second. I've said it before, and I'll say it again, flying chunks of exploded turbo are a bad thing... Let's review how a normal bypass/blowoff valve works. Here's a pic: Looking at the picture (thanks GReddy), we can see that the mechanism is quite simple. The large valve at the bottom (#3 in the pic), needs to open when we want to vent pressure. Simply pushing against a spring wouldn't do, because it wouldn't react quickly enough and may not allow the valve to open a sufficient amount to vent enough air. So the spring needs a little bit of help: in comes manifold pressure. Some bright mind knew that if we could help the spring stay closed with a little more pressure when under boost, and help the valve open when we close the throttle plate, we would have a happy medium. If we look at manifold pressure, that is to say after the throttle plate, we can easily see where the assistance comes in. When under WOT, the pressure in the manifold is the same as everywhere else after the turbo (yeah, yeah, not really; but for this it doesn't matter). When the throttle is closed, it forces the cylinders to suck air past a tiny slit created between the throttle plate and throttle bore. This huge restriction causes a partial vacuum. So why not seal that chamber where the spring is holding down the valving mechanism and utilize these two pressure scenarios? That is exactly what we do! We simply hook up (reference) the vacuum nipple (#1 in the pic) to a manifold vacuum fitting, and let the spring and pressure differential work together as a team. So under boost, we have spring pressure + boost pressure holding the valve closed. On the other side of the valve is only boost pressure, so we have a winning combination there. When we suddenly close the throttle plate, our Mazda's create somewhere around 10lbs of negative pressure (vacuum). So spring pressure + (-boost) gives us with spring pressure - 10lbs or so. If we utilize the correct spring in the happy medium, then we have a perfect setup. [This is why running the proper spring or having the spring tension set correctly is so important. If too tight, you won't vent pressure. If too loose, it could open too early and not close very quickly afterwards.] So I know that most of us know how a BPV works, but it never hurts to reiterate...Especially after what I've learned from the OE valve in question. Mazda, in all of their wisdom and money spending for creating the 2.3L turbocharged DISI, decided to reinvent the wheel (that wheel of course being the bypass valve). After repeatedly having the issue of fluttering boost, I suspected that the bypass valve was not closing correctly for some reason. After thinking about it, I figured that the BPV was not closing properly after shifting and venting air; and it would also open on occasion after staying in boost for a second or two. Why this could be: Looking at the workings of the valve above, I thought that it was either an extremely weak spring, or the boost pressure was not getting to the top of the spring due to a leak or restriction. I took the valve off, played with it, pressure tested it, and found some new and interesting things about the valve. My first test was to check spring tension with the precise method of pushing on it. I found that it has quite a bit of tension (no idea how much, don't care because it was quite a bit) so I dismissed the notion that it might be a weak spring. That thing was surprisingly strong, I was impressed. My second test was to pressure test the chamber which the spring resides in. I set up my compressor to put out only ~6.5lbs of pressure and I used my rubber cone tip to seal the against the BPV's vacuum nipple. As soon as I put pressure to the BPV, it leaked. The flange that connects the top half of the valve to the bottom half of the valve leaks pretty much all the way around. Not a huge amount, but definitely a leak that could cause a problem. Issue identified! BUT! I thought that the spring tension was quite high (higher than I expected anyway), so leaking a little bit wouldn’t cause the huge problems, would it? I suspect that the chamber is only losing a pound or two of pressure, at most. Which, even at 5lbs lost, I would have spring tension +10lbs of boost, which should be more than enough to close the valve and overcoming the boost pressure of 15lbs! So my investigation continued. Here’s the interesting part…As I tried to figure out how to take this thing apart (more about that later), I found a little hole on the bottom of the flange next to the valve which vents manifold pressure (#3 in the pic below depicts this)…I looked at it, looked in it, and didn’t think it went anywhere. But I was frustrated so I went ahead and pressure tested it. Lo and behold, the valve opened. The valve opened and there was not a leak anywhere. After consideration, I postulated that there must be a second chamber, below the spring chamber. Yes, a bi-chamber bypass valve. I searched google for quite a while and couldn't find anything that looked like this or any articles that referenced a valve that works like this...I presume that it looks something like this, but I don't have a valve to open [break] so that I can look at the actual construction. So what does this mean for us? Well if you consider the different pressure scenarios that we looked at earlier, you will see that it is a very impressive design. Under boost, the pressure at this orifice would be the same as the pressure in the intake manifold, making the pressure equal in both chambers of the valve. The net effect being zero, allowing only spring pressure to keep the valve closed against the pressure in the manifold pushing against the actual valve face (#2 in this pic). As soon as we close the throttle, we have vacuum at the top of the valve pulling it open, AND we have pressure on the bottom of the valve (because this is just attached to the intake manifold before the throttle body) and the pressure pushing against the valve face trying to escape!. This makes the valve react lightning fast and when there is only a small pressure differential. Well, we already know that the top chamber leaks! Now we have created a pressure differential at all times under boost! Now we have (spring pressure) + (top chamber boost pressure) - (bottom boost pressure chamber) - (valve face pressure). Which, if nothing leaked, would leave us with only spring pressure against the actual boost pressure on the face of the valve. When we are leaking from the top though, the bottom chamber pressure will be higher than the top chamber, which then allows the pressure differential to push the valve open! I believe this fits perfectly with what I, and others, have experienced! So how do we fix this? Well, I don't have a complete solution (yet). Simply sealing the flange where the top and bottom connect would be the simplest solution, but I don't think the valve can be opened (I didn't want to break mine). I considered removing the top chambers vacuum line, but then we wouldn't be helping the valve stay closed and the bottom chamber would just be pushing it open, anyway (actually, I did try this and it definitely does not work). We need the pressure in the top chamber to help keep the valve closed. SO, as a temporary, semi-solution, I decided to try and block the bottom chamber's boost reference. It WORKS. You are essentially turning this into a standard BPV when you remove the second chamber's boost reference. I used the top of a vacuum hose nipple cover cut long enough to be compressed into the hole when you bolt the BPV back to the manifold. See my caveat below, but I will be experimenting with different sized holes in the stopper to allow some airflow to the second chamber. After doing this, I went out and did some datalogs. Before, I was only making about 14lbs of boost on the top end near 6k rpms and my wastegate duty cycle was ridiculously high for most of the power band (like 80%+ from 4k-6.5k rpm). This told me that the turbo was doing all that it could to produce pressure, but it still wasn't building enough. It only makes sense that if the BPV is leaking (and thus, reducing pressure), that the turbo would have to work extra hard. After plugging the bottom chamber's orifice, I am now making about 20lbs of boost at 4k rpm and it tapers to about 16.5lbs on the very top end. All of this with duty cycles at their highest of 60% at 4krpm, and declining to about 40% at 6krpm. Boost went up a little bit, and wastegate duty cycle went down, by only taking the second chamber out of the equation. The power doesn't flutter and the car feels SOLID. CAVEAT: As a side effect, though, this does cause the turbo to flutter a little bit (yes, like the supra’s…). I must warn, though, so you don’t just run out and do this, that this IS potentially harmful for your turbo and that it may reduce the overall lifespan of the turbo. It is the turbo cavitating, and the pressure waves bouncing between the turbo and throttle plate. I’ve seen and heard of many cars that run like this for years. Numerous high power drag cars (think outlaw 10.5) and many street cars run bov’s with the spring tension just a little too high, which is what causes this. Many of these cars will never have a problem. [Actually, I don’t really know of anyone that has had an issue (i.e. breaking or bearing failure) when their turbo’s flutter a little bit.] But, I think that you should fully understand what is going on. Do this at your own risk. I am Now, we just need to figure out how to keep the top flange from leaking. -Phate .................................................. .................................................. ........................ discuss

I am not a turbo expert, but I've been working around engines and engineering unusual solutions for decades. This is exactly the question I was coming here to explore, but didn't see your thread at first as I was looking for threads with search key "wastegate" as oppose to BPV

It seems to me that one option that would not endanger the turbo and is low cost might be a mechanical fix, like you attempted, but to the flange. What do you think about simply coating the seam where the two pieces of metal are pressed together with ultra high temp gasket sealant and let it dry before testing?

It shouldn't be possible for that to harm anything, but I'm uncertain if it would work or not. It may not stop the leak, but rather reduce it ... a lot. that may be all you need.

I have something similar going one with a leaky wastegate on the 2003.5 speed protege causing my check engine light to prevent passing inspection. It was so hard to find the problem, I am jsut now working on a solution.

Naturally the dealer want $900 to totally replace the wastegate. That seems silly to me. They also seem to believe it can't be removed without disassembling the turbo. That I'm not certain about, but it doesn't sound like you had that problem. It would be difficult to seal the entire sem without removing the wastegate completely, but I might be able to seal what is needed, and you seem to be able to remove your BPV, so do you think that type of simple, but direct, approach is worth a try?

[btstarcher]

Time for another revival! Hey @phate; will the stocker hold 24psi ya think? My Forge is leaking around the piston, and I found my old OEM valve in the garage.

[phate]

I still run the stocker at whatever boost pressure the k04 can muster

[btstarcher]

Excellent! Then I can sell the Forge!

[Easter Bunny]

Originally Posted by btstarcher Time for another revival! Hey @phate; will the stocker hold 24psi ya think? My Forge is leaking around the piston, and I found my old OEM valve in the garage.

Only one way to find out.....

[MSMS3]

I can't say if the stock valve will hold 24, but mine holds 19 just fine and has 113,000 continuous use miles on it.

[speedfreak44]

I thought mine was holding boost fine but when I swapped in my cs bpv the car was noticeably faster when I would shift as in the power came on much quicker

[aackthpt]

Originally Posted by speedfreak44 I thought mine was holding boost fine but when I swapped in my cs bpv the car was noticeably faster when I would shift as in the power came on much quicker

Probably has nothing to do with holding boost statically (though it might to some extent). But for the most part it just was releasing earlier/quicker, slowing your turbot down less during the shift therefore allowing boost to build faster after the shift.

[speedfreak44]

this is true. My actual boost never really changed

[sheston]

My stocker will hold 28+ psi

I did this mod long ago, not sure it had any impact

[MSMS3]

Originally Posted by speedfreak44 I thought mine was holding boost fine but when I swapped in my cs bpv the car was noticeably faster when I would shift as in the power came on much quicker

Do you have any before and after datalogs?

Years ago, Grassroots Magazine compared the stock BPV on an MS3 to the Forge,and reported that while the properly adjusted Forge released faster, the stock BPV reached higher boost after the shift.

I stayed with the stock BPV for that reason, because it didn't leak, and was and is so smooth in operation with my mods.

[speedfreak44]

I saw no difference in peak psi oem vs corksport but I felt like the boost built much fast with cs vs oem. When I would get back on the throttle that car felt much more responsive. As far as logs are concerned I wouldn't even be able to tell you which ones were with oem or with cs. Like said previously they were holding the same boost pressure although I only target 19 psi

[btstarcher]

What about WGDC? Did it go down?

[speedfreak44]

Within 2 percentage of before. Probably more to do with weather/road. I don't know how stratified sets their tunes but I am always at near 100%.

[btstarcher]

Damn that turbo is working hard! What PSI? I'm around 90-100% at 24PSI. In the winter it was around 80 I think.

[speedfreak44]

I don't think that means the turbo is being worked hard it just means the wastegate is 100% open dumping exhaust into the downpipe and bypassing the turbo. I am running 19psi with spikes of 20-20.5ish.

[phate]

Originally Posted by speedfreak44 I don't think that means the turbo is being worked hard it just means the wastegate is 100% open dumping exhaust into the downpipe and bypassing the turbo. I am running 19psi with spikes of 20-20.5ish.

Other way around. More wgdc, more boost.

But 100% just means the solenoid is fully open, but not necessarily pushing the turbo any specific amount. That's why 3 port setups have to be retuned, or if you run a restrictor in the boost reference to the solenoid.