legacycentral bbs

Hey peeps
Just read this on the Autospeed website http://www.autospeed.com/cms/A_0670/art ... larArticle and im wondering if this could be aplicable to our cars. For my tuppeneth worth I recon that just having the pressure releif valve on its own is a good idea to speed the turbo spooling up but I just dont know enough about turbos to see what implications this would have (ie would the FCD come in if you set the boost level too high with the reg and would the ECU actually know whats going on, would boost theoretically get out of hand?). Think Ill do the releif valve bit anyway though as I could do with less lag and can get both the bits from work anyway. Have a read and tell me what you think coz the car aint as fast as id expected and the boost deliverys way too slow for my liking.

That's a weird setup.

I don't quite understand it. They seem to indicate that with just the pressure regulator you can control boost. I don't understand how that is the case; doesn't the regulator just act more or less as a pressure limiting device?

The funny part is that most good manual boost controllers are exactly the same thing as a good inline relief valve. You set your valve to open at the desired pressure, and the valve blocks pressure from the wastegate until you reach desired boost, and then it allows the wastegate to open when the valve opens.

Their claim that the regulator controls boost and the relief valve controls creep seems dubious. If you set the regulator for a pressure lower than your wastegate's spring setting, you'll get unlimited boost. If you set it higher, then you'll get boost determined by the relief valve setting, it seems to me.

I could imagine that the regulator makes the relief valve more predictable, though. And it makes sure that you don't push too hard on the wastegate actuator diaphragm, though I've never heard of that being a problem.

I did quite well with just the relief valve. Look for a post I made earlier titled "Ghetto MBC" where I describe how I made one for like $4 in parts from my local hardware store. I've since revised my design and added a few more dollars to make it a little better. It's actually a lot like most commercial MBCs now, but cost about $6.

If the ECU sees the engine maintain a boost level of above about 13psi for more than about three seconds, it will cut fuel. So basically you can run up to just below 13 psi nominal boost without hitting the overboost cutoff. You of course have to make sure the whole fuel system is up to the task of fueling for 13psi, though.

Right, on this subject, been doin a bit of research and it would seam that the legacy has an extremely weak diaphram in the wastegate (ie not enough to keep it closed at max standard boost) and that the actual boost that this sees is sent from a pipe with a solenoid in controling the pulses continually, dependant on a few factors, meaning that the wastegate never actually sees a continuous boost level (possibly good for controlling creep?). Right, went a bit tangenty there, back to the point, if the wastegate is rather weak anyway will this not mean that when the turbo gets to the max pressure youve set the (*pause to allow brain to catch up) releif valve to and this pressure goes thru the regulator and into the wastegate will you not suddenly lose boost as the wastegate diverts it all back round to the turbo inlet or does this mean that you have just capped the available boost level, or (takes long breath) does it mean that the turbo can in theory just run away with itself leading to massive overboost and dead tubo, holed pistons and a general big bang.
As i would understand the system the releif valve is just there to stop any creep from the wastegate whatever level youve got the reg set to. The regulator is there to control the absolutely maximum safe level of boost that you want the wastegate to ever see.
The problem with my thoughts is that I dont know quite what effect the wastegate has on the system, ie if it opens does it dump all the overboost or does it still allow the turbo to keep producing more boost, I suspect the first 1.

I haven't really read this entire thread.....but to answer your question.

Simple answer to your question is that if the wastegate opens, boost goes through it and not the turbo, so you loose boost.

The wastegate has a simple spring tension system that allows it to boost to like 4 or 5 psi....someone help me out with the exact pressure. Anyway.....the spring keeps the wastegate open. The diaphragm works against the spring and closes the wastegate to build boost.

If the diaphragm is weak or leaks, you'll get creap.....or leak in this case.....

vrg3 wrote: You of course have to make sure the whole fuel system is up to the task of fueling for 13psi, though.

So are you saying the stock fuel system on the Legacy turbo wont support
13 psi? At what point should a 255 FP and injector upgrade be performed?
Or is the FP a must and injectors are further down the road or what?

DOA wrote:Right, on this subject, been doin a bit of research and it would seam that the legacy has an extremely weak diaphram in the wastegate (ie not enough to keep it closed at max standard boost)

Wastegates don't have diaphragms. They have doors with springs that hold them shut. If you're talking about the spring in the wastegate, though, yes, I think you're right. It does appear that it is a bit too easy to push the wastegate open just with exhaust gas flow. This may be why Cobb Tuning noticed that divorced wastegate downpipes hurt how quickly boost builds up, and may be why Subaru chose to use the downpipe flange design it did.

This is not the normal way a wastegate is supposed to open. Like Josh says, a wastegate is meant to open when it's pressed on by the wastegate actuator. The actuator has a diaphragm in it, and when there's enough pressure on the diaphragm, it pushes the wastegate open. The spring on our cars is set to open fully around 5 or 6 psi, depending on the particular car.

and that the actual boost that this sees is sent from a pipe with a solenoid in controling the pulses continually, dependant on a few factors, meaning that the wastegate never actually sees a continuous boost level (possibly good for controlling creep?).

The pressure seen by the wastegate actuator's diaphragm comes from the compressor outlet. It does go through the boost control solenoid; but, the solenoid doesn't exactly pulse the line. The air always goes from the compressor outlet, through the body of the solenoid, to the wastegate actuator. The solenoid, when activated, bleeds that pressure back into the intake tubing before the compressor.

The ECU controls this solenoid to choose how much pressure to bleed. Since the solenoid can only be fully open or fully closed, the ECU basically cycles it open and closed at a fixed frequency (something like 20 times a second) but varies the duty cycle. Since it cycles so fast it effectively allows the ECU precise control of how much air is bled. The wastegate actuator doesn't actually see pulses since they are damped out by the volume of air. It just sees less than the full compressor outlet pressure.

So the ECU has the capability to reduce the pressure seen by the wastegate. As long is it keeps that pressure low, the wastegate stays closed. When it stops bleeding pressure, the wastegate opens. A little intelligent programming gives us the approximately 8.7 psi that our cars produce stock.

if the wastegate is rather weak anyway will this not mean that when the turbo gets to the max pressure youve set the (*pause to allow brain to catch up) releif valve to and this pressure goes thru the regulator and into the wastegate will you not suddenly lose boost as the wastegate diverts it all back round to the turbo inlet or does this mean that you have just capped the available boost level, or (takes long breath) does it mean that the turbo can in theory just run away with itself leading to massive overboost and dead tubo, holed pistons and a general big bang.

I'm not sure I understand... The wastegate doesn't divert anything round to the turbo inlet.

But you don't "lose boost" when the wastegate opens. In actuality, the wastegate doesn't just slam open, making the turbine stop. If it did, boost would drop and cause the wastegate to immediately open again. What really happens is an equilibrium condition is reached where the wastegate is partially open, and the turbine gets just enough exhaust gas to maintain the proper boost level.

If you put a relief valve inline between the compressor outlet and the wastegate actuator, you actually get the same kind of equilibrium condition happening inside the relief valve too.

It does seem to me that the regulator caps the pressure that the rest of the pressure circuit can see. And yes, if you set the regulator below wastegate spring pressure, you would effectively prevent the wastegate from ever opening, and get essentially unlimited boost (from the wastegate side of things anyway).

As i would understand the system the releif valve is just there to stop any creep from the wastegate whatever level youve got the reg set to.

Yes, that's what the Autospeed article seems to say. But I claim it is also the element responsible for increasing boost level.

The regulator is there to control the absolutely maximum safe level of boost that you want the wastegate to ever see.

Yeah, that's one of the things I was guessing it might be... What do you mean by "safe" though? Do you mean "low enough to not damage the wastegate actuator?"

The problem with my thoughts is that I dont know quite what effect the wastegate has on the system, ie if it opens does it dump all the overboost or does it still allow the turbo to keep producing more boost, I suspect the first 1.

Well, first make sure you understand that the wastegate does not dump boost. It routes exhaust gas around the turbine instead of through it.

Like I said above, it doesn't usually open fully, and it often isn't completely closed either. As long as it's not fully open, the turbo continues to produce some boost. The more open it is, the less exhaust gas the turbine gets, and so the less boost the compressor produces.

legacy92ej22t wrote:So are you saying the stock fuel system on the Legacy turbo wont support 13 psi? At what point should a 255 FP and injector upgrade be performed?
Or is the FP a must and injectors are further down the road or what?

I was intentionally sidestepping that issue. =)

The general consensus seems to be that 12psi at the least is safe with the stock fuel system if it's in good shape. Some of the reasoning behind this is probably just that "well, if Subaru put fuel cut at 13psi, it's gotta be okay as long as you're below 13psi, right?" which is kind of dubious reasoning...

Boostjunkie does seem to be able to take it up to more than 16psi, but he has a different turbo which is more efficient at those higher pressures.

There are people on this board more qualified than I am to address this issue.

That was quite the post It does seem like a lot of the turbo guys are running 13 or a little higher on the stock fuel system though.

Hey, you don't get recognition like what the subyluvr2212 quote in my sig by being concise!

Yeah, there do seem to be plenty of people pushing really high boost on stock management. It is a little iffy though because if you defeat the fuel cut that means the ECU doesn't see all the boost, and also that the stock turbo compressor is pretty much pushing a superheated plasma instead of air at that point. (I'm exaggerating of course, but it's pretty far out of its efficiency island.)

Just to clear up a common misconception -- the ECU not seeing all the boost does not mean the ECU doesn't see all the air. It does see the airflow, and will still try to fuel accordingly. It won't run lean. Some people might just not feel it runs rich enough.

hey DOA, I notice you're from UK....this would mean you have an intercooled 2.0 yeah? the US car is a non-intercooled 2.2.........the Audi boost control system mentioned in the Autospeed article has been very successfully used on a Legacy/Liberty RS so you should have no problems.....just remember to cap the maximum boost at a safe level!

Jolly good, youve answered all my questions. Sorry about the diaphram confusion but I was thinking of the sytem as a whole whereby the actuator diaphram efectively sees a boost pressure and at this pressure (as set by the spring) it opens the wastegate, I just wast sure if it just went bang wide open or gradually varied the amount it opens, looks like the latter. So I would presume that when for instance you set the releif valve to 10 psi it will allow the boost level to stay at that value instead of suddenly dumping ALL the available boost (was thinking of dump valves lol).
The solenoid valve explanaition I put was more a case of mixed up terminology and me being mildly drunk at the time of writing .
Looking at it with all the info in my head Id sort of agree that the releif valve effectively caps the level of boost that you can get (without any wastegate creep leading to, wehay less lag) but the regulator could effectively smooth out any sudden spikes due to any time delays if you were to set it to a slightly higher level than the releif valve and also provide a sort of overboost protection. However from what you say about the wastegate never fully opening then if you were to set it higher than the releif valve then you would effectively get to a boost level (possibly higher than the releif valve is set to) when the wastegate is fully open and the compressor wont be able to pass anymore boost (needs experimentation methinks). oh and by safe I mean the maximum boost level your happy with the turbo providing (think id stay at around 10 PSI coz I want the car to last).
Cheers for your explanations as its cleared up all the bits I was worried about, I dont mind doing mods by myself but I do rather like to know the theory behind what Im trying to do lol and youve helped me along the way .
Saying all that I cant help thinking that it would be better to rechip the ECU to allow more boost (as I dont like the idea of the mixture and timing settings being wrong) with a releif valve in place and set to the ECU max boost to stop the creep occuring.

Oops, forgot 1 last thing. From what I understand about turbos, as you go through the rev range the boost will actually drop off towards the top end, would the releif valve stop this happening and keep the boost at the 10 theoretical PSI right to top revs instead of dropping off (assuming of course you bypass the control solenoid anyway).

DOA wrote:I was thinking of the sytem as a whole whereby the actuator diaphram efectively sees a boost pressure and at this pressure (as set by the spring) it opens the wastegate, I just wast sure if it just went bang wide open or gradually varied the amount it opens, looks like the latter.

Right. It's basically just that pressure on the diaphragm pushes the actuator rod. So it can definitely partially open it.

So I would presume that when for instance you set the releif valve to 10 psi it will allow the boost level to stay at that value instead of suddenly dumping ALL the available boost (was thinking of dump valves lol).

You mean you presumed that before, not anymore, right?

The solenoid valve explanaition I put was more a case of mixed up terminology and me being mildly drunk at the time of writing .

=)

Looking at it with all the info in my head Id sort of agree that the releif valve effectively caps the level of boost that you can get (without any wastegate creep leading to, wehay less lag)

Right, okay, I guess... I think of it more as the relief valve raises the level of boost that you can get. Ordinarily it's just the wastegate actuator that controls the level of boost. By blocking pressure below a threshold from the actuator, you get to raise the level. As a bonus, you avoid creep, since the wastegate doesn't partially open at low boost.

but the regulator could effectively smooth out any sudden spikes due to any time delays if you were to set it to a slightly higher level than the releif valve

Yeah, it could possibly. I guess it would depend on all the dynamics of the whole system.

and also provide a sort of overboost protection.

I don't understand how it could provide any overboost protection. The regulator limits the amount of boost seen by the wastegate through the relief valve, right? So the only thing it can do, as far as I can see, is prevent the wastegate from opening. That is the opposite of overboost protection.

However from what you say about the wastegate never fully opening then if you were to set it higher than the releif valve then you would effectively get to a boost level (possibly higher than the releif valve is set to) when the wastegate is fully open and the compressor wont be able to pass anymore boost (needs experimentation methinks).

I don't follow. As long as the regulator is set to a pressure above the wastegate actuator's pressure setting, the wastegate will have the potential to fully open, and so it will still reach an equilibrium state with boost being controlled by the actuator. And if you stick the relief valve in, boost will be controlled by that valve instead. Right?

I mean, the regulator can't have a higher output pressure than input pressure, right? At least not for any significant amount of time...

oh and by safe I mean the maximum boost level your happy with the turbo providing (think id stay at around 10 PSI coz I want the car to last).

Well, once you start doing upgrades to reduce underhood temperatures (like an intercooler or an exhaust) you can justify increasing boost, and it's possible to do this without drastically affecting reliability.

Cheers for your explanations as its cleared up all the bits I was worried about, I dont mind doing mods by myself but I do rather like to know the theory behind what Im trying to do lol and youve helped me along the way .

Thanks. Yeah, I think it's really important to understand why you're doing what you're doing too.

Saying all that I cant help thinking that it would be better to rechip the ECU to allow more boost (as I dont like the idea of the mixture and timing settings being wrong) with a releif valve in place and set to the ECU max boost to stop the creep occuring.

The ECU is actually meant to provide quick spoolup. It bleeds off most of the pressure from the actuator before full boost is reached. You shouldn't need a relief valve (also known as a ball-and-spring manual boost controller) if the ECU does its job right.

Note that the ECU's boost control and fuel/ignition control portions are semantically separate. It controls fueling and timing based on what sensors read, not based on its boost control output. So with respect to mixture and timing it doesn't matter whether boost is controlled by the ECU or by other means.

From what I understand about turbos, as you go through the rev range the boost will actually drop off towards the top end, would the releif valve stop this happening and keep the boost at the 10 theoretical PSI right to top revs instead of dropping off (assuming of course you bypass the control solenoid anyway).

It's not clear why this happens. One possible reason is just that the stock turbo is so darn small, in which case boost controllers won't help much. But I think the ECU does actually reduce boost a little as you approach redline, for reasons known only to the Subaru engineers who designed it. I have found that a ball-and-spring valve helped maintain full boost through the rev range, though it still dropped off towards redline.

SubbyRS, you say you've heard of the Autospeed design being used on a BC/BF... Can you shed any light on the confusion surrounding how it works?

This thread is a damn novel...LOL
Anyways, i hit 22psi the other day, and there was no knock and the A/F showed nothin but green, so I'm guessing that the fuel system is a lot stronger than most poeple think...

With the 2nd point, I was thinking that the turbo will still pass this amount off boost but will have the wastegate open by a certain percentage (like your saying it will find an equilibrium).

With the 4th point im thinking more that when I get round to trying it Ill do it on stock boost levels but your right (assuming the point above is right) that it should also raise your boost level.

Obviously thinking wrong with the overboost protection then lol.

Being a UK car Ive already got the intercooler and the exhaust isnt standard either, which is nice . Im gonna have a look at the heat exchanger for the intercooler though after reading a few other sites that say (and it makes sense) that the water in the intercooler acts more as a heatsink than to actively cool the air and then passes part of this heat back into the airstream at a later time, so I think some thought needs to be put into making the pump run fast all the time and maybe a fan if there isnt one.

OOhtage regarding the ECU fuel control, that cheers me up a bit as they cost in the region of 500 pounds over here.

I can sort of understand why the boost drops at very high revs as the turbos probably past its eficient working range and cant actually shove any more air into the engine, but its nice to know that a releif valve will stabilise this for longer.

I think Ill try the system with and without the regulator to see what difference it makes, just gotta sort a knock sensor out first and then nick the bits from work . Just btw, the knock sensors the small black thing atached to the right hand (as stood looking at the front of the car) cylinder bank at the front right?

Thanks again for the answers, Your help is definately apreciated.

91White-T wrote:This thread is a damn novel...LOL

Heheh... Well, it doesn't help that I'm making such liberal use of the quoting feature.

Anyways, i hit 22psi the other day, and there was no knock and the A/F showed nothin but green, so I'm guessing that the fuel system is a lot stronger than most poeple think...

22psi?! I find that incredible. Did you just momentarily peak at 22? Because if you did, the oxygen sensor may have missed the momentary lean condition....

Im gonna have a look at the heat exchanger for the intercooler though after reading a few other sites that say (and it makes sense) that the water in the intercooler acts more as a heatsink than to actively cool the air and then passes part of this heat back into the airstream at a later time, so I think some thought needs to be put into making the pump run fast all the time and maybe a fan if there isnt one.

Yes. All intercooled street cars have their intake charge cooled by air. With an air/water intercooler you just use water to transfer the heat to a place where it is more easily absorbed by passing air. It's still relying on the air to cool the charge.

I think Ill try the system with and without the regulator to see what difference it makes

Cool, I am interested in hearing the results.

Just btw, the knock sensors the small black thing atached to the right hand (as stood looking at the front of the car) cylinder bank at the front right?

Yes, it is.

Thanks again for the answers, Your help is definately apreciated.

Thanks for discussing this stuff and pointing the rest of us to that article.

Just btw, the knock sensors the small black thing atached to the right hand (as stood looking at the front of the car) cylinder bank at the front right?

Yes, it is.

Wouldn't the knock sensor be on the rear side, about in the middle of the right hand side of the engine...

You can find it by looking directly down ontop of the engine just a bit in front and to the right of the throttle body.

MK

Wait, yeah, you're right, MK. I got confused by the backwards directions.

DOA, in general, the left side of the car is the passenger side in the UK and the driver side in the US. The right of the car is the driver side in the UK and the passenger side in the US.

And like MK says, the sensor is below the rear left intake runner.

I guess maybe you were looking at the camshaft angle sensor?

Found it after taking the cam angle sensor out and thinkin hang on, I recon thats the cam angle sensor lol. Anyway, tinkered with the knock sensor, gave it a good clean and heres hoping its just a wiring fault as their £90 (over $100).

DOA wrote:Anyway, tinkered with the knock sensor, gave it a good clean and heres hoping its just a wiring fault as their £90 (over $100).

It's not the same part number as on US Legacies? Here you can get them for more like $30-$40.

Here there was a documented problem with early sensors so there is a redesigned sensor which is recommended.

Its probably the same part as the dealership told me theres a possible update (probably the one your saying) and thats still £80. Your in some ways lucky in coming from the states as over here they shaft you on all prices for the simple reason that we pay them weras you lot probably wont. Saying all that though having taken it off, cleaned it and the conector up its started working again, think it had something to do with it having a bad earth to the block with it being only a single wire sensor and judging by the crud I had to scrape off its metal face and the block. Im now back to a nice free boosting car, no hesitations .

Cool, maybe it was just cruddy connections and stuff. I think one of the common problems on the original sensor design had to do with small cracks developing in the ceramic allowing it to vibrate and set off the knock detector even when there was no knock.

If you can find out the part number from your dealer for the updated sensor, and it's 22060AA031, then maybe you could get one of us in the colonies to source you a sensor cheaper. =)

The cracks happened to the sensor on my new engine, i swapped over my old sensor and the cel went away.

MK

Might have a look into that, I know a lad who works at the nearest main dealer so should be able to get the number (wont be going there in person as its a 70 mile round trip lol). Any chance of getting me some plugs as well, theyre about £5 a piece over here (only jokin lol but they are £5! ).

5 pounds?! What kind of plugs are they? That kind of price gets platinum plugs here normally.

I do beleive iridium and platinum ones are about £16 a piece and the fiver ones are just v grooved ngks, last set i stuck in my bike were £7 a piece but theyre a slightly odd size and type. As I say we get regally shafted over here, bet your cars only cost IRO $15k wheras mine cost £20kish brand new.