Extracting the most from your setup... Fluid Dynamics, thermal insulation, etc

[kenamond]

If the bleeder hole size in a normal ball/spring MBC matters for fine-tuning boost control, couldn't you thread a nipple into the MBC body and run that to a bleeder MBC to fine tune the ball/spring bleed? Then you could vary the spring, MBC screw and bleeder settings, right? And you could send the bleeder output back to the intake pipe if you're draw-through.

 

[dsm-onster]

If the bleeder hole size in a normal ball/spring MBC matters for fine-tuning boost control, couldn't you thread a nipple into the MBC body and run that to a bleeder MBC to fine tune the ball/spring bleed? Then you could vary the spring, MBC screw and bleeder settings, right? And you could send the bleeder output back to the intake pipe if you're draw-through.

That's an excellent idea!!! Someone with spike issues should try this! I suspect the bleeder section of th ball-and-spring responds to VERY fine changes. So a valve that changes flow in very small incriments would be neccesary. Like a bleeder valve used for boost. . . Where's Matt (Mr Peepers) . He can drum up an experiment...

Black_Bullet, if you're not spiking, I'd try shimming the wastegate actuator. Or get a Peepers wastegate actuator so you can adjust the arm length to see if you get quicker spool. But, you already have a decent ball-spring control from what I see. It depends on the hole diameter and spring rates available. You'll likely see no difference unless your actuator spring is weak.

 

[kenamond]

That's an excellent idea!!! Someone with spike issues should try this! I suspect the bleeder section of th ball-and-spring responds to VERY fine changes. So a valve that changes flow in very small incriments would be neccesary. Like a bleeder valve used for boost. . . Where's Matt (Mr Peepers) . He can drum up an experiment...

Black_Bullet, if you're not spiking, I'd try shimming the wastegate actuator. Or get a Peepers wastegate actuator so you can adjust the arm length.

He's probably at UPS shipping a turbo to me so I can try my first rebuild

 

[Mr Peepers]

^ Why yes, yes I was I'll be sending you a PM shortly.

That's an excellent idea!!! Someone with spike issues should try this! I suspect the bleeder section of th ball-and-spring responds to VERY fine changes. So a valve that changes flow in very small incriments would be neccesary. Like a bleeder valve used for boost. . . Where's Matt (Mr Peepers) . He can drum up an experiment...

I have some results from a ways back from a thread I'm still looking for. Here is one similar: http://www.dsmtuners.com/forums/newbie-forum/266779-boost-spiking-ok.html#post151243706 -this was after I had a guinea pig in another thread It helped but didn't completely cure it, a step in the right direction.

BTW, the early Joe P.(the one that looked like it was from the plumbing department) is a ball and spring MBC and it is known to spike from its small bleeder hole- 1/16". This was the MBC that was tested.

I will ABSOLUTELY be running a MBC with an adjustable bleeder hole. The quickest spool up can be obtained just before it starts to spike.

EDIT: found the original thread, http://www.dsmtuners.com/forums/newbie-forum/262063-boost-spike-fade-leak.html

Cut the spike down by 4-5 psi just by going one step larger

 

[Black_Bullet]

That's an excellent idea!!! Someone with spike issues should try this! I suspect the bleeder section of th ball-and-spring responds to VERY fine changes. So a valve that changes flow in very small incriments would be neccesary. Like a bleeder valve used for boost. . . Where's Matt (Mr Peepers) . He can drum up an experiment...

Black_Bullet, if you're not spiking, I'd try shimming the wastegate actuator. Or get a Peepers wastegate actuator so you can adjust the arm length to see if you get quicker spool. But, you already have a decent ball-spring control from what I see. It depends on the hole diameter and spring rates available. You'll likely see no difference unless your actuator spring is weak.

I see about a 2-3 psi spike in 4th and 5th gear. 3rd will get that 2-3 psi spike also
if its cold out.

It makes it really hard to tune on the edge with the car spiking inconsistently.
Thats why external gate was important to me.

I want 24psi thru all gears but no more than that.

If i set it to were 2nd and 3rd gear are right were i want it, then 4th and 5th are boosting too much and pulling timing. If i drop the boost down to were 4th and 5th are were i want it, then 1st, 2nd, and 3rd are too weak in my opinion...

^ Why yes, yes I was I'll be sending you a PM shortly.



I have some results from a ways back from a thread I'm still looking for. Here is one similar: http://www.dsmtuners.com/forums/newbie-forum/266779-boost-spiking-ok.html#post151243706 -this was after I had a guinea pig in another thread It helped but didn't completely cure it, a step in the right direction.

BTW, the early Joe P.(the one that looked like it was from the plumbing department) is a ball and spring MBC and it is known to spike from its small bleeder hole- 1/16". This was the MBC that was tested.

I will ABSOLUTELY be running a MBC with an adjustable bleeder hole. The quickest spool up can be obtained just before it starts to spike.

EDIT: found the original thread, http://www.dsmtuners.com/forums/newbie-forum/262063-boost-spike-fade-leak.html

Cut the spike down by 4-5 psi just by going one step larger

Im going to check that bleeder hole out but its definitly looking like a 1/16th.
Ill tap it slightly bigger and see what happens. Im not worried about spool up, it spools fast enough as it is,
i just want consistent boost while holding to redline for as long as possible.
I know theoretically that should hurt it slightly in holding to redline longer but oh well.

And i definitly have the old Joe P. mbc, I mean like 3 years old kind of old.

 

[dsm-onster]

So going bigger on the bleeder oriface helps spike. I've used ebay ones with larger holes and used the spring/ball from my gillis and spike was still noticably larger with the ebay unit. I thought that the hole bleed too much down as boost was rising allowing the actuator not to see boost until too late thus the spiking. . . But I havn't experimented thoroughly with a rubber ducky and shaving cream. So my scientific study is likely skewed .

I'm brainstorming a way to run my ball/spring valve with "high spike" using my external and counter the spike with another valve. Maybe there's a way for an on/off without having to run a vacuum solenoid.

...I also think chosing intake runner length, and exhaust runner length is a detail that gets overlooked. Boosted applications usually don't see much benefit since the charge characteristics change so much. But, it's something that Porsche has been monkeying with for years in their turbocharged enginebays. They also have done "revolutionary" work in plenum design for a turbocharged application.

2008 Porsche 911 GT2 intake manifold (I cannot remember at all where I got this photo and caption). EDIT-- Here is where this came from.

 

[Mr Peepers]

So going bigger on the bleeder oriface helps spike. I've used ebay ones with larger holes and used the spring/ball from my gillis and spike was still noticably larger with the ebay unit. I thought that the hole bleed too much down as boost was rising allowing the actuator not to see boost until too late thus the spiking. . . But I havn't experimented thoroughly with a rubber ducky and shaving cream. So my scientific study is likely skewed

Shhhh....

Here's my take on it, although I may be repeating what I've said in previous posts:

You can trick the WG into opening later two ways with a boost controller:
1- Bleed off pressure from the line
2- Block it from seeing any pressure until a certain psi is reached

A ball and spring type uses both of these methods since a bleeder hole is required to keep it from acting like a check valve and holding pressure within the WGA, holding the flapper open. So, to reach the same desired boost you have a mix of these two methods, and depending on how much of each is applied will ultimately affect how the WGA responds.

Either tighten down the spring and bleed less air or loosen the spring and bleed more air. The more the spring is tightened, the higher the psi is reached(closer to the desired value) before the WG cracks open to begin controlling boost. The larger the bleeder hole, the looser the spring will have to be to achieve the same desired boost. This will allow the WG to see a pressure signal at a lower psi and start opening more progressively as boost builds, giving it a better chance of "catching" the boost before it overshoots the desired value. Of course this will sacrifice some spool in the process. Likewise, blocking the signal for too long will tend to spike before the WGA can respond in time.

You can also look at it like this. When you first install a MBC and have it all the way out, your boost will inevitably rise. Why? It is acting like a bleeder valve; lets say this is 10 psi over the WG spring 8. Now, you up the boost to 15 psi- everything is still great. When boost comes on, it hits noticeably harder than stock. Turn it up to 20 psi. Occasionally spikes a couple times, but has never hit so hard. Dial in your tune and squeeze out 24 psi. Hits hard but is spiking more and more. Why? The amount of air bled remained the same, but you're blocking the signal more and more as you tightened the spring; upsetting the balance and leaning more towards "blocking" rather than "bleeding". Sure it hits hard as hell, but the spike is now affecting your on the edge tune.

IMO a certain sized bleeder hole is only ideal at a certain boost pressure. Up that boost pressure and it spikes, lower it and you aren't getting the best spool up possible- a very fine line. THAT is why I'm going to make mine adjustable.

Note: The check valve affect of making the bleeder hole too small likely won't be reached until after spike is present.

Keep in mind that there are other factors to boost spike in the design and materials of the MBC to begin with, I'm just stating the effects of spool-up/spike given the same set up and same MBC. A lighter ball, larger volume of flow(lines/internal bottlenecks), and even spring rate may play a role in boost spike.

Phew. Anyway... I hope this made sense.

 

[kenamond]

An alternative to my original idea of adding a nipple to the MBC bleeder hole would be to plug the bleeder hole and tee into the MBC-to-WGA line and run that to a bleeder.

I agree completely with Matt (not because I got a turbo cheap, either). But I feel I must elaborate

Donmagicjuan and I had a nice discussion about ball/spring MBC function a while back. When the ball starts to crack, a bit of air flows past it but right out the bleeder hole...but there will be a bit of pressure in the WGA (air doesn't move out a bleeder unless it's pushed out which means there's pressure on the down-wind side of the ball). As the flow past the ball increases, the bleeder flows the same amount, but only because the pressure pushing air out has gone up (the bleeder requires more and more pressure to flow the same as the ball/orifice). Eventually, the bleeder requires 8psi of pressure to equal the flow of the ball/orifice and the WGA opens.

A bathtub experiment would demonstrate this. Take a cup, poke a hole in the side at the bottom. Start filling the cup up and notice how the water shoots out more and more the deeper the cup gets. The hole is the bleeder, the water you're pouring into it is (sort of) the air leaking past the ball, and the depth of the water in the cup is the pressure the WGA sees. In that experiment, the inflow of water doesn't equal the flow out the hole, but it still demonstrates that the bleeder will only flow more if more pressure is applied.

I like to understand things by looking at the extreme cases. Looking at bleeder size and MBC spring preload, we have 4 cases:

1) Low spring preload, small bleeder: When the ball cracks, the bleeder gets quickly overwhelmed and the WGA basically sees boost pressure even when it's way below 8psi. The WGA is getting a constantly increasing pressure as boost rises, opens smoothly, and boost levels smoothly. Acts like running no MBC (vac line from compressor to WGA).

2) Low spring preload, large bleeder: When the ball cracks, the bleeder leaks easily. The flow past the ball has to be much higher to eventually reach 8psi and open the WGA. Boost level is dominated by the bleeder hole size and might not even change if the spring is preloaded quite a bit more. Basically, the WGA starts seeing pressure at very low boost and that signal rises smoothly with boost so that the WGA opens smoothly and boost levels smoothly. Acts like a bleeder.

3) High spring preload, small bleeder: When the ball cracks, the bleeder is immediately overwhelmed and the WGA sees compressor output pressure. WG slams open right after the ball cracks. Me thinks it'd spike here, as there isn't much of a warning to the WG so the boost feedback delay leads to some extra boost.

4) High spring preload, large bleeder: When the ball cracks, the bleeder can handle the flow, but as boost increases, the flow past the ball increases and eventually pressurizes the WGA to 8psi. If the ball cracks at 10psi but it takes several more psi to overwhelm the bleeder, the WGA gets a nice warning and can smoothly open, and boost levels smoothly.

So the bleeder gives the WGA a smoothly increasing signal allowing it to smoothly open and reduce spiking. The spring preload determines when the WGA will start seeing this increasing signal. So you want the spring to hide boost completely up to a point for spoolup and then use the bleeder size to control how much higher than that cracking pressure the boost should be while feeding the WGA a smoothly increasing signal (giving it time to respond and limit spikes). More bleeder gives earlier WGA signal but less spike, less bleeder gives later WGA signal and more spike.

So I definitely think that an adjustable bleeder would be great for high-boost applications. Maybe some MBCs out there SUCK for low-boost but would be perfect for 24psi on an 8psi WGA...because the bleeder is too big for low-boost. If you can control both, you can raise the cracking pressure as high as you can and then increase the bleeder to decrease the spikes. If you can't bleed enough, lower the spring preload a bit and try again. I imagine you're still going to be limited such that you either spike in 4th and 5th or degrade boost response in 1st through 3rd. That ties into the feedback time the WG has on the boost signal which is short in 1st and long in 5th.

Ideally, you'd have 5 MBCs, one for each gear. 5th would have the highest spring preload and smallest bleeder (things happen slowest, so the WGA needs the least warning), and 1st would have the lowest spring preload and largest bleeder (things happen fastest, so the WGA needs the most warning).

But then what do you do when you're at 5k in 3rd and tromp it? You might spike there but not if you did a pull in 3rd.

I think I'm starting to appreciate the EBC a bit more now

 

[Mr Peepers]

I'm glad we are on the same page, and we think alike in regards to going to each extreme to have a better understanding on how things work. All hell breaks loose when there are two exponential variables fighting though

I kind of alluded to it in the article, and I completely agree a larger bleeder hole is required as the boost pressure rises to keep it from spiking. I really like having two viewpoints on the same subject. Explaining what's on my mind isn't exactly my strong point I tend to edit waaay too much.

These little buggers are more complicated than most people think! I never really thought about the effects between building boost from boost threshold or from lag at an rpm above the threshold. That would be factored out from how it behaves between shifts though I reckon. I'll do some more pondering.

 

[Black_Bullet]

4) High spring preload, large bleeder: When the ball cracks, the bleeder can handle the flow, but as boost increases, the flow past the ball increases and eventually pressurizes the WGA to 8psi. If the ball cracks at 10psi but it takes several more psi to overwhelm the bleeder, the WGA gets a nice warning and can smoothly open, and boost levels smoothly.

So the bleeder gives the WGA a smoothly increasing signal allowing it to smoothly open and reduce spiking. The spring preload determines when the WGA will start seeing this increasing signal. So you want the spring to hide boost completely up to a point for spoolup and then use the bleeder size to control how much higher than that cracking pressure the boost should be while feeding the WGA a smoothly increasing signal (giving it time to respond and limit spikes). More bleeder gives earlier WGA signal but less spike, less bleeder gives later WGA signal and more spike.

K, well I am going to try to take my bleeder hole one size bigger without getting carried away.

You guys have a point about the more your tighten the MBC the more likely spike will be due to delayed response. A year ago I used to only run 15 psi daily and I spiked to 17lbs in upper gears. I started running 20 psi and there would be no spike ever ( Until I did my CAI ) Now turning the boost up beyond 20lbs it spikes harder than ever.
The more I tighten it the worse It gets huh.
Partial due to the fact that im more than doubling stock boost ( WGA spring )

Like I said im only spiking about 2 psi in 4th, and 5th, but thats still limits me to run the boost at only 22psi in 2nd and 3rd gear so that the upper gears wont exceed 24.
Id like to feel 24psi in 2nd and 3rd without worrying about 4th and 5th.
My tune is perfect for 24psi and more on the edge, that spike pulls timing, and I've found that once knock starts it can chase you thru your whole pull.

What do some of you guys think about stiffer waste gate actuators
( Like the 17psi HRC unit for example )

Another memeber on here noticed boost held to redline longer on his T28 with it.
Should I throw the shimming of the WGA idea out the window at this point In terms of keeping the flapper closed longer, will that be counter productive to me trying to limit spike?

Thanks for all the great posts !

 

[MIdsm]

Browser Warning
Browser Warning

 

[Mr Peepers]

Browser Warning
Browser Warning

I was going to say, "I get it now!" but nope; checked again and the regulator setup is leading to the bottom port of the wastegate. Would have made perfect sense if it was going to the top (and a line was leading directly from a pressure source to the bottom).

Not saying its not plausible, but its not clicking for me. The whole problem is the regulator setup and how the whole thing controls boost once it reaches "x" psi above WG spring pressure.

The wastegate will see boost 1:1(disregard the pressure relief valve for now) until it reaches the set output of the regulator, and obviously the regulator output(seen by the WG) must be set below the WG spring pressure before it will start to effectively raise boost. Once that happens, the wastegate won't see any more boost than what the regulator is set for, no matter what the actual boost continues rising to. This means the WG will stay "x" degrees opened, and the only thing perhaps controlling boost is the exhaust gas pushing it open further. This doesn't seem very stable. In a "normal" feedback system, if the boost raises just a tiny bit, the WG opens a tiny bit more- lowering boost back down and maintaining a balance. This all happens simultaneously and WHAMO! a steady boost pressure is obtained. With a regulator keeping the pressure seen by the WG at a fixed amount, no matter what the actual amount changes to, I can't wrap my mind around how it will stay at a somewhat stable boost pressure; it is no longer a feedback system. The only thing I can see is the pressure of the exhaust gas against the WG regulating boost at that point. Maybe that's enough, I don't know.

EDIT: Hmmm... maybe that's the whole point. The boost just slowly "creeps" up to a certain pressure on its own from the WG stopping at a certain point then lowers back down; judging by the curve compared to stock. This would likely be highly dependent on the turbo and boost pressure trying to be run; similar to boost creep or dropping boost. "Controlled (boost)creep."

 

[kenamond]

I was going to say, "I get it now!" but nope; checked again and the regulator setup is leading to the bottom port of the wastegate. Would have made perfect sense if it was going to the top (and a line was leading directly from a pressure source to the bottom).

Not saying its not plausible, but its not clicking for me. The whole problem is the regulator setup and how the whole thing controls boost once it reaches "x" psi above WG spring pressure.

The wastegate will see boost 1:1(disregard the pressure relief valve for now) until it reaches the set output of the regulator, and obviously the regulator output(seen by the WG) must be set below the WG spring pressure before it will start to effectively raise boost. Once that happens, the wastegate won't see any more boost than what the regulator is set for, no matter what the actual boost continues rising to. This means the WG will stay "x" degrees opened, and the only thing perhaps controlling boost is the exhaust gas pushing it open further. This doesn't seem very stable. In a "normal" feedback system, if the boost raises just a tiny bit, the WG opens a tiny bit more- lowering boost back down and maintaining a balance. This all happens simultaneously and WHAMO! a steady boost pressure is obtained. With a regulator keeping the pressure seen by the WG at a fixed amount, no matter what the actual amount changes to, I can't wrap my mind around how it will stay at a somewhat stable boost pressure; it is no longer a feedback system. The only thing I can see is the pressure of the exhaust gas against the WG regulating boost at that point. Maybe that's enough, I don't know.

EDIT: Hmmm... maybe that's the whole point. The boost just slowly "creeps" up to a certain pressure on its own from the WG stopping at a certain point then lowers back down; judging by the curve compared to stock. This would likely be highly dependent on the turbo and boost pressure trying to be run; similar to boost creep or dropping boost. "Controlled (boost)creep."

I found it an interesting read, but I didn't really understand what the components do specifically enough to ponder the details of the design. Seemed a bit pricey, but I don't know what the exchange rate for Aussie currency was back when that was written.

 

[Black_Bullet]

I tapped out the bleeder hole on the mbc slightly bigger today. I barely made it any bigger just to be safe and I cant say I really notice a difference but I know with stuff like this any tiny change could be significant.
Idk if I had to find something to point out, " Maybe "
boost doesnt build as fast, but just maybe... There was still a 2psi difference between 2nd and 5th gear;
but it didnt seem like a spike anymore; it rose up to whatever boost level it was going to peak at and held there until the VE of the engine caused the turbo to decrease in psi.

Make any sense?

I didnt spend long testing it or playing with the mbc, ill take another look at it tommorow when I get back from the dyno.

EDIT: The car definitly builds boost a bit slower with the bleeder hole enlarged.
I cant tell about the spike yet, but its still there just not as bad i believe.

A little run up.

Set at 20psi I fall to about 16-17 psi at redline.
Set to 23psi I fall to about 19-20 psi at redline.
Set to 24-25 psi I fall to about 20-21 psi by redline depending how cold ambient temps are.

The boost falls so fast that when I hit 25psi and start falling it looks more like a boost spike but Im coming to realize its not spiking its just dropping boost faster, because if I hit 25psi at only partial throttle it will remain 25psi until I lay into it.