As per Ed's reccomendation, I'm now running a Hallman Pro RX Kit MBC (he runs one too). Boost tapers as it should. Spoolup and response are remarkably better.

Here's some supporting data from a log I took today.

I had Ed look at my logs and they look good. He thinks it's probably running a bit rich in open loop based on my low EGT's at full throttle and next time he's up here, he'll bring his wideband and lean it out a bit for this elevation. But this has nothing to do with me adding the MBC. My tune was good before and after adding the MBC, no tweaking necessary. Ed says you can probably run this MBC on any tuned map, maybe even Cobb's canned maps. DO NOT RUN AN MBC ON A STOCK MAPPED SUBARU EVER.

He used one of my logs (maybe the one above) to make the following graph. Numbers aren't great, but this is at ~5,200 feet of elevation. I'm just glad to get rid of all the lag I was seeing at this elevation.

I am not disputing your results/impressions, just trying to understand Ed's logic.

If the ECU dynamically controls an electronic boost controller, why can't the levels that you are fixing in the MBC just be programmed into the ECU and controlled by the EBC?

It just seams counter intuitive to use something mechanical when you have a precision electronic option available.

It's because the stock boost solenoid is slow to respond. Upgrading to a better solenoid helps, but apparently nothing responds like an MBC.

That's because a MBC *can't* respond. It's a spring. And by your log, it looks like it's hooked up incorrectly, because you're still tapering boost at high RPMs. Do you have both the MBC and the EBC hooked up?

Here's what MBC's usually do:

1000 2
1500 3
2000 4
2500 8
3000 14
3500 16
4000 16
4500 16
5000 16
5500 16
6000 16
6500 16
7000 16

And it's that 5500 to 7000 16psi that ends up killing the motor... since that turbo can't put out that sort of boost efficiently at that rate of flow, you end up running really hot.

Also, what are your EGT's running around with a MBC, specifically at partial throttle closed loop 16psi. Your data is from an open loop WOT pull, no? Aren't you concerned about running an MBC w/o changes to your closed loop programming, since it's now possible to hit full boost w/o running with open loop fuel levels?

Call me a stick in the mud, but switching to a MBC is a step backwards IMO. Reeks of the MBC/UniChip days.

My thoughts exactly.

Just because it uses a spring doesn't mean it can't respond. It's just a much more simple system. The stock solenoid is completely disconected from all vaccum lines and I will probably just remove it completely as Ed suggested. The MBC is connected properly.

Perhaps you have used an MBC that doesn't taper, but this one does.

Full boost, partial throttle EGT's are slightly higher. Under load on the freeway on a cool morning its gone as high as 1525*F whereas before it would only go to 1500*. However its rare I see EGT's above 1400 and I've been watching the gauge like a hawk. Ed says these temps are perfectly fine.

Yes, that data is from a WOT 3rd gear pull.

Apparently once you remove the programming on the stock map that is designed to save on emissions, you program the ECU to compensate AFR's for full boost at partial throttle.

Counter-intutive? Yes. But the results speak for themselves.

Please explain to me how the spring in your MBC gets softer as the RPMs increase.

Slow to respond where? If the solenoid is programmed not to cycle until 16PSI and then to cycle to maintain 16PSI, how is that different than a MBC?

If it does not cycle at all until 16PSI, the boost should be indistinguishable from a MBC under low RPM conditions, so lag, etc. should be identical. Thus I am confused by your comment of "Spoolup and response are remarkably better." If that is the boost curve you wanted, why could it not be configured in the ECU? I would think not cycleing at all should not be an issue even if the solinoid is "slow".

Again, I am only trying to understand, not dispute.

I am also trying to understand the taper. Is there some other connection to the MBC so it can sense RPM?

I show it a picture of Jonny's Mom. :P

I don't know Dude. Perhaps its because it responds to cubic feet/minute and not PSI?

I don't know exactly why the stock setup is so slow but it is.

The MBC is connected to two of the three points that the EBC was connected to, the turbo and the wastegate actuator. You plug the connection from the intake with an included cap and it is not used with the MBC.

I'll ask Ed to explain how the taper works. I asked him before but I didn't really get it. It has somthing to do with the bleeder hole on this MBC being the right size I think.

Telekinesis, duh.

so a leaky MBC creates the boost taper on the top end ?

My head hurts. Man I hope my 255's fit!

Ok guys... there are a couple major gaps in the info here.

First of all, the stock boost control solenoid raises boost by pulsing and bleeding boost before it gets to the wastegate actuator. So the solenoid needs to pulse in order to make boost, not the other way around like someone here stated. This is in essence a failsafe system so that if the solenoid goes bad (and they have been known to), you don't get uncontrollable boost that grenades the motor but rather just end up with wastegate spring pressure (~7psi).

This brings us to the issue of response. The stock solenoid can only bleed so much boost, so Subaru uses a series of restrictors to help it out and to try to smooth out the operation. These work ok for stock boost levels, but once you start running higher boost the restrictor size becomes inapropriate. This can certainly be tuned around, although some cars require a replacement restrictor to reach desired boost levels. Using the appropriate restrictor and propper tuning, you can get pretty much any boost curve you want... that's not really the issue here.

The real issue is when you you're just cruising along and decide to hammer on the gas. The ECU sees the TPS increase, then looks up the WDC for that RPM/TPS, ramps the solenoid to that duty cycle, looks at the boost level reached, then makes incremental adjustments to the duty cycle to try to hit the target boost for that RPM/TPS. This is how any closed loop system works. The issue is how quickly the ECU can actually ramp up the duty cycles while maintaining stable boost without spiking and how quickly it can make the small changes to reach target boost. Depending on setup, external conditions, and target boost levels, the system can hit boost pretty quickly and accurately or can take its sweet time about it and introduce a good deal of lag on quick throttle transitions. Remember that the ECU is having to respond to many different inputs to zero in on its target boost. Combine this with the inherent limitations of a bleeder type solenoid setup and you get the idea of why this system is not ideal for all out response.

One solution here is to replace the stock 2-port solenoid with a 3-port solenoid that can be set up in interrupt mode rather than as a bleeder. In this case you don't need to use any restrictors and instead of bleeding boost before it reaches the actuator, the solenoid actually interrupts the signal before the actuator sees any boost at all. In essence this type of setup acts much more like a ball and spring MBC and allows significantly quicker response than the stock bleeder solenoid. Response is, however, still limited by the fact that it is a closed loop system and the ECU must still ramp boost up at a controlled rate in order to reduce spiking, while also having to make small adjustments constantly to get closer to target boost. Granted, the response of this system is much more precise and much better overall than the stock setup, but there are still limitations. Even considering this, I recomend this setup to about 95% of my customers because its a great overall solution that comprimises a bit of response for a system that's fully controlled by the ECU and requires no user intervention.

This brings us to the remaining 5% of my customers who would benefit from the best possible response and aren't afraid of learning about something a bit different. For such people, like Cody and myself, I recomend using a good quality MBC such as the Hallman. I've tried many MBC's in the past and this unit has proven to be the best by far. Some older, low quality MBC's only used a ball and spring and resulted in significant spiking. There were also some units that only worked as a bleeder and resulted in inconsistant boost levels with varying conditions and less that ideal boost response. The newer units use a combination of the two systems. While the ball and spring is the main method of boost control, they also have a small bleeder hole to stabalize the boost curve. These units provide unmatched boost response with very consistant boost levels. While boost levels will in fact vary slightly based on temperature, these variances are very appropriate and very similar to what you would see while using the stock boost control system. These variances actually occure moreso from the efficiency of the IC rather than varying performance of the actual MBC. Remember that the MBC is referencing boost before the IC, so while it may be maintaining totally consistant boost levels pre-IC, if the TMIC is heatsoaked manifold pressure will appropriately be lower.

Now lets talk about boost tapering. No MBC I've ever seen runs X boost all the way to redline... atleast not on a stock turbo. The ball and spring is not a binary on/off system. It reacts much in the same way as your wastegate actuator spring. When it sees some boost, it opens a bit, when it sees more boost, it opens more allowing the actuator to see more boost. This results in the actuator seeing a fairly constant pressure variance under boost which results in an almost constant wastegate opening angle. As RPM's increase, the turbo runs less efficiently and requires a higher wastegate opening angle to run the same boost levels. Since the wastegate remains at a fairly constant angle, boost tapers off naturally with RPMs. This taper can be manipulated by the size of the bleeder, and of course will vary quite a bit with turbo size. In general, however, this taper is very appropriate and results in a boost curve that I would end up running using a solenoid for boost control.

The great thing about an MBC is that its an extremely simple mechanical pressure regulating system that does a much better job in most cases than the relatively complex closed loop solenoid system. The MBC also acts much more precisely and naturally than a solenoid which pulses air in small quick bursts. I've always been a fan of good MBC's for this reason.

Now lets talk about partial throttle high boost situations. While this is a real issue with a stock ECU, most tuners choose to reprogram the OL/CL switchover so that they can also run higher boost levels at low throttle angles. This makes the car feel torquier and more fun to drive. Cobb does this in their base maps and every tuner I know does the same thing in varying degrees. This means that the ECU is ready for the higher boost levels at part throttle conditions and runs the appropriately programmed open loop fueling levels to accomodate this. This was always the big issue with MBC's on Subarus and where many of the misconceptions originated. Before I could afford full blown tuning software, I actually designed a simple controller that would watch boost levels and force the ECU into open loop once a pre-set threshold was crossed. This allowed 02/03 WRX's to run MBC's on the stock ECU with no danger and smooth operation. I ran this setup on my own car for 60k miles and sold about 60 units without a single reported issue.

Phew... ok I covered a lot here and I hope I didn't miss anything. If I did miss something or didn't explain something well, please let me know and I'll try to explain more clearly.

Thanks
-- Ed

I certianly understand your reasoning far better than I did this morning. Thanks.

So are you actually lowering the cutoff for closed loop, and running open loop? If not, what does the ECU do about not seeing expected responses to duty cycle changes for the EBC during closed loop?

Thanks for the info Ed.

I'm still missing the advantage to the MBC. I understand the response limitations of the stock EBC, but I guess I'm just not seeing them as actual limitations. What sort of response gain are you getting? On my 2.0L w/ stock turbo, I was making full boost before 3000 rpm, and had zero driveability/throttle response issues... how much more can you ask for?

Also, when I was running a MBC, back in the pre-reflash era, I was seeing spiking to 18psi, PTFB issues, no taper at redline, and tremendous swings in max boost due to weather. For example, the car would hold 17psi in the winter at our altitude with the MBC adjusted to it's lowest setting, I was forced to keep my foot out of the throttle just to prevent over boosting, and eventually just hooked the stock controller back up for the winter.

So, I understand that a MBC can be made to work well (though I would certainly get things tuned specifically for it), I just don't understand why anyone would replace a well designed OEM system with a $2 spring (and pay $170 for it). In my experience, the factory setup is far superior to a manual solution, even with the "limitations" inherent to the design.

$170? I paid $79.99 plus shipping for a new one. MSRP is $120 IIRC. Comes with everything you need.

Oh sorry, they're $180, not $170:

http://www.boostcontroller.com/index...26category%3D5

No, like I said, the MSRP is $120

http://www.boostcontroller.com/index...26category%3D5

You're linking to a kit that includes adjustment from the drivers seat.

I got mine off Ebay for $79.99.

Good for you. That doesn't negate the fact that what is functionally only a $2 spring is for sale for $180.

The spring is probably worth less than $2. The entire device is worth every penny IMHO. Don't hate.

I was being kind and giving the spring the benefit of the doubt.

As far as "hating", I'm more unimpressed than anything... I don't see a significant advantage to the MBC, and in my past experience they're nothing but problems. The whole issue seems like a solution to a manufactured problem. Show me some back to back dyno pulls that demonstrate a useful low-end power/torque gain and you might start to convince me, but for now I still think it's spending money to take a step backwards.

I am surprised there is not a BC based on a stepper motor or similar actuator instead of vacume slenoids, and springs, or maybe there is. It would seam that the best of both worlds could be achieved by an ECU/computer controlled stepper motor contraption of some form.

With the reflash, the switchover to open loop occures at lower throttle angles which is apropriate for the high boost, low throttle condition seen with the new boost map as well as with an MBC.

Thanks
-- Ed

There are very significant response gains to be had with an MBC over the stock boost control system. While the low RPM boost threshold is lowered a bit because boost can ramp much quicker, the main gains are in the form of boost response during quick throttle transitions. Here's a little experiment for you guys to try with the stock BCS: Get in 3rd or 4th gear around 3500rpm, boost at part throttle to about 9-10psi for couple seconds, then quickly go full throttle and watch the rate at which your boost climbs. Now get to the same RPM and from cruising quickly go to full throttle and watch the rate of boost increase. You'll see that the ECU has to ramp boost much more slowly when starting from a bit of boost.

There are also very significant gains in the lower gears because with the MBC you can hit significantly higher boost levels than with the stock system due to the improved ramp rate.

I certainly would never suggest running an MBC with a stock ECU, but with a reflash that is designed to handle part throttle, high boost conditions, there is no danger as long as the boost levels are appropriate. From your description, it sounds like you had one of the poorer designed MBC's out there. I have some experience with such units and I'll agree that their performance is no where near satisfactory. A unit like the Hallman really can't be compared to those old MBC's put together out of plumbing parts. The Hallman's springs are chosen for consistant, progressive operation, their RX model uses a ceramic ball which is much lighter than the standard metal ball, and the overall design of the body and the adjustment is wonderful. It doesn't require any tools to adjust, doesn't have any lock nuts, you just turn the knob and it stays put with even the harshest vibrations. I know this must sound like a commercial, but I have been really impressed with the overall quality, design, and performance of this unit compared to every other MBC I've used.

The hallman unit can be had for about $80 which IMO is very well worth the performance gains for anyone who would benefit from the best possible response which includes pretty much any motorsport competitor or even just the "enthusiastic" driver.

Thanks
-- Ed