Arduino Electronic Boost Controller Project
04-12-2015, 09:59 PM
#91
Rennlist Member
Thread Starter
I do like refresh951's build a lot, for sure.
Not that I am any kind of expert of anything, but just looking at it, I think it's all sized Goldilocks, just right. 3" exhaust is about right, as well as 2.5" compressor to intercooler pipe. The stock core can handle the flow and heat well enough, as long as one modifies those end tanks to produce a lower pressure drop -- which he's done. Etc.
All cost effective, too. Anyone can blow a bunch of money on a car and make it really fast, but I have a special appreciation for builds that make a ton of reliable power that could be replicated for a reasonable amount of money.
There are a couple of things that I'd do differently, I suppose. First, at the same 500 rwhp power level, the whole system would be a lot less stressed with the 4-valve head. Less boost needed, more compression allowed, everything more efficient. Now, it might not be as cost effective as his other mods, but I'd personally definitely go with a 4-valve head.
Second, my personal preference would have been to run an rpm-specific boost profile and retain longer intake runners. If one runs a constant boost pressure across the rpm range, then I can see the appeal of the short-runner intake. You'll lose maybe 15% of torque under 3000 rpm, but then keep the torque curve flat instead of falling on its face at high rpms and make about 10% more power at 6500 rpm or so. I am making up ball park numbers there, of course.
The whole runner length calculus however changes when the boost is allowed to vary per rpm with an electronic controller. Now, in my experience you want to retain the long runners to get the low rpm torque up. When the long runners and small cams are about to fall on their face at high rpms, one can then increase the high rpm boost to compensate this worse cylinder filling. It's thermally inefficient at those rpms but much less so than what one would first think once one recognizes the cooling effect of "off tune" intake manifold runners.
Still, all things considered, this is one of my favorite 951 builds that I've read about.
Not that I am any kind of expert of anything, but just looking at it, I think it's all sized Goldilocks, just right. 3" exhaust is about right, as well as 2.5" compressor to intercooler pipe. The stock core can handle the flow and heat well enough, as long as one modifies those end tanks to produce a lower pressure drop -- which he's done. Etc.
All cost effective, too. Anyone can blow a bunch of money on a car and make it really fast, but I have a special appreciation for builds that make a ton of reliable power that could be replicated for a reasonable amount of money.
There are a couple of things that I'd do differently, I suppose. First, at the same 500 rwhp power level, the whole system would be a lot less stressed with the 4-valve head. Less boost needed, more compression allowed, everything more efficient. Now, it might not be as cost effective as his other mods, but I'd personally definitely go with a 4-valve head.
Second, my personal preference would have been to run an rpm-specific boost profile and retain longer intake runners. If one runs a constant boost pressure across the rpm range, then I can see the appeal of the short-runner intake. You'll lose maybe 15% of torque under 3000 rpm, but then keep the torque curve flat instead of falling on its face at high rpms and make about 10% more power at 6500 rpm or so. I am making up ball park numbers there, of course.
The whole runner length calculus however changes when the boost is allowed to vary per rpm with an electronic controller. Now, in my experience you want to retain the long runners to get the low rpm torque up. When the long runners and small cams are about to fall on their face at high rpms, one can then increase the high rpm boost to compensate this worse cylinder filling. It's thermally inefficient at those rpms but much less so than what one would first think once one recognizes the cooling effect of "off tune" intake manifold runners.
Still, all things considered, this is one of my favorite 951 builds that I've read about.
On your first point, I would say I must agree. If I had it to do again I would go with a 968 4V head. The benefits are many and the cost is really not that bad. The great thing is that we have found some really cheap ways to improve the 8V 951 head and they can be found in abundance.
On your second point, I guess I can see why you may view it that way. However, I am absolutely convinced the short runner/large plenum intake was the best approach for my setup. You really need to drive both setups to feel the difference. Early on I focused on maximizing bottom end torque and it looks good on a dyno chart but it is really not very usable. The car has more than enough torque down low especially with the 95mm stroke. The short runner/large plenum intake (along with the cam) really makes for a free breathing top end that pulls unbelievably quick and keeps on pulling well past 6500 rpm. Time does not show up on the dyno but it matters and if you could drive both setups I think you would appreciate the acceleration at the top end. My feeling is that I could run more boost up top with a longer runner and improve the dyno chart but I do not think the car would be as fast.
I do appreciate your input however and thanks again for the compliments.
04-13-2015, 06:49 AM
#92
Nordschleife Master
Thanks for the compliments! I have also read quite a bit about your build, very impressive.
On your second point, I guess I can see why you may view it that way. However, I am absolutely convinced the short runner/large plenum intake was the best approach for my setup. You really need to drive both setups to feel the difference. Early on I focused on maximizing bottom end torque and it looks good on a dyno chart but it is really not very usable. The car has more than enough torque down low especially with the 95mm stroke. The short runner/large plenum intake (along with the cam) really makes for a free breathing top end that pulls unbelievably quick and keeps on pulling well past 6500 rpm. Time does not show up on the dyno but it matters and if you could drive both setups I think you would appreciate the acceleration at the top end. My feeling is that I could run more boost up top with a longer runner and improve the dyno chart but I do not think the car would be as fast.
On your second point, I guess I can see why you may view it that way. However, I am absolutely convinced the short runner/large plenum intake was the best approach for my setup. You really need to drive both setups to feel the difference. Early on I focused on maximizing bottom end torque and it looks good on a dyno chart but it is really not very usable. The car has more than enough torque down low especially with the 95mm stroke. The short runner/large plenum intake (along with the cam) really makes for a free breathing top end that pulls unbelievably quick and keeps on pulling well past 6500 rpm. Time does not show up on the dyno but it matters and if you could drive both setups I think you would appreciate the acceleration at the top end. My feeling is that I could run more boost up top with a longer runner and improve the dyno chart but I do not think the car would be as fast.
I don't disagree with you that your car is faster with a short-runner intake if the boost is held constant. That's clear from your dyno graphs. The average hp between shift points is higher now than it was with the long-runner intake manifold.
What I am suggesting that this might no longer be the case when the boost can be varied with the rpm. The way I see it is that once that GT3582R has spooled, you'll be like Ernie "Big Ern" McCracken at the end of Kingpin: Finally above the law! ;-) With the turbine spooled and with the ability to vary boost setpoint with rpm, you can do anything and compensate for any sort of deficiencies in the intake manifold (within reason) especially with your mild factory cams. I am not saying that I am sure that long runners would be better with the rpm specific boost control, all I am saying is that it's a possibility to take seriously.
"Big Ern, a million dollars. What are you going to do with it?
"l don't know. All l know is, l finally got enough money... that l can buy my way out of anything. l can do anything l want when l get my money later. And l won! Finally, Big Ern is above the law! lt's a great feeling."
"l don't know. All l know is, l finally got enough money... that l can buy my way out of anything. l can do anything l want when l get my money later. And l won! Finally, Big Ern is above the law! lt's a great feeling."
If you can get Arduino based cheap controller to include the features of Eboost2, including closed loop control with rpm-specific set points, it'll be a great device.
04-13-2015, 06:59 AM
#93
Boost cannot compensate for a poor VE, which is directly affected by the geometry of the intake manifold. If the VE is poor, increasing boost will add stress without increasing output by any useful margin.
04-13-2015, 09:48 AM
#94
Nordschleife Master
For the most part, pump gas turbo cars are knock limited above the boost threshold rpms. Perhaps this is not the case for the E85 engines as much, but I am convinced that it's the case for almost all pump gas turbo cars including most hot-rodded 951s.
Knock limit itself is determined by my understanding primarily by three things. First, intake charge temperature. Second, the pressure of the fresh charge in the cylinder. Third, the residual exhaust gas left in the cylinder when the exhaust valve closes.
Suppose that a car has a restrictive intake that leads to say 80% natural VE instead of 100% natural VE at some high RPM. Suppose further that the car has a very effective intercooler. If one compensates for the lower VE with higher boost, two things happen. First, exhaust manifold pressure has to increase (to provide the power to drive the compressor harder). Second, the compressor outlet temperature increases. Both are bad.
However, let's not forget the intercooler. If the intercooler is very efficient and brings the charge temperature to close to ambient, then we now have high pressure cool charge in the intake manifold. When the 80% VE engine ingests that air, the pressure has to and will drop because of the lower than 100% VE. With the pressure drop, the charge temperature will drop as well -- theoretically below ambient temperatures are possible. (Airplanes use a loosely related air-cycle intercooler to run the cabin air conditioning, by the way.) This will allow the engine to burn more charge mass before the knock onset holding everything else constant.
So why not make the natural VE really low to make most power? There's a limit to the benefits of this strategy, mainly because the rising exhaust back pressure will result in more burned hot exhaust gas staying the combustion chamber, which in turn increases knock. So this is not one of those cases where if little is good, more is better. Where the optimum is for the 951, I have no idea. Just making a conceptual point.
For my 928 S4 which has 4-valve head, the strategy of compensating for falling VE with higher intercooled boost has been very effective in extending the torque curve. Similarly, for the 997 Turbo, the "expansion intake manifold" that deliberately reduces the VE at certain rpm points results in more power, better efficiency, and broader torque curve when run on pump gas: http://www.porsche.com/usa/models/91...take-manifold/
04-14-2015, 05:08 AM
#95
I agree with the global theory which may be attractive to apply on a properly-breathing 4V head but I should have precised that the stock 2V heads flow so little it's pretty easy to reach their limit, by which increasing boost will add excessive stress in the sense that compressed air will blow headgaskets as it has nowhere else to go.
We can improve VE with a more opened camshaft then use a larger turbo to fill the void up top, so to speak, but that will reduce the width of the usable powerband.
A flat torque curve imo highlights how restricted an engine can be regarding to its top end hardware. I would always favour an engine that's got a normally-aspirated power display, and if turbocharging greatly helps flattening the torque curve, I see little interest in flattening it per se, the most modern iteration of this approach to maximise efficency being electric motors and their dullness. This does not belong in a Porsche imo.
Just my 2 cents.
We can improve VE with a more opened camshaft then use a larger turbo to fill the void up top, so to speak, but that will reduce the width of the usable powerband.
A flat torque curve imo highlights how restricted an engine can be regarding to its top end hardware. I would always favour an engine that's got a normally-aspirated power display, and if turbocharging greatly helps flattening the torque curve, I see little interest in flattening it per se, the most modern iteration of this approach to maximise efficency being electric motors and their dullness. This does not belong in a Porsche imo.
Just my 2 cents.
Last edited by Thom; 04-14-2015 at 05:31 AM.
04-14-2015, 10:18 AM
#96
Nordschleife Master
I agree with the global theory which may be attractive to apply on a properly-breathing 4V head but I should have precised that the stock 2V heads flow so little it's pretty easy to reach their limit, by which increasing boost will add excessive stress in the sense that compressed air will blow headgaskets as it has nowhere else to go.
We can improve VE with a more opened camshaft then use a larger turbo to fill the void up top, so to speak, but that will reduce the width of the usable powerband.
A flat torque curve imo highlights how restricted an engine can be regarding to its top end hardware. I would always favour an engine that's got a normally-aspirated power display, and if turbocharging greatly helps flattening the torque curve, I see little interest in flattening it per se, the most modern iteration of this approach to maximise efficency being electric motors and their dullness. This does not belong in a Porsche imo.
Just my 2 cents.
We can improve VE with a more opened camshaft then use a larger turbo to fill the void up top, so to speak, but that will reduce the width of the usable powerband.
A flat torque curve imo highlights how restricted an engine can be regarding to its top end hardware. I would always favour an engine that's got a normally-aspirated power display, and if turbocharging greatly helps flattening the torque curve, I see little interest in flattening it per se, the most modern iteration of this approach to maximise efficency being electric motors and their dullness. This does not belong in a Porsche imo.
Just my 2 cents.
The good news for you guys is that if this Arduino projects is successful, you can set the torque curve to whatever shape you want (within reason). It's really valuable tuning tool, I think the best $600 I've spent on my car is eboost2 with rpm specific boost control. Can't underestimate the benefit of rpm specific boost profile for a street car, it's huge!
Some off topic questions and comments:
Why does the 951 blow head gaskets?
Why does the pressure drop from compressor outlet to intake port factor into blowing head gaskets? Is it because
more restrictive intake side -> more boost needed at the compressor outlet -> more exhaust back pressure needed -> poorer cylinder evacuation of exhaust gasses -> more knock -> blown head gasket
? Or some other mechanism?In terms of cams, how much overlap the engine likes depends on the (total) pressure differential between intake and exhaust ports around the overlap. Covering up restrictive charge air pipes, intercooler, and intake manifold with more boost and accepting the pressure drop from the compressor to intake port means that one can't use cams with much overlap. Free flowing intake side (and turbine, of course) will allow one to be much more aggressive with the cams. This is all theory at this point, but I do have two sets of cams now that have an adjustable center sprocket so one of these days (years?) I'll get around to testing that theory.
04-16-2015, 04:42 AM
#97
On a stock-ish 951 I would say the main issue is the vibration of cylinders under excessive stress that end up cutting the headgaske, in addition to the heat pocket at the back of the head. The KKK turbine, the cam and the intake are arguably designed to give a satisfying compromise between off boost performance and the limited on-boost performance. With the same kind of restrictive turbine used with a 16v head/cams I would guess the issue would worsen.
Inversely, freeing up the turbine breathing with an accordingly sized compressor but retaining the stock cam with little overlap may not help with peak power if the head/cam cannot let the turbo flow as much as it can.
It is all probably easier to find a suitable out of the box modern turbo when using a 16v head/cams which may be close to modern flow standards, but getting an 8v set up "right" may not be quite as straightforward.
Inversely, freeing up the turbine breathing with an accordingly sized compressor but retaining the stock cam with little overlap may not help with peak power if the head/cam cannot let the turbo flow as much as it can.
It is all probably easier to find a suitable out of the box modern turbo when using a 16v head/cams which may be close to modern flow standards, but getting an 8v set up "right" may not be quite as straightforward.
Last edited by Thom; 04-16-2015 at 05:01 AM.
08-04-2015, 10:34 PM
#102
Addict
Lifetime Rennlist
Member
Lifetime Rennlist
Member
Join Date: May 2001
Location: Hotlanta - NE of the Perimeter
Posts: 12,269
Received 266 Likes
on
153 Posts
Guess the question now becomes
Is the Arduino worth the wait?
I've read this thread a couple of times, and have had the honorable pleasure of experiencing the controlled boost in Refresh951's car (
). Now it seems as though this will be a direct competitor to the eBoost2 system, which admittedly I know little about at this point.
So, is the Arduino better? Are the controls better thought out?
It appears that the feature of the different boost levels via PB are worth the price of admission alone from where I sit.
However, I would add that from where I sit, the time to market for this product is growing long in the tooth.
Is the Arduino worth the wait?
I've read this thread a couple of times, and have had the honorable pleasure of experiencing the controlled boost in Refresh951's car (
). Now it seems as though this will be a direct competitor to the eBoost2 system, which admittedly I know little about at this point.So, is the Arduino better? Are the controls better thought out?
It appears that the feature of the different boost levels via PB are worth the price of admission alone from where I sit.However, I would add that from where I sit, the time to market for this product is growing long in the tooth.
08-05-2015, 12:17 AM
#103
Rennlist Member
Guess the question now becomes
Is the Arduino worth the wait?
I've read this thread a couple of times, and have had the honorable pleasure of experiencing the controlled boost in Refresh951's car ( ). Now it seems as though this will be a direct competitor to the eBoost2 system, which admittedly I know little about at this point.
So, is the Arduino better? Are the controls better thought out? It appears that the feature of the different boost levels via PB are worth the price of admission alone from where I sit.
However, I would add that from where I sit, the time to market for this product is growing long in the tooth.
Is the Arduino worth the wait?
I've read this thread a couple of times, and have had the honorable pleasure of experiencing the controlled boost in Refresh951's car (
). Now it seems as though this will be a direct competitor to the eBoost2 system, which admittedly I know little about at this point.So, is the Arduino better? Are the controls better thought out?
It appears that the feature of the different boost levels via PB are worth the price of admission alone from where I sit.However, I would add that from where I sit, the time to market for this product is growing long in the tooth.
08-05-2015, 06:18 PM
#104
Burning Brakes
I've spent a lot of time talking/working with Joshua as he was developing his A & M Tunes and know that if he an Shawn do get this done it will be worth the wait. To me what could make this great is to integrate it with the M-tune, so you select a different tune and you get a different boost map, too. That's like the holy grail of engine management, for my moderate use, and it looks like it is in site!
