IPD Plenum - best performance mod?

[maytag]

Originally Posted by 03boxster-S

Hello, new member here. I can't find any detailed instructions for the IPD plenum install on a 986, only for the 987? I think I'm missing something, watched all the videos on YouTube and my car won't start.

Wait..... you followed all the YouTube videos, and it isn't working?

Color me shocked! Haha

Ok. Tell us what you've done, where you're at in the project, and what're the symptoms.

Sent from my SM-G970U using Tapatalk

[SpIcEz]

Btw, this is another option to the IDP plenum.

https://uniwerksdesign.com/product/porsche-986-carbon-fiber-intake-plenum/

Build thread found here on the forum.

[maytag]

Originally Posted by SpIcEz

Btw, this is another option to the IDP plenum.

https://uniwerksdesign.com/product/porsche-986-carbon-fiber-intake-plenum/

Build thread found here on the forum.

And an excellent option, at that.

Sent from my SM-G970U using Tapatalk

[michaelpetersen3]

Just posted a YouTube video highlighting the dyno results of the FabSpeed air intake coupled with the IPD Plenum and larger throttle body on a Boxster S, chassis 986.

Wanted to share it in hopes that it helps the next person looking for the information… (Hint, it didn’t add any horsepower overall, in fact the average horsepower between 3500k-7000k rpm was reduced).

Here’s the link https://youtu.be/zRUHnGPJC0Q

[husker boxster]

Nice video MP. Now we have some actual numbers vs all the butt dynos.

In some regards I'm not surprised by your results. I have an IPD on my CSS and my butt dyno didn't feel any HP advantage. I did feel some mild mid-range grunt, which your #s show. You don't mention doing any software tuning in your experiment and I don't have a tune either. I think that's why the HP #s look like they do. With no tune, you're just adding air to the equation. The fuel injectors are still performing at their previous levels. I'm not an expert and I didn't stay at a Holiday Inn, but I would think diluting the fuel / air mixture with addl air explains the loss of HP. The factory mixture Porsche chose would hopefully be the best ratio possible. With a tune, you could get the fuel side richened to match the extra air and produce a more explosive cocktale. Somebody with more knowledge than my guesses can chime in.

When my engine rebuild is complete, the shop that pulled my engine is going to do a tune and dyno pull on it to maximize the setup after a proper break in session. I'm looking fwd to that (hopefully sometime this month).

Sure don't miss the Kram days.

[blue62]

Originally Posted by husker boxster

Nice video MP. Now we have some actual numbers vs all the butt dynos.

In some regards I'm not surprised by your results. I have an IPD on my CSS and my butt dyno didn't feel any HP advantage. I did feel some mild mid-range grunt, which your #s show. You don't mention doing any software tuning in your experiment and I don't have a tune either. I think that's why the HP #s look like they do. With no tune, you're just adding air to the equation. The fuel injectors are still performing at their previous levels. I'm not an expert and I didn't stay at a Holiday Inn, but I would think diluting the fuel / air mixture with addl air explains the loss of HP. The factory mixture Porsche chose would hopefully be the best ratio possible. With a tune, you could get the fuel side richened to match the extra air and produce a more explosive cocktale. Somebody with more knowledge than my guesses can chime in.

When my engine rebuild is complete, the shop that pulled my engine is going to do a tune and dyno pull on it to maximize the setup after a proper break in session. I'm looking fwd to that (hopefully sometime this month).

Sure don't miss the Kram days.

As to the no tune:
You are not just adding air.
The DME/ECU adjusts the fuel/air mixture for each individual cylinder on each individual intake stroke. So at higher RPM that is hundreds of times per second.
If the car is not throwing codes for having reached it's enrichment limit then short term and learned (long term) fuel strategies are adjusting properly for the additional air supplied by the IPD.

[michaelpetersen3]

Great discussion guys, a couple of thoughts to enrich the conversation. For background, early in my career I was a powertrain engineer at Ford Motor Company, so I have enough tuning experience to have an educated point of you, but I do not know the specifics of how Porsche calibrated the 986/996, and the logic on their program. I’m also posting a copy of the dyno runs that are shown in the video, for easier view ability.

First, the car does have a “tune” from Fabspeed. They built the tune for the higher flow exhaust system, but it has not been updated for the intake system. Even the first tune from Fabspeed took several iterations for them to find additional horsepower, and I did not settle on it until we verified it on the dyno. They were able to extract an additional 5 hp from the tune but no more. Remember, if you are tuning a turbo engine, you can increase the boost and find significant gains in horsepower. On the other hand, if you are tuning naturally aspirated you can only adjust the spark table and the air fuel ratio. Nothing more.

1) Typically, the spark table is advanced to the point where the knock sensors detect early pre-detonation and then the spark is backed off slightly. I am assuming that is what is happening in the Porsche calibration but I am not certain. There is no easy way to measure this on a typical Dyno set up.
2) The other variable is the air fuel ratio. When an engine is operating under normal conditions the O2 sensors are used to keep the air fuel ratio at the stoichiometric ratio, 14.7:1. This is done primarily for emissions, but it also aids in fuel economy. This condition is considered a “closed loop” operation meaning the ECU is constantly adjusting the fuel flow based on the signals from the O2 sensors to keep the engine operating at this ideal stoichiometric condition. However, when you move to wide-open throttle (WOT), the engine typically goes into a open loop parameter, whereas the fuel flow rate is simply a mathematical relationship (called a transfer function) where the ECU looks at the airflow meter in the intake system and calculates the right amount of fuel to inject into the cylinders. In some cases, these transfer functions are fixed, and can be adjusted by the tuner, in other situations these transfer functions will be automatically modified overtime based on the performance of the engine, this is not managed by the tuner but managed by the logic within the ECU. It is my assumption that given the age of this car, that it is a fixed transfer function. You will note at the bottom of the chart that the air fuel ratio is monitored during the runs through the output of the exhaust system through a wideband sensor. So, we know if the tune is generally correct for the engine, or a further programming needs to be done. In general, a leaner engine (ie higher A/F ratio) delivers the best horsepower. However, in reality, programmers need to richen the ratio (lower A/F ratios) to cool down the cylinder and prevent the engine from pre-detonation. At these WOT limits it’s a tradeoff between spark advance, and air fuel ratio, to uncover the best overall performance. Keeping in mind that the knock detectors can automatically pull spark out at any point if it detects pre-detonation. Cutting through all of the gibberish, if you look at the chart the car is running near an AF ratio of 12.0 for all three runs. The exhaust-only tune is running slightly leaner by 0.5, but that small delta is likely only making a few hp difference, at most.

This means… that a custom tune might be able to find a few horsepower at most, but no significant gains. I did inquire on the cost/timing of a new tune from Fabspeed, but learned that many tuners are now getting out of the business as the EPA is putting pressure on larger aftermarket shops as these tunes have not been tested by the EPA and are therefore not compliant. I think we will see fewer tuning shops in existence in the future. Once a shop starts to pick up business momentum the EPA will come in and quickly shut them down. Furthermore, to properly tune a car you want to bring it to an shop that has a dyno and can make fine-tuning adjustments on the car based on multiple, often dozens, of dyno pulls to optimized all of the variables. Keep in mind every car is slightly different due to the manufacturing tolerances, and the best tune for a specific car is not necessarily the best tune for another car even if it is exactly the same make model and year. And, once you start changing the variables such as intake and exhaust, the only way to give it a proper tune is with 6-8 eight hours of dyno time back-and-forth with the programmer. This is not realistic for the majority of people putting together a hobby car as the car and the programmer are in two different cities and do not have the luxury of dyno testing as they optimize the tune.

All of this being said, I do not think we will see significant gains on the intake system. If somebody has better data to prove me wrong, I’m open to seeing it. As I was expecting, and even hoping, to see more significant improvements from the intake mods.

Oh, BTW.
Husker Boxster: butt dynos are historically bad. You mentioned that you thought you felt a little mid-range grunt, but I think you looked at the chart incorrectly. The dyno shows that with the intake system it actually lost up to 20hp in the mid-range.

Blue 62: the car does not adjust each individual cylinder; it looks at the left bank and the right bank and treats each bank as a family. It does not have information to adjust an individual cylinder. And while you could say that it optimizers every stroke, in reality it adjust the parameters more gradually, and may take 30 or 100 strokes to bring the spark or air fuel ratio back in line with the program. If you look at some of my commentary above, at wide open throttle it is likely an closed loop program and is doing nothing more than dumping in a predefined amount of fuel based on the transfer function, running spark off of the spark table, and pulling spark out if it detects detonation on either bank of cylinders.

There may be some Porsche tuners in the community that can add some further detail to how Porsche manages their programming on their ECU. If you have first-hand experience here, and can add some color to the above, I would welcome the insights, as some of my insights are based on Fords programming, and it may be slightly different for Porsche. But I doubt it will be much different as every engine is still bound by the same laws of physics!

[blue62]

Originally Posted by michaelpetersen3

Great discussion guys, a couple of thoughts to enrich the conversation. For background, early in my career I was a powertrain engineer at Ford Motor Company, so I have enough tuning experience to have an educated point of you, but I do not know the specifics of how Porsche calibrated the 986/996, and the logic on their program. I’m also posting a copy of the dyno runs that are shown in the video, for easier view ability.

First, the car does have a “tune” from Fabspeed. They built the tune for the higher flow exhaust system, but it has not been updated for the intake system. Even the first tune from Fabspeed took several iterations for them to find additional horsepower, and I did not settle on it until we verified it on the dyno. They were able to extract an additional 5 hp from the tune but no more. Remember, if you are tuning a turbo engine, you can increase the boost and find significant gains in horsepower. On the other hand, if you are tuning naturally aspirated you can only adjust the spark table and the air fuel ratio. Nothing more.

1) Typically, the spark table is advanced to the point where the knock sensors detect early pre-detonation and then the spark is backed off slightly. I am assuming that is what is happening in the Porsche calibration but I am not certain. There is no easy way to measure this on a typical Dyno set up.
2) The other variable is the air fuel ratio. When an engine is operating under normal conditions the O2 sensors are used to keep the air fuel ratio at the stoichiometric ratio, 14.7:1. This is done primarily for emissions, but it also aids in fuel economy. This condition is considered a “closed loop” operation meaning the ECU is constantly adjusting the fuel flow based on the signals from the O2 sensors to keep the engine operating at this ideal stoichiometric condition. However, when you move to wide-open throttle (WOT), the engine typically goes into a closed loop parameter, whereas the fuel flow rate is simply a mathematical relationship (called a transfer function) where the ECU looks at the airflow meter in the intake system and calculates the right amount of fuel to inject into the cylinders. In some cases, these transfer functions are fixed, and can be adjusted by the tuner, in other situations these transfer functions will be automatically modified overtime based on the performance of the engine, this is not managed by the tuner but managed by the logic within the ECU. It is my assumption that given the age of this car, that it is a fixed transfer function. You will note at the bottom of the chart that the air fuel ratio is monitored during the runs through the output of the exhaust system through a wideband sensor. So, we know if the tune is generally correct for the engine, or a further programming needs to be done. In general, a leaner engine (ie higher A/F ratio) delivers the best horsepower. However, in reality, programmers need to richen the ratio (lower A/F ratios) to cool down the cylinder and prevent the engine from pre-detonation. At these WOT limits it’s a tradeoff between spark advance, and air fuel ratio, to uncover the best overall performance. Keeping in mind that the knock detectors can automatically pull spark out at any point if it detects pre-detonation. Cutting through all of the gibberish, if you look at the chart the car is running near an AF ratio of 12.0 for all three runs. The exhaust-only tune is running slightly leaner by 0.5, but that small delta is likely only making a few hp difference, at most.

This means… that a custom tune might be able to find a few horsepower at most, but no significant gains. I did inquire on the cost/timing of a new tune from Fabspeed, but learned that many tuners are now getting out of the business as the EPA is putting pressure on larger aftermarket shops as these tunes have not been tested by the EPA and are therefore not compliant. I think we will see fewer tuning shops in existence in the future. Once a shop starts to pick up business momentum the EPA will come in and quickly shut them down. Furthermore, to properly tune a car you want to bring it to an shop that has a dyno and can make fine-tuning adjustments on the car based on multiple, often dozens, of dyno pulls to optimized all of the variables. Keep in mind every car is slightly different due to the manufacturing tolerances, and the best tune for a specific car is not necessarily the best tune for another car even if it is exactly the same make model and year. And, once you start changing the variables such as intake and exhaust, the only way to give it a proper tune is with 6-8 eight hours of dyno time back-and-forth with the programmer. This is not realistic for the majority of people putting together a hobby car as the car and the programmer are in two different cities and do not have the luxury of dyno testing as they optimize the tune.

All of this being said, I do not think we will see significant gains on the intake system. If somebody has better data to prove me wrong, I’m open to seeing it. As I was expecting, and even hoping, to see more significant improvements from the intake mods.

Oh, BTW.
Husker Boxster: butt dynos are historically bad. You mentioned that you thought you felt a little mid-range grunt, but I think you looked at the chart incorrectly. The dyno shows that with the intake system it actually lost up to 20hp in the mid-range.

Blue 62: the car does not adjust each individual cylinder; it looks at the left bank and the right bank and treats each bank as a family. It does not have information to adjust an individual cylinder. And while you could say that it optimizers every stroke, in reality it adjust the parameters more gradually, and may take 30 or 100 strokes to bring the spark or air fuel ratio back in line with the program. If you look at some of my commentary above, at wide open throttle it is likely an closed loop program and is doing nothing more than dumping in a predefined amount of fuel based on the transfer function, running spark off of the spark table, and pulling spark out if it detects detonation on either bank of cylinders.

There may be some Porsche tuners in the community that can add some further detail to how Porsche manages their programming on their ECU. If you have first-hand experience here, and can add some color to the above, I would welcome the insights, as some of my insights are based on Fords programming, and it may be slightly different for Porsche. But I doubt it will be much different as every engine is still bound by the same laws of physics!

michaelpetersen3:

Thank you for the feedback on my comments.
Sometimes I over simplify things and sometimes I over think things.
After reading your feedback and a moments thought
Yes I see the the system adjusts fuel/air ratio on a per bank basis.
It would need an O2 sensor for each cylinder (exhaust runner) to have the info to adjust fuel/air ratio on a per cylinder basis. (at the least.)

husker boxster:

Sorry about my miss-information in post #46.

[husker boxster]

Originally Posted by blue62

husker boxster:

Sorry about my miss-information in post #46.

Hey, no problem. My diagnosis wasn't very accurate. We're here to learn.

[azlvr]

Would a larger motor have more benefits from this intake say 3.6 or 3.8? I have both a 74mm and a 81mm waiting for when I build a 3.8.

[michaelpetersen3]

Husker / Blue - All good. We're all in the learning mode, thx for your comments ;-)

Azlvr, the larger TB on the larger motor is directionally correct to improve HP. But it might not make any improvements if the TB is not in the critical path of the air flow resistance. You can test this either on the dyno, or on a flow bench (less common but cheaper way to test intakes).

If you have both TB's, use the larger 81mm, but don't expect much, and don't spend big $ to incorporate it. My 3.2L tests suggest you will find the most benefits from the exhaust flow improvements.

[PaulE]

I'm enjoying following this thread. As some additional annecodatal info, my shop generally advises against the IPD Plenum and larger throttle bodies on the M96/M97 Porsches. They have a dyno and have tested lots of cars before and after swapping plenums and throttle bodies. They say they haven't seen increases in power from the larger throttle bodies and plenums, and in some cases, have actually seen decreases in power. They advised me against adding the IPD plenum and larger throttle body, both when my Boxster S was a stock 3.2, and again when they rebuilt it as a 3.6 with a custom tune.

They use a Dynapack dyno, which doesn't have rollers, the wheels are taken off and the dyno is bolted directly to the driven wheel hubs. My 3.6 produced 254.7 HP at the hubs on that. I think I could get more HP with a more open exhaust, i.e. catless headers and losing the stock muffler, but I'm happy where I am, and the dyno time costs money!

[ike84]

Michael,

Very nice writeup about the tuning. I can't stand when people talk about all the power you get out of custom tunes, and how tunes are written for specific "parts" like catless headers. There may be some room for advancing spark in this motor but as soon as you knock the dme will pull the timing. You can alter the AFR target in closed loop operations, as it looks like yours has been, but your gonna pay for it one way or the other (either with more heat or more fuel).

A huge barrier to tuning is the ME 7.2. Even on the nefmoto site, which is probably the internet's best group of euro tuning junkies, the consensus is to not try to mess with it. Too complicated, no aftermarket support, no technical education. All guess work, and lots of risk. FVD claims that their tuners are trained in ME 7.2 but amazingly enough I could never get one of these guys on the phone.

Im surprised that you managed to get any bump out of a tune with fabspeed. I spoke with them directly about developing a tune last fall and let's just say I was fairly convinced they didn't know they first thing about tuning.

There is another thread where Pedro's plenum was dyno test and showed nearly identical number to yours. The theory with both of these plenums is that the Y shape helps to improve laminar flow around the split, but clearly this is not the case.

There is always the interesting war between CAI and short ram fans. One thing that I think is true that some engines do better with one over the other, and I think that the M96 may like short rams.

The reason I say this is because of my fuel trim data on my base 986. I run a 996 tune with a 996 plenum, 82mm TB, and short ram 3.5" dual filter intake that runs straight into my trunk. It's about as short and fat as you can make it.

My long term fuel trims are -17% on both banks. What's interesting here is that a 2.7L is 20.6% smaller than the 3.4. But my setup (which in closed loop runs at 14.7+-.2 at all times, I have a WB sensor installed) is managing to pull more air. By my calculations, I'm getting 4%ish more air than what I should proportional to displacement. And that's actually with my intake temps running hot because of the heat in the trunk, about 100F. I have no idea if this has translated into more power but I found it interesting.


Sent from my SM-G970U1 using Tapatalk

[michaelpetersen3]

Hi Paul, I really like the idea of the direct connect dyno. It takes one more variable (wheel rolling resistance) out of the equation. Should give you more consistent results. Im pulling 79HP/L on my 3.2. You should be able to pull 285HP from your 3.6 w/ exhaust work, maybe 290/295HP if you eliminate rolling resistance. BTW, if you want to see the details on my exhaust mods and dyno results, I posted a vid several years ago here: https://youtu.be/MdCQg2tJFYg.

Hi Ike, Sounds like you have more depth on the Porsche tune than I. We ran the car back to back with the stock tune vs Fabspeed. It was an honest 4-5 HP. Not huge, but honest. Looked like the achieved it through a leaner A/F ratio. That said, when I spoke to Fabspeed they clearly outsourced their tune, but I do not know to whom. Fabspeed acted as the intermediary. It took them 2 tries to find the few HP, but they did. As I mentioned, they are no longer offering tunes as the EPA shut them down. Disappointing, as the number of tuned cars is 0.001% of the overall car population, and most of these cars see little street time. They are chasing mice nuts IMO. Government agencies finding more reach than value (I'll stop my political rant now, sorry).

[azlvr]

Originally Posted by michaelpetersen3

Husker / Blue - All good. We're all in the learning mode, thx for your comments ;-)

Azlvr, the larger TB on the larger motor is directionally correct to improve HP. But it might not make any improvements if the TB is not in the critical path of the air flow resistance. You can test this either on the dyno, or on a flow bench (less common but cheaper way to test intakes).

If you have both TB's, use the larger 81mm, but don't expect much, and don't spend big $ to incorporate it. My 3.2L tests suggest you will find the most benefits from the exhaust flow improvements.

I just installed the Fabspeed Maxflo cat back system and plan to go with their headers before bigger motor. I'll have 987 airbox and tubing with 81mm going into it.

[ike84]

Originally Posted by michaelpetersen3

Hi Paul, I really like the idea of the direct connect dyno. It takes one more variable (wheel rolling resistance) out of the equation. Should give you more consistent results. Im pulling 79HP/L on my 3.2. You should be able to pull 285HP from your 3.6 w/ exhaust work, maybe 290/295HP if you eliminate rolling resistance. BTW, if you want to see the details on my exhaust mods and dyno results, I posted a vid several years ago here: https://youtu.be/MdCQg2tJFYg.

Hi Ike, Sounds like you have more depth on the Porsche tune than I. We ran the car back to back with the stock tune vs Fabspeed. It was an honest 4-5 HP. Not huge, but honest. Looked like the achieved it through a leaner A/F ratio. That said, when I spoke to Fabspeed they clearly outsourced their tune, but I do not know to whom. Fabspeed acted as the intermediary. It took them 2 tries to find the few HP, but they did. As I mentioned, they are no longer offering tunes as the EPA shut them down. Disappointing, as the number of tuned cars is 0.001% of the overall car population, and most of these cars see little street time. They are chasing mice nuts IMO. Government agencies finding more reach than value (I'll stop my political rant now, sorry).

I'm 100% on board with your rant. ****************, you cant even buy a new tractor without a def system on it. I recently stripped all my emissions gear on my Duramax 3500 - not because I ever roll coal, but because the rules are so stringent that when a part breaks it puts the truck into limp mode with permanent dtcs that can't be cleared without a gm reader.

Our forum has a lot of super knowledge guys and gals with tons of experience, but the thing we don't have among us are engine builders. They just don't really thrive in the Porsche world. I'm certainly not one, but I can tell you this - there are only two ways to make REAL power gains on a NA motor - built engine with lighter/stronger components and more aggressive cams OR conversion to forced induction. Unfortunately there aren't great kits for either of these available for the M96 motors. The x51 option was a "built" version of these motors but was never offered as an aftermarket kit so that's out the window. TPC made the SC and now the TC kits for these cars but I wouldn't trust them. So, people are left with intake/exhaust mods or engine swaps.

I will say one thing though, twin mitsu td06h are perfectly sized to run a low boost setup on the 2.7s good for 300 hp

Sent from my SM-G970U1 using Tapatalk

[edc]

It good to see some actual independent data on these mods. I went the 'value' route using the 996 plenum doing base run and then subsequent runs with the exhaust and intake mods. Each step was small an incremental. Ok, with the and money it could have been done better ie hub Dyno, all on the same day etc etc. But I did it for myself not anyone else so satisfied myself with at least using the same Dyno, and same operator.

[ike84]

I wrote up the process in the DIY section, titled 82mm TB. Lots of pics there. There are two vents in the rear sidewalls of the trunk that are normally for venting heat out. I punched out those vents so it can pull air from under the car. Eventually I'm either going to remove the rear OEM spoiler and cover that opening with a louvered cover, or drill a big cutout in the trunk lid and cover it with a scoop. Life's been pretty busy so I've put it in hold for now though.

Sent from my SM-G970U1 using Tapatalk

[michaelpetersen3]

Hi Ike,

Im onboard with everything you say. When you finish your intake project, send pics and results. Very interested to see!

[ike84]

Here it is as it currently exists - 996 plenum, 82mm TB, 45 deg down, 45 deg up, 3.25" maf housing (not in pic but under rear engine compartment lip just before firewall) through the firewall, and then to the big boi. The last pic shows those vents knocked out.

I am very strongly leaning toward a twin turbo install this winter, so I probably won't make any more mods until I do that. I will say though that my aftermarket rear wing works beautifully so the OEM spoiler can go without problem at this point.

Sent from my SM-G970U1 using Tapatalk