BPM vs. ESS

[M3takesNYC]

Originally Posted by img

Your entire premise seems built upon the assumption of this equality. But what if they aren't equal? Have you thought about that?



I made a very similar suggestion. Not only would they get results that others could analyze (the vBox files themselves), but you could also re-animate the race just like this video below.

You see, a buddy already did a vBox test to see how small differences in shift speed affected real world performance. So he took a DCT car into the middle of the desert late at night on a completely empty road. He ran the car 12 times. Six times with DCT in D2 mode, and six times with DCT in D5 mode. The runs were made at exactly the same spot in the road each time, in the exact same direction each time, and only minutes apart.

So I thought about this and asked the guy if he would loan me the vBox files and make a video that would show what happens if the same exact car in DCT-D2 raced himself in DCT-D5. I think the results shocked all of us. Take a look.

So watch the video then ask yourself this question: How much difference will different cars, different weights, different wheels, different mods, and different drivers make? Watch the video and then tell me what you think.

This is pretty silly as why would anyone be going all out in D2 mode? It shifts more slowly so yes a car would be slower. Pretty silly comparison. Both people in D5, stock cars equal (if they are not equal than obviously we won't get to the tuned test so of course we assume this premise right now) and mash the gas. Who needs the vbox analysis.

99 percent of people want to see a rung what ya brung-head to head, out of the box tunes on two equally fast cars that are proved to be equally fast.

This is getting ridiculous the nitpicking. People who are not interested in the race as its being described maybe should build their own test and start a new thread

[STi_traitor]

Originally Posted by img

Your entire premise seems built upon the assumption of this equality. But what if they aren't equal? Have you thought about that?



I made a very similar suggestion. Not only would they get results that others could analyze (the vBox files themselves), but you could also re-animate the race just like this video below.

You see, a buddy already did a vBox test to see how small differences in shift speed affected real world performance. So he took a DCT car into the middle of the desert late at night on a completely empty road. He ran the car 12 times. Six times with DCT in D2 mode, and six times with DCT in D5 mode. The runs were made at exactly the same spot in the road each time, in the exact same direction each time, and only minutes apart.

So I thought about this and asked the guy if he would loan me the vBox files and make a video that would show what happens if the same exact car in DCT-D2 raced himself in DCT-D5. I think the results shocked all of us. Take a look.

So watch the video then ask yourself this question: How much difference will different cars, different weights, different wheels, different mods, and different drivers make? Watch the video and then tell me what you think.

Yes I have thought about that and if they are not equal than so be it. I still had fun driving my car 130MPH+++ legally!

Cool video bro!

This isn't my first time comparing two tuned cars with similar mods. My old friend and I had the same car (STi) with the exact same mod and tuner, we were tuned on the same day after finalizing near exact modifications to our cars. After the tuning the cars never left each others side. Prior to that I had the most the modifications and he had just a few of them, I took him. This is something I will base my judgement off because I have witnessed and experienced it myself, a real before and after of what a tune did for him. There is nothing wrong with comparing cars this way, it has been done for years. Years before your vbox was established. Lets just say I am old fashioned.

Originally Posted by M3takesNYC

This is pretty silly as why would anyone be going all out in D2 mode? It shifts more slowly so yes a car would be slower. Pretty silly comparison. Both people in D5, stock cars equal (if they are not equal than obviously we won't get to the tuned test so of course we assume this premise right now) and mash the gas. Who needs the vbox analysis.

99 percent of people want to see a rung what ya brung-head to head, out of the box tunes on two equally fast cars that are proved to be equally fast.

This is getting ridiculous the nitpicking. People who are not interested in the race as its being described maybe should build their own test and start a new thread

+1,000,000

Still enjoyable to hear everybody's comments!

[BMRLVR]

Originally Posted by M3takesNYC

This is pretty silly as why would anyone be going all out in D2 mode? It shifts more slowly so yes a car would be slower. Pretty silly comparison. Both people in D5, stock cars equal (if they are not equal than obviously we won't get to the tuned test so of course we assume this premise right now) and mash the gas. Who needs the vbox analysis.

99 percent of people want to see a rung what ya brung-head to head, out of the box tunes on two equally fast cars that are proved to be equally fast.

This is getting ridiculous the nitpicking. People who are not interested in the race as its being described maybe should build their own test and start a new thread

The only one being silly here is you!!! img is not saying anyone will be starting in D2, this was just used to illustrate how something so insignificant such as 120ms difference in shift speeds can greatly affect the final out come. One of the main reasons for showing this is that the two forum members who are doing this "test" were planning on shifting themselves which can potentially add a lot more than 120ms. This would make the results skewed right off the get go........ Couple this with different chassis, different supporting mods etc. and it makes the entire premises of this test worthless....... That is what myself, img, swamp, Porschefile were pointing out.

You see this was posted in the technical area of the forum and some people like to keep it just that. I for one am not trying to stir up shit but rather my interest is in the betterment of this community by sharing my knowledge (and the others I mentioned are only doing the same) so that others can learn and benefit from it. This whole thing will be great fun I am sure, and I would love to be able to be there with them just for the fun of it. All any of us wanted to do was assist the OP's in this thread by suggesting ways to eliminate some of the many variables that were inevitably going to exist in this "Test" since the OP's started off the thread implying that it would be just that, a test.

So I am sure that the guys are going to have fun doing this, I know I would. If the OP's are actually interested in getting some concrete and tangible info out of this, there are people on here that would love to help with setting out some parameters to make the info a lot more valid. Even if it can't be 100% accurate since 2 different cars are being used, at least the info could be recorded in a way that something useful could be taken out of it......... I thought that is what the OP's were after, tangible, useful info on the tunes in question!?

[STi_traitor]

Originally Posted by BMRLVR

The only one being silly here is you!!! img is not saying anyone will be starting in D2, this was just used to illustrate how something so insignificant such as 120ms difference in shift speeds can greatly affect the final out come. One of the main reasons for showing this is that the two forum members who are doing this "test" were planning on shifting themselves which can potentially add a lot more than 120ms. This would make the results skewed right off the get go........ Couple this with different chassis, different supporting mods etc. and it makes the entire premises of this test worthless....... That is what myself, img, swamp, Porschefile were pointing out.

You see this was posted in the technical area of the forum and some people like to keep it just that. I for one am not trying to stir up shit but rather my interest is in the betterment of this community by sharing my knowledge (and the others I mentioned are only doing the same) so that others can learn and benefit from it. This whole thing will be great fun I am sure, and I would love to be able to be there with them just for the fun of it. All any of us wanted to do was assist the OP's in this thread by suggesting ways to eliminate some of the many variables that were inevitably going to exist in this "Test" since the OP's started off the thread implying that it would be just that, a test.

So I am sure that the guys are going to have fun doing this, I know I would. If the OP's are actually interested in getting some concrete and tangible info out of this, there are people on here that would love to help with setting out some parameters to make the info a lot more valid. Even if it can't be 100% accurate since 2 different cars are being used, at least the info could be recorded in a way that something useful could be taken out of it......... I thought that is what the OP's were after, tangible, useful info on the tunes in question!?

The only way this will be valid for you, img, swamp and Porschefile is by using the vbox. I am not purchasing the vbox for this. UBER and I will fine tune our variable of 120ms.......I mean holy sh*t, we are debating about 120ms now.

[UBER ///M3]

Thanks for all the suggestions we both are taking them into account the main thing is lets see how it goes, do our runs, and if the forum/mass (including us) think we can better it or something was that far off then the vbox idea will probably be used...No ones ideas are being bashed or cut down non of us are professionals we just share the same passion and are learning along the way

[CanAutM3]

Originally Posted by STi_traitor

The only way this will be valid for you, img, swamp and Porschefile is by using the vbox. I am not purchasing the vbox for this. UBER and I will fine tune our variable of 120ms.......I mean holy sh*t, we are debating about 120ms now.

120ms at 130mph equates to 23 feet. Not negligible...

[CanAutM3]

Originally Posted by img

Your entire premise seems built upon the assumption of this equality. But what if they aren't equal? Have you thought about that?



I made a very similar suggestion. Not only would they get results that others could analyze (the vBox files themselves), but you could also re-animate the race just like this video below.

You see, a buddy already did a vBox test to see how small differences in shift speed affected real world performance. So he took a DCT car into the middle of the desert late at night on a completely empty road. He ran the car 12 times. Six times with DCT in D2 mode, and six times with DCT in D5 mode. The runs were made at exactly the same spot in the road each time, in the exact same direction each time, and only minutes apart.

So I thought about this and asked the guy if he would loan me the vBox files and make a video that would show what happens if the same exact car in DCT-D2 raced himself in DCT-D5. I think the results shocked all of us. Take a look.

So watch the video then ask yourself this question: How much difference will different cars, different weights, different wheels, different mods, and different drivers make? Watch the video and then tell me what you think.

Thanks for sharing the video again. I guess missed it the first time. It was very interesting to watch. Don't want to go OT, but it would be great to compare the various "S" modes. I am curious to find out if there are acceleration benefits at using S6.

[BMRLVR]

Originally Posted by STi_traitor

The only way this will be valid for you, img, swamp and Porschefile is by using the vbox. I am not purchasing the vbox for this. UBER and I will fine tune our variable of 120ms.......I mean holy sh*t, we are debating about 120ms now.

Not debating, just using it to prove how something as insignificant as 120ms makes a big difference as speeds climb. You don't have to purchase a vbox but if you intend to do this test of the software, you both need to do at least these things for both stock and performance software runs: 1)The same wheels and same tire sizes on both cars, 2)use the same DCT mode (run in D mode not S mode to eliminate driver error), 3)fill up with the same fuel type (brand and octane rating) and burn a few tanks of it before the test to allow the DME to calibrate for timing, 4)*Optional*clear adaptations if you can get access to a scan tool with the ability to do it.

Yes a vbox would be a nice addition to the test however no one expects you to purchase one if you don't want to.......maybe someone will loan you guys one if someone on the forums has one in that area.

One other thing that would be nice would be to weigh the cars with drivers and ballast the lighter one to the same weight as the heavier one...... I realize that having access to a scale to weigh the cars may be hard to do.

All of these suggestions are only to allow the "TUNES" to be compared as equally as you guys have the resources to do so.

Originally Posted by UBER ///M3

Thanks for all the suggestions we both are taking them into account the main thing is lets see how it goes, do our runs, and if the forum/mass (including us) think we can better it or something was that far off then the vbox idea will probably be used...No ones ideas are being bashed or cut down non of us are professionals we just share the same passion and are learning along the way

Thanks for realizing that the comments were not personal attacks but rather constructive criticism/FYI kinds of comments. Have fun and be safe!

Originally Posted by CanAutM3

120ms at 130mph equates to 23 feet. Not negligeable...

You always add useful info to the forum and I enjoy your professionalism.

Originally Posted by Team Plutonium

This is a technical forum, and members share their technical expertise. If that advice is dismissed... the OP should've posted this in the off-topic section. Or create a title that says: "I'm gonna race my buddy's car and see who is faster". In any case, I'm sure you'll have fun. Be safe.

+1000000 Phillip!

[M3takesNYC]

With all due respect as I know you are an expert in engines and you do that for a living, there are a few things which are just plain wrong.

1. Why would wheels and tires matter if a stock baseline is established where equality is established? I agree if no stock runs were established than this would be a bare minimum but for whatever reason, one car may be a bit faster than the other but maybe its heavier wheels will equalize it to the slightly slower car with stock wheels or vice versa (just an example). The bottom line is if they can establish equal stock runs than all of that stuff does not matter.

If during a stock run, one car is faster than the other than they need to equalize the stock runs. Maybe adding a passenger in the faster car or adding weight etc.

But once stock is equality is determined than none of those variables matter and there is absoolutely no need for someone to change their wheels and tires. You need to start looking at the end metric here which is at these speeds whether the two cars run side-by side. "how" they do it is irrelevent in terms of one may be heavier than the other. You are getting caught in the fine details and missing the big picture.

2. Running through a few tanks of gas to have them adapt? The cars fully adapt after a couple pulls of full throttle. That makes no sense that it would take even one full tank for an advanced ionic knock system to finally adapt to timing. it would defeat the purpose. Its able to adapt within a run or two to maximize timing and maximize engine output for all variables. Adaptations happen within pulls, they don't happen over tanks of gas.

Again you are creating problems here that complicate this but don't even contribute to being more accurate.

The only important thing is the cars having the same mods in terms of response to tuning. Catless setup really is the only one that matters (or HFC) which both have. That really is the only factor that a tune may respond differently to. Besides octane which I agree is a huge factor as hitting the timing targets of a tune with a car down on octane would potentially prevent the car from getting as much out of a tune as a car with higher octane. So catless setup and equal octane is the only two factors in addition to equal stock runs that matters. You could put 5 pound wheels on one car with 2 passengers and it may be equal to a stock wheel car with no passengers and as long as you are only changing the ECU/tune variable given same octane nad catless, than it does not change the metric of whether the tune is adding x amount of hp

[luis_m3]

Originally Posted by UBER ///M3

yeah man I am and the whole ///M Flight crew are 100% and im sure STI TRAITOR will also...15,000 people and its a huge showing with lts of nice cars...come for sure youll enjoy it

I am not part of M Flight, but I'll come along for sure!

[BMRLVR]

Originally Posted by M3takesNYC

With all due respect as I know you are an expert in engines and you do that for a living, there are a few things which are just plain wrong.

1. Why would wheels and tires matter if a stock baseline is established where equality is established? I agree if no stock runs were established than this would be a bare minimum but for whatever reason, one car may be a bit faster than the other but maybe its heavier wheels will equalize it to the slightly slower car with stock wheels or vice versa (just an example). The bottom line is if they can establish equal stock runs than all of that stuff does not matter.

If during a stock run, one car is faster than the other than they need to equalize the stock runs. Maybe adding a passenger in the faster car or adding weight etc.

But once stock is equality is determined than none of those variables matter and there is absoolutely no need for someone to change their wheels and tires. You need to start looking at the end metric here which is at these speeds whether the two cars run side-by side. "how" they do it is irrelevent in terms of one may be heavier than the other. You are getting caught in the fine details and missing the big picture.

2. Running through a few tanks of gas to have them adapt? The cars fully adapt after a couple pulls of full throttle. That makes no sense that it would take even one full tank for an advanced ionic knock system to finally adapt to timing. it would defeat the purpose. Its able to adapt within a run or two to maximize timing and maximize engine output for all variables. Adaptations happen within pulls, they don't happen over tanks of gas.

Again you are creating problems here that complicate this but don't even contribute to being more accurate.

The only important thing is the cars having the same mods in terms of response to tuning. Catless setup really is the only one that matters (or HFC) which both have. That really is the only factor that a tune may respond differently to. Besides octane which I agree is a huge factor as hitting the timing targets of a tune with a car down on octane would potentially prevent the car from getting as much out of a tune as a car with higher octane. So catless setup and equal octane is the only two factors in addition to equal stock runs that matters. You could put 5 pound wheels on one car with 2 passengers and it may be equal to a stock wheel car with no passengers and as long as you are only changing the ECU/tune variable given same octane nad catless, than it does not change the metric of whether the tune is adding x amount of hp

1) centrifugal force and inertia are the reasons why the wheels and tires need to be the same size and weight..... Remember this was a software test and not a test of which car is faster! Centrifugal force and inertia act differently then adding weight to the chassis of the car... i.e passengers.

2) burning a couple of tanks of fuel ensures that the tanks have fuel that is not a mixture of a few different octane ratings which give a cocktail that would be impossible to determine without sampling it and testing it for the AKI or RON number.
Since these engines are so dependant on timing to make power, fuel is one of the most important factors here. That ionic knock sensing system can be your friend or your enemy depending on fuel quality. One example of the octane rating being skewed would be if any of the cars had ethanol blended fuel in them (although not as common in Europe it is still available). Ethanol has a tendency to settle to the bottom of the tank which could put the pump in pure ethanol which has an RON rating of 108.6 and an AKI of 99, pretty easy to tell what effect that would have on timing, right? So again the couple tanks of fuel are just to ensure what is in the tank as an assurance of fuel type an octane rating of the fuel in both cars.

I am not going to waste any more time discussing this with you than what is above. Advise the OP's on the proper testing method since you think you know that should be and none of the rest of us do!!!

Also, kindly overlook any of my posts as I won't reply if all you want to do is try and discredit. You haven't attempted to add any value to this thread whatsoever, you were just there to jump on someone you didn't agree with. I have already established a reputation on this forum and I hold a government issued license that is recognized world wide saying that i am qualified to work on engines, not to mention hundreds of hours of technical training and tens of thousands of hours of experience. Fortunately I don't need your blessing to uphold either my licence or my reputation.

To the OP's I will reply to any questions and I hope this ends up being a fun and informative process for you guys.

*NOTE: I wrote this post on my iPhone and apologize in advance of any grammatical or spelling errors in. I will clean it up when I get in front of a computer.

[swamp2]

Originally Posted by img

You see, a buddy already did a vBox test to see how small differences in shift speed affected real world performance. So he took a DCT car into the middle of the desert late at night on a completely empty road. He ran the car 12 times. Six times with DCT in D2 mode, and six times with DCT in D5 mode. The runs were made at exactly the same spot in the road each time, in the exact same direction each time, and only minutes apart.
...
So I thought about this and asked the guy if he would loan me the vBox files and make a video that would show what happens if the same exact car in DCT-D2 raced himself in DCT-D5. I think the results shocked all of us. Take a look.

Off topic... but

Absolutely fantastic. The program itself and this particular application. It causes me to recall an old and heavily contested thread here long ago where multiple individuals claimed that D2 is measurably faster than higher D modes. As you might guess I argues against that. I would say that this settles that argument once and for all and provides us some very nice insights into Drivelogic modes and how they affect performance. I believe I also recall many claims that Drivelogic modes will overall not make a significant effect on real world performance. I think this settles that as well. One additional test I'd love to see is S5 with tracton control off vs. S5 with traction control on. My contention has always been that the power shift effect, wheel spin and additional flywheel angular momentum being dumped into the drive train simply MUST give this mode a measurable advantage.

[M3takesNYC]

Originally Posted by BMRLVR

1) centrifugal force and inertia are the reasons why the wheels and tires need to be the same size and weight..... Remember this was a software test and not a test of which car is faster! Centrifugal force and inertia act differently then adding weight to the chassis of the car... i.e passengers.

2) burning a couple of tanks of fuel ensures that the tanks have fuel that is not a mixture of a few different octane ratings which give a cocktail that would be impossible to determine without sampling it and testing it for the AKI or RON number.
Since these engines are so dependant on timing to make power, fuel is one of the most important factors here. That ionic knock sensing system can be your friend or your enemy depending on fuel quality. One example of the octane rating being skewed would be if any of the cars had ethanol blended fuel in them (although not as common in Europe it is still available). Ethanol has a tendency to settle to the bottom of the tank which could put the pump in pure ethanol which has an RON rating of 108.6 and an AKI of 99, pretty easy to tell what effect that would have on timing, right? So again the couple tanks of fuel are just to ensure what is in the tank as an assurance of fuel type an octane rating of the fuel in both cars.

I am not going to waste any more time discussing this with you than what is above. Advise the OP's on the proper testing method since you think you know that should be and none of the rest of us do!!!

Also, kindly overlook any of my posts as I won't reply if all you want to do is try and discredit. You haven't attempted to add any value to this thread whatsoever, you were just there to jump on someone you didn't agree with. I have already established a reputation on this forum and I hold a government issued license that is recognized world wide saying that i am qualified to work on engines, not to mention hundreds of hours of technical training and tens of thousands of hours of experience. Fortunately I don't need your blessing to uphold either my licence or my reputation.

To the OP's I will reply to any questions and I hope this ends up being a fun and informative process for you guys.

*NOTE: I wrote this post on my iPhone and apologize in advance of any grammatical or spelling errors in. I will clean it up when I get in front of a computer.

Its ok, you are just not addressing the points I bring up. Rotational weight or chassis weight of course matters. However not if two cars run equally. Equal is equal. How it runs equally matters not unless the reason would cause the car to respond differently to a tune which weight obviously would not. I am not saying rotational weight would not have a much bigger benefit than chassis weight on a single cars performance but when looking at compariing two, as long as they accelerate the exact same, it does not matter how their weight or setup is distributed for a straight line test..

If you both fill up at the same station, do a few pulls and both have spark plugs changed within the same relative period, than also given the stock cars pulled equally, multiple tanks is simply not at all necessary.

Not discrediting you but you also fail to actually address the point but rather give another arguement for a different situation. You seemingly can't see the point and assumption that the first step is two cars running equal on a stock ECU. I obviously agree this step needs to happen or else subsequent tests are meaningless but when and if they are equal all of that other stuff about wheels and distribution of weight is meaningless

[CanAutM3]

Originally Posted by swamp2

Off topic... but

Absolutely fantastic. The program itself and this particular application. It causes me to recall an old and heavily contested thread here long ago where multiple individuals claimed that D2 is measurably faster than higher D modes. As you might guess I argues against that. I would say that this settles that argument once and for all and provides us some very nice insights into Drivelogic modes and how they affect performance. I believe I also recall many claims that Drivelogic modes will overall not make a significant effect on real world performance. I think this settles that as well. One additional test I'd love to see is S5 with tracton control off vs. S5 with traction control on. My contention has always been that the power shift effect, wheel spin and additional flywheel angular momentum being dumped into the drive train simply MUST give this mode a measurable advantage.

+1



All we need is someone with a vbox willing to do some runs in the various "S" modes with DSC on and off.

I could do it, but would need to wait until spring when my summers are back on .

[BMRLVR]

Originally Posted by M3takesNYC

Its ok, you are just not addressing the points I bring up. Rotational weight or chassis weight of course matters. However not if two cars run equally. Equal is equal. How it runs equally matters not unless the reason would cause the car to respond differently to a tune which weight obviously would not. I am not saying rotational weight would not have a much bigger benefit than chassis weight on a single cars performance but when looking at compariing two, as long as they accelerate the exact same, it does not matter how their weight or setup is distributed for a straight line test..

If you both fill up at the same station, do a few pulls and both have spark plugs changed within the same relative period, than also given the stock cars pulled equally, multiple tanks is simply not at all necessary.

Not discrediting you but you also fail to actually address the point but rather give another arguement for a different situation. You seemingly can't see the point and assumption that the first step is two cars running equal on a stock ECU. I obviously agree this step needs to happen or else subsequent tests are meaningless but when and if they are equal all of that other stuff about wheels and distribution of weight is meaningless

I was and still am replying on my iPhone so I apologize I am not hitting the proper points. I will try and do so when I am back in front of my computer.

[m33]

Originally Posted by STi_traitor

The only way this will be valid for you, img, swamp and Porschefile is by using the vbox. I am not purchasing the vbox for this. UBER and I will fine tune our variable of 120ms.......I mean holy sh*t, we are debating about 120ms now.

I went from 6 MT to DCT for these milliseconds

[BMRLVR]

Originally Posted by M3takesNYC

Its ok, you are just not addressing the points I bring up. Rotational weight or chassis weight of course matters. However not if two cars run equally. Equal is equal. How it runs equally matters not unless the reason would cause the car to respond differently to a tune which weight obviously would not. I am not saying rotational weight would not have a much bigger benefit than chassis weight on a single cars performance but when looking at compariing two, as long as they accelerate the exact same, it does not matter how their weight or setup is distributed for a straight line test..

If you both fill up at the same station, do a few pulls and both have spark plugs changed within the same relative period, than also given the stock cars pulled equally, multiple tanks is simply not at all necessary.

Not discrediting you but you also fail to actually address the point but rather give another arguement for a different situation. You seemingly can't see the point and assumption that the first step is two cars running equal on a stock ECU. I obviously agree this step needs to happen or else subsequent tests are meaningless but when and if they are equal all of that other stuff about wheels and distribution of weight is meaningless

What you are failing to see is that any test that is accepted as valid has to have controls. Just getting a baseline by running the cars side by side and seeing if they are the same is not enough. Theoretically that would make all things equal but honestly I don't think they will be exactly equal (maybe they will though!?!?). Look at all of the different results we have seen in the dyno database from cars with the exact same mods, or from stock untouched m3's. Very seldom are the dyno numbers exactly the same even thought they should be.

You are correct in the fact that if they run exactly the same for the baseline runs then they "should" be apples and apples. I found that the effects of rotational mass appears to be independent of vehicle power after doing some further reading (equivalent mass). The one thing that different wheel/tire sizes will effect is gear ratios so it is imperative that they be the same.

Anyhow, in comparing "TUNES" one needs to eliminate all variables possible. In comparing the "CARS" to one and other that is not important. The for the software to work the (DME) ECM has to recieve inputs from the various input devices (sensors) on the vehicle and make outputs to the various actuators (cam phasers, fuel injectors, ignition coils, etc). By testing two different cars slight differences in signals from the input devices (which are inevitable and uncontrollable due to production variances) and slight differences in the output actions by the actuators (also inevitable and uncontrollable) are going to effect the outcome numbers of the software flash when installed on two different cars. Although all of these components are very precisely made, there is always an allowable range of their operation be it electrical (the actual voltage it sends to the ecm which will be effected by variances in internal resistance and the PTC or NTC bias built into some thermistors) or linear/rotational (actual amount of movement from a cam phaser for a given oil pressure, probably accurate to fractions of a degree, but still any variance will have an effect) or the amount of fuel delivered by an injector at a specified duty cycle.

If you can see all of the slight variances that can exist hopefully you can see the importance of having software tested on the same car for 100% validity...... Granted the variances are usually extremely minute but they exist none the less and this is why seemingly identical stock and modified cars never make exact numbers.

[swamp2]

Originally Posted by STi_traitor

The only way this will be valid for you, img, swamp and Porschefile is by using the vbox.

Just to be clear you are misstating this. I never said that such a test would be invalid. What would help greatly in addition to many of the other suggestions is to:

1. Swap drivers (I think that was already mentioned)
2. Repeat the testing many times both before and after the new software.

Statistics and very good at separating systematic (real tune gains) effects from random effects.

[M3takesNYC]

Originally Posted by BMRLVR

What you are failing to see is that any test that is accepted as valid has to have controls. Just getting a baseline by running the cars side by side and seeing if they are the same is not enough. Theoretically that would make all things equal but honestly I don't think they will be exactly equal (maybe they will though!?!?). Look at all of the different results we have seen in the dyno database from cars with the exact same mods, or from stock untouched m3's. Very seldom are the dyno numbers exactly the same even thought they should be.

You are correct in the fact that if they run exactly the same for the baseline runs then they "should" be apples and apples. I found that the effects of rotational mass appears to be independent of vehicle power after doing some further reading (equivalent mass). The one thing that different wheel/tire sizes will effect is gear ratios so it is imperative that they be the same.

Anyhow, in comparing "TUNES" one needs to eliminate all variables possible. In comparing the "CARS" to one and other that is not important. The for the software to work the (DME) ECM has to recieve inputs from the various input devices (sensors) on the vehicle and make outputs to the various actuators (cam phasers, fuel injectors, ignition coils, etc). By testing two different cars slight differences in signals from the input devices (which are inevitable and uncontrollable due to production variances) and slight differences in the output actions by the actuators (also inevitable and uncontrollable) are going to effect the outcome numbers of the software flash when installed on two different cars. Although all of these components are very precisely made, there is always an allowable range of their operation be it electrical (the actual voltage it sends to the ecm which will be effected by variances in internal resistance and the PTC or NTC bias built into some thermistors) or linear/rotational (actual amount of movement from a cam phaser for a given oil pressure, probably accurate to fractions of a degree, but still any variance will have an effect) or the amount of fuel delivered by an injector at a specified duty cycle.

If you can see all of the slight variances that can exist hopefully you can see the importance of having software tested on the same car for 100% validity...... Granted the variances are usually extremely minute but they exist none the less and this is why seemingly identical stock and modified cars never make exact numbers.

great info. I completely agree and as a scientist myself or an engineer or engine builder we value objective and specific data. However most people, including myself really want to get the bottom line about which tune delivers the fastest experience. The minute variables although when logged and measured would be a ton of info, some rolling freeway rolls really should not be effected to a great extent and I think the ONLY outcome of this test is whether one car reallly pulls ahead or they are both dead equal with a tune.

Those two scenarios will answer the OP's original question with level of specificity and utility that is tangible and beneficial to the average end user.

Bottom line this is cleary and unscientific but still nicely thought out and well controlled for what it is and should be a great watch. Sadly seems like these kind of posts never actually lead to the race or test actually happening. Will be nice if it does!

[STi_traitor]

Originally Posted by swamp2

Just to be clear you are misstating this. I never said that such a test would be invalid. What would help greatly in addition to many of the other suggestions is to:

1. Swap drivers (I think that was already mentioned)
2. Repeat the testing many times both before and after the new software.

Statistics and very good at separating systematic (real tune gains) effects from random effects.

Sounds good, we will incorporate these ideas.

[STi_traitor]

Originally Posted by BMRLVR

1) centrifugal force and inertia are the reasons why the wheels and tires need to be the same size and weight..... Remember this was a software test and not a test of which car is faster! Centrifugal force and inertia act differently then adding weight to the chassis of the car... i.e passengers.

2) burning a couple of tanks of fuel ensures that the tanks have fuel that is not a mixture of a few different octane ratings which give a cocktail that would be impossible to determine without sampling it and testing it for the AKI or RON number.
Since these engines are so dependant on timing to make power, fuel is one of the most important factors here. That ionic knock sensing system can be your friend or your enemy depending on fuel quality. One example of the octane rating being skewed would be if any of the cars had ethanol blended fuel in them (although not as common in Europe it is still available). Ethanol has a tendency to settle to the bottom of the tank which could put the pump in pure ethanol which has an RON rating of 108.6 and an AKI of 99, pretty easy to tell what effect that would have on timing, right? So again the couple tanks of fuel are just to ensure what is in the tank as an assurance of fuel type an octane rating of the fuel in both cars.

I am not going to waste any more time discussing this with you than what is above. Advise the OP's on the proper testing method since you think you know that should be and none of the rest of us do!!!

Also, kindly overlook any of my posts as I won't reply if all you want to do is try and discredit. You haven't attempted to add any value to this thread whatsoever, you were just there to jump on someone you didn't agree with. I have already established a reputation on this forum and I hold a government issued license that is recognized world wide saying that i am qualified to work on engines, not to mention hundreds of hours of technical training and tens of thousands of hours of experience. Fortunately I don't need your blessing to uphold either my licence or my reputation.

To the OP's I will reply to any questions and I hope this ends up being a fun and informative process for you guys.

*NOTE: I wrote this post on my iPhone and apologize in advance of any grammatical or spelling errors in. I will clean it up when I get in front of a computer.

Originally Posted by BMRLVR

What you are failing to see is that any test that is accepted as valid has to have controls. Just getting a baseline by running the cars side by side and seeing if they are the same is not enough. Theoretically that would make all things equal but honestly I don't think they will be exactly equal (maybe they will though!?!?). Look at all of the different results we have seen in the dyno database from cars with the exact same mods, or from stock untouched m3's. Very seldom are the dyno numbers exactly the same even thought they should be.

You are correct in the fact that if they run exactly the same for the baseline runs then they "should" be apples and apples. I found that the effects of rotational mass appears to be independent of vehicle power after doing some further reading (equivalent mass). The one thing that different wheel/tire sizes will effect is gear ratios so it is imperative that they be the same.

Anyhow, in comparing "TUNES" one needs to eliminate all variables possible. In comparing the "CARS" to one and other that is not important. The for the software to work the (DME) ECM has to recieve inputs from the various input devices (sensors) on the vehicle and make outputs to the various actuators (cam phasers, fuel injectors, ignition coils, etc). By testing two different cars slight differences in signals from the input devices (which are inevitable and uncontrollable due to production variances) and slight differences in the output actions by the actuators (also inevitable and uncontrollable) are going to effect the outcome numbers of the software flash when installed on two different cars. Although all of these components are very precisely made, there is always an allowable range of their operation be it electrical (the actual voltage it sends to the ecm which will be effected by variances in internal resistance and the PTC or NTC bias built into some thermistors) or linear/rotational (actual amount of movement from a cam phaser for a given oil pressure, probably accurate to fractions of a degree, but still any variance will have an effect) or the amount of fuel delivered by an injector at a specified duty cycle.

If you can see all of the slight variances that can exist hopefully you can see the importance of having software tested on the same car for 100% validity...... Granted the variances are usually extremely minute but they exist none the less and this is why seemingly identical stock and modified cars never make exact numbers.

Very informational! I wish we could be more exact in this comparison but we are working with what we have. I am learning a great deal from this thread.

I still think that if both are even stock this should be a great comaprison for the tunes. Just to see if one simply pulls harder than the other post tune. I doubt this may happen but that is what we are wanting to find out.

[STi_traitor]

Originally Posted by UBER ///M3

Thanks for all the suggestions we both are taking them into account the main thing is lets see how it goes, do our runs, and if the forum/mass (including us) think we can better it or something was that far off then the vbox idea will probably be used...No ones ideas are being bashed or cut down non of us are professionals we just share the same passion and are learning along the way

+1