BMW M3 Forum (E90 E92)

BMW Garage BMW Meets Register Search Today's Posts Mark Forums Read


Go Back   M3Post - BMW M3 Forum > M3 (E90 / E92 / E93) > M3 vs....
 
BPM
Post Reply
 
Thread Tools Search this Thread
      08-04-2008, 11:15 PM   #23
swamp2
Lieutenant General
swamp2's Avatar
United_States
312
Rep
10,371
Posts

Drives: E92 M3
Join Date: Sep 2006
Location: San Diego, CA USA

iTrader: (1)

footie: Good one

OC: Sorry, got your PM, it was so much info I forgot to get back to it and check it over, will do, rght now.... really working in real time now. Accel would likely yield some usable results but it is not the ideal kind. It has no DC capability. Cool stuff, that is work related for you, right?

lucid: Sounds good!
Appreciate 0
      08-05-2008, 12:52 AM   #24
footie
Major General
footie's Avatar
No_Country
204
Rep
7,507
Posts

Drives: ????????????
Join Date: Oct 2007
Location: BMW M3 will get a V6TT

iTrader: (0)

Garage List
2008 E92 M3  [0.00]
Quote:
Originally Posted by lucid View Post
Well, that Nerd speak will help us shed light on the highly speculative yet precise statements you have been making about shift times as the one below from the other thread:
Lucid,

I was only being a little light-hearted, there was not bad feeling meant by it, Swamp seen the humour in it.

I know I have made big claims for the shift time and no one hopes their right more than me because it will be me with egg on his face if proved wrong, but I will dust myself off and move on because at least we will know a lot more about how the gearbox is doing it's magic and why we don't perceive the shifts at all (apart from the surge that is present in the upper modes).
Appreciate 0
      08-05-2008, 04:39 AM   #25
lucid
Major General
lucid's Avatar
United_States
141
Rep
8,034
Posts

Drives: E30 M3; Expedition
Join Date: Apr 2007
Location: USA

iTrader: (0)

Quote:
Originally Posted by footie View Post
Lucid,

I was only being a little light-hearted, there was not bad feeling meant by it, Swamp seen the humour in it.

I know I have made big claims for the shift time and no one hopes their right more than me because it will be me with egg on his face if proved wrong, but I will dust myself off and move on because at least we will know a lot more about how the gearbox is doing it's magic and why we don't perceive the shifts at all (apart from the surge that is present in the upper modes).
I was kind of joking/pulling your leg, and kind of reminding you that you often provide precise numbers in the absence of data. Sometimes, they sound like facts rather than speculation. You might very well turn out to be right though!
__________________
Appreciate 0
      08-05-2008, 04:49 AM   #26
swamp2
Lieutenant General
swamp2's Avatar
United_States
312
Rep
10,371
Posts

Drives: E92 M3
Join Date: Sep 2006
Location: San Diego, CA USA

iTrader: (1)

More thinking...

We need to quantify lag as well as shift times and shift dynamics. The best way to do so is adding a clamp on current probe to the data acquisition system. The probe will (non-destructively) sense the signal in the paddle/lever and synchronously gather this switch signal and the acceleration signal. It should require only popping the transmission tunnel cover to get access to wiring and no cutting, crimping, soldering, nor anything else invasive/destructive. Decent low current ones seem to be about $375 minimum. Anyone know exactly how much current is flowing in a transmission control switch? I'm guessing in the mA range. Anyone have access to a mA range DC current probe that they can loan to this project. It needs to have outputs for a data acquisition system (i.e. not just a digital display). Hope I don't have to keep begging for funding...
Appreciate 0
      08-05-2008, 05:52 AM   #27
footie
Major General
footie's Avatar
No_Country
204
Rep
7,507
Posts

Drives: ????????????
Join Date: Oct 2007
Location: BMW M3 will get a V6TT

iTrader: (0)

Garage List
2008 E92 M3  [0.00]
Swamp,

Would a stereo/car alarm work shop not have this type of equipment. (Only a guess)
Appreciate 0
      08-05-2008, 06:44 AM   #28
lucid
Major General
lucid's Avatar
United_States
141
Rep
8,034
Posts

Drives: E30 M3; Expedition
Join Date: Apr 2007
Location: USA

iTrader: (0)

Quote:
Originally Posted by swamp2 View Post
More thinking...

We need to quantify lag as well as shift times and shift dynamics. The best way to do so is adding a clamp on current probe to the data acquisition system. The probe will (non-destructively) sense the signal in the paddle/lever and synchronously gather this switch signal and the acceleration signal. It should require only popping the transmission tunnel cover to get access to wiring and no cutting, crimping, soldering, nor anything else invasive/destructive. Decent low current ones seem to be about $375 minimum. Anyone know exactly how much current is flowing in a transmission control switch? I'm guessing in the mA range. Anyone have access to a mA range DC current probe that they can loan to this project. It needs to have outputs for a data acquisition system (i.e. not just a digital display). Hope I don't have to keep begging for funding...
Yes, I was asking about this on the other thread. If we want to measure the "lag" in the different systems, we will need to detect the shift signal somehow. I'll ask around for a probe as well although it would be a good start just to be able to capture the shifting action itself.

Regardless of what I can come up with temporarily, it sounds like we have the following members who are willing to donate a $100 to this noble cause so far:

Swamp
Lucid
TLud
T Bone
OC Kid (said he's in, but not sure if meant cash as well as equipment)

Come on, this is a bargain for the Euro members given the current exchange rates!
__________________
Appreciate 0
      08-05-2008, 01:24 PM   #29
T Bone
Brigadier General
T Bone's Avatar
226
Rep
4,021
Posts

Drives: 2008 335xi Coupe
Join Date: Feb 2007
Location: The land where we kill baby seals

iTrader: (0)

Quote:
Originally Posted by swamp2 View Post
More thinking...

We need to quantify lag as well as shift times and shift dynamics. The best way to do so is adding a clamp on current probe to the data acquisition system. The probe will (non-destructively) sense the signal in the paddle/lever and synchronously gather this switch signal and the acceleration signal. It should require only popping the transmission tunnel cover to get access to wiring and no cutting, crimping, soldering, nor anything else invasive/destructive. Decent low current ones seem to be about $375 minimum. Anyone know exactly how much current is flowing in a transmission control switch? I'm guessing in the mA range. Anyone have access to a mA range DC current probe that they can loan to this project. It needs to have outputs for a data acquisition system (i.e. not just a digital display). Hope I don't have to keep begging for funding...

You should measure lag at the paddle level and not at the tranny level because this is what the driver would experience. A simple switch would suffice here and would be intrusive.
__________________
"Aerodynamics are for people who cannot build engines"......Enzo Ferrari
Appreciate 0
      08-05-2008, 03:14 PM   #30
footie
Major General
footie's Avatar
No_Country
204
Rep
7,507
Posts

Drives: ????????????
Join Date: Oct 2007
Location: BMW M3 will get a V6TT

iTrader: (0)

Garage List
2008 E92 M3  [0.00]
Quote:
Originally Posted by T Bone View Post
You should measure lag at the paddle level and not at the tranny level because this is what the driver would experience. A simple switch would suffice here and would be intrusive.

I believe that is what is already being suggested.
Appreciate 0
      08-05-2008, 06:07 PM   #31
T Bone
Brigadier General
T Bone's Avatar
226
Rep
4,021
Posts

Drives: 2008 335xi Coupe
Join Date: Feb 2007
Location: The land where we kill baby seals

iTrader: (0)

Quote:
Originally Posted by footie View Post
I believe that is what is already being suggested.
See: "and synchronously gather this switch signal and the acceleration signal."

Not necessary IMHO
__________________
"Aerodynamics are for people who cannot build engines"......Enzo Ferrari
Appreciate 0
      08-05-2008, 06:23 PM   #32
swamp2
Lieutenant General
swamp2's Avatar
United_States
312
Rep
10,371
Posts

Drives: E92 M3
Join Date: Sep 2006
Location: San Diego, CA USA

iTrader: (1)

Quote:
Originally Posted by T Bone View Post
See: "and synchronously gather this switch signal and the acceleration signal."

Not necessary IMHO
Incorrect.

If you want any meaningful information on the delay you have to acquire both signals simultaneously. The accel data will show you the clutch timing and perhaps a bit about phasing but NOTHING about delay. If you want the exact time between the paddle/lever click and the beginning of the change in acceleration (i.e. the beginning of clutching) you must also record the signal from the paddle/lever at the same time. You then simply look for the delay between the two events. All data has to be gathered at a fairly high frequency when we are talking about 1/10th or 1/100th of seconds.
Appreciate 0
      08-05-2008, 07:05 PM   #33
T Bone
Brigadier General
T Bone's Avatar
226
Rep
4,021
Posts

Drives: 2008 335xi Coupe
Join Date: Feb 2007
Location: The land where we kill baby seals

iTrader: (0)

Quote:
Originally Posted by swamp2 View Post
Incorrect.

If you want any meaningful information on the delay you have to acquire both signals simultaneously. The accel data will show you the clutch timing and perhaps a bit about phasing but NOTHING about delay. If you want the exact time between the paddle/lever click and the beginning of the change in acceleration (i.e. the beginning of clutching) you must also record the signal from the paddle/lever at the same time. You then simply look for the delay between the two events. All data has to be gathered at a fairly high frequency when we are talking about 1/10th or 1/100th of seconds.

Disagree. You want to measure paddle, clutch and then acceleration. The middle step is completely unnecessary since we want to measure the time from paddle activation to the actual change in acceleration.

More data is not needed. Good data is preferred.
__________________
"Aerodynamics are for people who cannot build engines"......Enzo Ferrari
Appreciate 0
      08-05-2008, 09:32 PM   #34
swamp2
Lieutenant General
swamp2's Avatar
United_States
312
Rep
10,371
Posts

Drives: E92 M3
Join Date: Sep 2006
Location: San Diego, CA USA

iTrader: (1)

Quote:
Originally Posted by T Bone View Post
Disagree. You want to measure paddle, clutch and then acceleration. The middle step is completely unnecessary since we want to measure the time from paddle activation to the actual change in acceleration.

More data is not needed. Good data is preferred.
You are wrong. Most others seem to get it. You need to think this over some more. To quantify the delay you need to measure the trigger of the begining of the delay period, which is the paddle pull. I can not be much more clear.

I'll still take your $100 .
Appreciate 0
      08-05-2008, 10:02 PM   #35
T Bone
Brigadier General
T Bone's Avatar
226
Rep
4,021
Posts

Drives: 2008 335xi Coupe
Join Date: Feb 2007
Location: The land where we kill baby seals

iTrader: (0)

Quote:
Originally Posted by swamp2 View Post
You are wrong. Most others seem to get it. You need to think this over some more. To quantify the delay you need to measure the trigger of the begining of the delay period, which is the paddle pull. I can not be much more clear.

I'll still take your $100 .

I know you like to make things complicated.....try something simple...


Paddle Pull -> x ms -> changes in acceleration

All else is irrelevant. The g changes take into consideration, throttle management and clutch engagement.

Find x.
__________________
"Aerodynamics are for people who cannot build engines"......Enzo Ferrari
Appreciate 0
      08-06-2008, 02:38 AM   #36
footie
Major General
footie's Avatar
No_Country
204
Rep
7,507
Posts

Drives: ????????????
Join Date: Oct 2007
Location: BMW M3 will get a V6TT

iTrader: (0)

Garage List
2008 E92 M3  [0.00]
I agree with TB that we all want/need to know how long the actual time it takes from the paddle click to the start of the actual shift taking place and in all of it's varying situations, be it part throttle, constant throttle and full throttle. But I disagree that it is the most important part of the data needing collected, the total time of the shift and how much it affects to acceleration is in my opinion the most important of all data.

It's this data compared to a manual that determines why the DCT can pull away from the manual car in acceleration.
Appreciate 0
      08-06-2008, 03:05 AM   #37
TLud
Colonel
TLud's Avatar
United_States
44
Rep
2,280
Posts

Drives: '12 Golf R
Join Date: Jun 2008
Location: Texas

iTrader: (3)

Quote:
Originally Posted by footie View Post
I agree with TB that we all want/need to know how long the actual time it takes from the paddle click to the start of the actual shift taking place and in all of it's varying situations, be it part throttle, constant throttle and full throttle. But I disagree that it is the most important part of the data needing collected, the total time of the shift and how much it affects to acceleration is in my opinion the most important of all data.

It's this data compared to a manual that determines why the DCT can pull away from the manual car in acceleration.
I agree 100%. I think the curiousity in shift-response time has been blown a little out of proportion by all the lag talk that flies around this forum daily. Certainly, addressing this issue is significant, and of special interest to the M3 community, but it does not affect the actual shift time and resulting loss in acceleration, which is ultimately what we're trying to measure and compare.
Appreciate 0
      08-06-2008, 04:15 AM   #38
swamp2
Lieutenant General
swamp2's Avatar
United_States
312
Rep
10,371
Posts

Drives: E92 M3
Join Date: Sep 2006
Location: San Diego, CA USA

iTrader: (1)

Quote:
Originally Posted by T Bone View Post
I know you like to make things complicated.....try something simple...


Paddle Pull -> x ms -> changes in acceleration

All else is irrelevant. The g changes take into consideration, throttle management and clutch engagement.

Find x.
My patience is being tried. Not sure how much more simple it can get. Imagine your scenario with NO data acquisition from the paddle operation being taken. You have an accelerometer and are taking vehicle accel data before during and after a shift. ALL YOU SEE ON THE ACCEL. TRACE AS A FUNCTION OF TIME IS:
-the slightly tapering off but mostly steady state acceleration before the clutching operations
-the deceleration during the (very brief) clutching operations which appears as a very short downward blip/spike in the trace
-the acceleration going back to positive after all shift operations are complete
-you examine the trace, subtract the end event time from the begining event time and get the shift time (with the very specific definition that shift time here is NOTHING other than the clutch swap time)

PLEASE tell me how you can find the delay from the paddle click to the beginning of the deceleration data?

Do I have to make a sketch for you?

PLEASE think before you post, you are starting to look really bad.
Appreciate 0
      08-06-2008, 06:00 AM   #39
lucid
Major General
lucid's Avatar
United_States
141
Rep
8,034
Posts

Drives: E30 M3; Expedition
Join Date: Apr 2007
Location: USA

iTrader: (0)

Not sure why this got so heated. I thought we had an understanding here as to what we are trying to measure. My intention is in agreement with Swamp's. Assuming we have the instrumentation to detect the click, there are two distinct intervals:

Interval 1. The Delay: Time between paddle click and first sign of drop in acceleration (drop in acceleration signals clutch disengagement).

Interval 2. The Shift: Time between the first sign of drop in acceleration and the return to somewhat even acceleration (return to somewhat even acceleration signals the second clutch engagement and return to torque delivery in higher gear).

So we need to have 3 time measurements to be able to identify these two intervals.

Some might find a third interval meaningful as well:

Total Time from Shift Command to Shift Completion = Interval 1 + Interval 2

That's just algebra. Without the intermediate time measurement, we won't know anything about how long the mechanical action takes, which is obviously of interest because that represents the interruption in torque delivery. Also, Interval 2 is our only objective reference for determining how much faster DCT is than MT.

If we don’t have the instrumentation to detect the click, we can only measure Interval 2.
__________________
Appreciate 0
      08-06-2008, 10:16 AM   #40
T Bone
Brigadier General
T Bone's Avatar
226
Rep
4,021
Posts

Drives: 2008 335xi Coupe
Join Date: Feb 2007
Location: The land where we kill baby seals

iTrader: (0)

Quote:
Originally Posted by swamp2 View Post
My patience is being tried. Not sure how much more simple it can get. Imagine your scenario with NO data acquisition from the paddle operation being taken. You have an accelerometer and are taking vehicle accel data before during and after a shift. ALL YOU SEE ON THE ACCEL. TRACE AS A FUNCTION OF TIME IS:
-the slightly tapering off but mostly steady state acceleration before the clutching operations
-the deceleration during the (very brief) clutching operations which appears as a very short downward blip/spike in the trace
-the acceleration going back to positive after all shift operations are complete
-you examine the trace, subtract the end event time from the begining event time and get the shift time (with the very specific definition that shift time here is NOTHING other than the clutch swap time)

PLEASE tell me how you can find the delay from the paddle click to the beginning of the deceleration data?

Do I have to make a sketch for you?

PLEASE think before you post, you are starting to look really bad.

It seems you have become a rabid dog on this issue and when people disagree with you, you like to make silly statements like "Please think before you post, you are starting to look really bad".

Let's pull up your old posts where you claim near instant shifts. Relax.

If you have a switch at the paddle level, a click will "start" the clock, the end of the cycle is when acceleration resumes. Depending on the resolution of the accelerometer, you can may also see when the deceleration begins too.

The measurement of the start and end of the cycle is exactly what a driver would feel. What a driver would feel is what the testing methodology should try to replicate.
__________________
"Aerodynamics are for people who cannot build engines"......Enzo Ferrari
Appreciate 0
      08-06-2008, 10:16 AM   #41
TLud
Colonel
TLud's Avatar
United_States
44
Rep
2,280
Posts

Drives: '12 Golf R
Join Date: Jun 2008
Location: Texas

iTrader: (3)

Very well put, lucid.
Appreciate 0
      08-06-2008, 10:18 AM   #42
T Bone
Brigadier General
T Bone's Avatar
226
Rep
4,021
Posts

Drives: 2008 335xi Coupe
Join Date: Feb 2007
Location: The land where we kill baby seals

iTrader: (0)

Quote:
Originally Posted by lucid View Post
Not sure why this got so heated.

I dunno why either....SWAMP2 likes to drop in words like INCORRECT, or phrases like YOU ARE WRONG and imply that people are idiots if they disagree with him.

Take the issue up with him.
__________________
"Aerodynamics are for people who cannot build engines"......Enzo Ferrari
Appreciate 0
      08-06-2008, 10:27 AM   #43
TLud
Colonel
TLud's Avatar
United_States
44
Rep
2,280
Posts

Drives: '12 Golf R
Join Date: Jun 2008
Location: Texas

iTrader: (3)

Quote:
Originally Posted by T Bone View Post
It seems you have become a rabid dog on this issue and when people disagree with you, you like to make silly statements like "Please think before you post, you are starting to look really bad".

Let's pull up your old posts where you claim near instant shifts. Relax.

If you have a switch at the paddle level, a click will "start" the clock, the end of the cycle is when acceleration resumes. Depending on the resolution of the accelerometer, you can may also see when the deceleration begins too.

The measurement of the start and end of the cycle is exactly what a driver would feel. What a driver would feel is what the testing methodology should try to replicate.
TB, I think you're most interested in what lucid refers to as "Total Time from Shift Command to Shift Completion" (Interval 1 + Interval 2). I agree that we should try to measure this, but I think that what Swamp is trying to say (and I agree) is that measuring Interval 2 ultimately is more important because that is what will be compared to the 6MT to determine relative shift times and acceleration loss. As Enigma pointed out in the other thread, when the car shifts after your trigger pull is not as important as how fast the gear change itself occurs.

Swamp, let's keep this civil. I appreciate your initiation and efforts on this, but if it's going to be successful, we all need to be open to critical analysis and opinion.
Appreciate 0
      08-06-2008, 10:34 AM   #44
T Bone
Brigadier General
T Bone's Avatar
226
Rep
4,021
Posts

Drives: 2008 335xi Coupe
Join Date: Feb 2007
Location: The land where we kill baby seals

iTrader: (0)

Quote:
Originally Posted by TLud View Post
TB, I think you're most interested in what lucid refers to as "Total Time from Shift Command to Shift Completion" (Interval 1 + Interval 2). I agree that we should try to measure this, but I think that what Swamp is trying to say (and I agree) is that measuring Interval 2 ultimately is more important because that is what will be compared to the 6MT to determine relative shift times and acceleration loss. As Enigma pointed out in the other thread, when the car shifts after your trigger pull is not as important as how fast the gear change itself occurs.

Thank you for succinctinly highlighting the differences in the proposed testing methods.

For me, it is interesting to see how quickly the MDCT shifts once it initiates the shift but based on the "lag" videos we have seen for the MDCT, this metric may not be as relevant in the real world as measuring the shift from the paddle actuation. Additionally, the driver only feels the shift from paddle actuation to the resumption of acceleration.

When the alleged software update comes in September, I hope that we have a before and after dataset for comparison.
__________________
"Aerodynamics are for people who cannot build engines"......Enzo Ferrari
Appreciate 0
Post Reply

Bookmarks

Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT -4. The time now is 04:40 PM.




m3post
m3post
Powered by vBulletin® Version 3.7.0
Copyright ©2000 - 2016, Jelsoft Enterprises Ltd.
1Addicts.com, BIMMERPOST.com, E90Post.com, F30Post.com, M3Post.com, ZPost.com, 5Post.com, 6Post.com, 7Post.com, XBimmers.com logo and trademark are properties of BIMMERPOST