M-DCT Lag - What's the reason?

[mkoesel]

Originally Posted by nitramsen

Exactly my point for a mechanical issue. Why would they introduce a lag via software if this is not a physical related phenomen?

I don't know that I understand your logic.

One theory, for example, is that the fix to the stalling issue resulted in the lag issue. I've explained in the past how this could be the case. Granted that still doesn't rule out some underlying hardware issue, but it certainly does not confirm one either. So, again, I don't see your logic.

But how many of you have actually tried to trace the error via OBDII or CAN?

We haven't, but apparently there is an aftermarket solution in the works (the thread was linked a ways back). Presumably these folks have done this, or at least tried, because they claim to have a solution to the lag. Of course, their software is not yet available so we don't have proof yet.

I assume few so why don't you Swamp2 use your Math,Physics with limited development exprience to tried to get some hard facts for your postulates.

Probably because he hasn't had the time and resources. I don't suppose you can back up your theory with facts either? Remember, the Ferrari California uses this same transmission with no lag, as reported by at least one forum member. Now, are we sure the transmission didn't get some hardware changes? No, we cannot be 100% certain of that. But it is reportedly the same part number.

[nitramsen]

Originally Posted by mkoesel

I don't know that I understand your logic.

One theory, for example, is that the fix to the stalling issue resulted in the lag issue. I've explained in the past how this could be the case. Granted that still doesn't rule out some underlying hardware issue, but it certainly does not confirm one either. So, again, I don't see your logic.



We haven't, but apparently there is an aftermarket solution in the works (the thread was linked a ways back). Presumably these folks have done this, or at least tried, because they claim to have a solution to the lag. Of course, their software is not yet available so we don't have proof yet.



Probably because he hasn't had the time and resources. I don't suppose you can back up your theory with facts either? Remember, the Ferrari California uses this same transmission with no lag, as reported by at least one forum member. Now, are we sure the transmission didn't get some hardware changes? No, we cannot be 100% certain of that. But it is reportedly the same part number.

If you assume that BMW is so stupid that they cannot fix one issue without testing for others (Rerunning their regression test) why did you buy a BMW?

My simple logic is that BMW actually knows how to build and maintain cars but budgets and time limits prevents them from finishing a product completely. I think that there is a good chance that this is just as always a mechanical issue that will be dealt with later.

Assuming (More on that later) that the lag was introduced by the first update it could be that for some reason the pressure was not high enough to hold the clutch thereby stalling the car by slamming the clutch. See fixing this would mean that you need to either chance slower (Lag) or keeping a higher gear (V33). I know this is just speculation but the reasoning is just as sound as simply saying I know someone how thinks that his FC does not lag and therefore it must be software.

So there is one person that says his FC does not have a lag and by the away what is the partnumber for the high pressure pump on that gearbox (I would like to know if there is any relation to my theory)?

I bet that I can find 100+ BMW M3 DCT drivers who thinks their car works just fine. Does that mean that the bug is not there? At least according to you, me and a lot of other people definitely not.

Are we 100% sure that this lag was not there from the beginning (This may actually support a software fix)? I have dealt with software development and administration for +20 years. I can tell you that I have tried countless times that a user finds a bug that was known before a fix for another issue is released. When you explain to them that this has always been the case they go into tilt mode. I am sure that a lot of people on this board working within IT will agree to this.

Get any IT supporter to read this thread and he will doubt that the issue was not there from the beginning (I claim it was not there from the beginning but I assume that I could be stupid user in the case) because it has 99% of classic user denial.


And yes I do have Accelerator pedal position, Absolute throttle position, RPM and speed read outs while a lag Attack. I took me less than 2 hours to make the program including chart readout for this and 58USD for the hardware.

I am not 100% sure of anything but I find it amazing that a few guys here keeps trying to find excuses why BMW did not complete mischaracterized the functionality of DCT in their advertisement.

The only facts we know for sure is that BMW (Or at least my dealer speaking for BMW, Local Importer and a SIB) is saying this is by design and that no fix will be available for this.
Still you and a few others know better…….

[mkoesel]

Originally Posted by nitramsen

If you assume that BMW is so stupid that they cannot fix one issue without testing for others (Rerunning their regression test) why did you buy a BMW?

I bought a BMW because I like their cars. I sure didn't buy a BMW because I believe that they have 100% bug-free software. How bout you? Why did you buy a BMW if you believe they would install a faulty transmission in a $70k performance car?

My simple logic is that BMW actually knows how to build and maintain cars but budgets and time limits prevents them from finishing a product completely.

That is simple logic indeed. Simple enough, in fact, to cover the software possibility too.

... just as sound as simply saying I know someone how thinks that his FC does not lag and therefore it must be software.

...

I bet that I can find 100+ BMW M3 DCT drivers who thinks their car works just fine. Does that mean that the bug is not there? At least according to you, me and a lot of other people definitely not.

If you go back and actually read, the person who reported on the California owns an M3 M-DCT and knows exactly how to reproduce the lag. He tried to do so on the Ferrari but said that he could not get it to lag.

...and by the away what is the partnumber for the high pressure pump on that gearbox (I would like to know if there is any relation to my theory)?

No idea, you'd have to look it up.

Are we 100% sure that this lag was not there from the beginning (This may actually support a software fix)?

No.

I am not 100% sure of anything but I find it amazing that a few guys here keeps trying to find excuses why BMW did not complete mischaracterized the functionality of DCT in their advertisement.

I'm not making excuses, and I can't recall anyone else doing so.

The only facts we know for sure is that BMW (Or at least my dealer speaking for BMW, Local Importer and a SIB) is saying this is by design and that no fix will be available for this. Still you and a few others know better…….

What does this have to do with hardware vs. software? You're seem to be attacking me for no reason. I don't "know better" than anybody else. All I know is that the lag is an issue on this car, and I'd like it fixed.

[GT3 Tim]

FYI -- My car had lag from day one. It also stalled several times too (that seems to have been corrected). In my opinion, the lag was not a side-effect from fixing the stalling, because mine has always had the low speed lag issue. 7/08 production.

[mkoesel]

Originally Posted by GT3 Tim

FYI -- My car had lag from day one. It also stalled several times too (that seems to have been corrected). In my opinion, the lag was not a side-effect from fixing the stalling, because mine has always had the low speed lag issue. 7/08 production.

Thanks for that info Tim.

Would you say that the lag was worse after the original software update? I think the commonly held belief is that the update did not cause the lag, but did exacerbate it. Now, I only owned my car a very short time before the I got the update. And I only rarely get lag as it is, since I use S mode and tend to downshift early. So I'm far from the best source for a before/after comparison. But I do remember that as reports on the update rolled in from forum members, more than one person seemed to think that this lag was made worse (some seemed to think much worse, as I recall it) by the first software update. And I think that is where the basis for the link between the update and the lag originated. Granted, it seems no one really can say for sure, because I doubt anyone actually compared the two with a timer or whatever instrumentation, especially on this specific issue.

Of course, in the scheme of things, its not really an important point, except possibly to BMW who is (or who hopefully is) trying to debug this thing. And I suppose by now they will know whether the other update has any part in the lag or not.

[aus]

Originally Posted by nitramsen

Exactly my point for a mechanical issue. Why would they introduce a lag via software if this is not a physical related phenomen?

You are talking about a COMPANY with massive processing power in both human and hardware form so if this was a pure software bug this would have been solved long ago. I assume BMW knows about regression test and therefore this points in my direction not your.

And to Swamp2 I have read the thread and I see no proof of pre March 2009 DCT working. I assume that if BMW realizes that they made a mechanical mistake they will try to fix it in later versions. Like they have done in all previous M's with mechanical issues where they fix it but did not provide early version with any corrections.(I still remember the high oil burn of early E39 m5 compared to after 2001, did BMW provide a fix for pre 2001?)

But how many of you have actually tried to trace the error via OBDII or CAN?
I assume few so why don't you Swamp2 use your Math,Physics with limited development exprience to tried to get some hard facts for your postulates.

You give BMW WAY too much credit for doing the right thing on their own.
They REPEATEDLY DENIED there was a problem with the S54, saying it was operator error, saying people over reving and abusing the car... until the SMG equiped cars started blowing up too.
Then they REPEATEDLY DENIED there was a problem with the N54 AFTER a software update that introduced lag. It took the internet forums to force them to finally admit there was lag introduced into the software. They claim there's a fix now, but people still don't feel the car is the same as it was when they bought it.

BMW usually EVENTUALLY does the right thing, but they need to be pushed and sometimes forced to do the right thing.

My last question is do you OWN a E9x M3 with DCT? Or are you just another 6 speed guy who's trying to stir the pot, and should be IGNORED.

[Montoya]

First: please understand this post is in the interest of getting to the bottom of the issue!

Why is it that some of us have the lag and some don't? I have tried every method mentioned on this thread to duplicate the lag to no effect (believe me I have really tried multiple times!). Everytime the car either changes gear and accelerates or just accelerates in the current gear.

I know that I had a recent update to my software, but... the problem was not present before. A lot of speculation so far is that the problem is learned when you do a lot of drive in D modes, which I do not. But, reading through the most recent threads software vs. hardware, I'm beginning to think that the issue is hardware variation. In other words some vehicles have certain items like clutch and sensors that cause the software to cause a lag, while others like me have the same sensors etc, more "in spec" that allows normal operation.

My point is that it isn't universal and that some (many?) don't have the problem, so it makes the troubleshooting that much harder. I'm not trying to pick sides on the debate, but just saying maybe it is a little of both issues- hardware and software. Ultimately the fix might be either.

OK, now I'm lazy for not doing a search, but what has been the result when one of you with the lag has demonstrated the issue to the BMW regional tech rep? Has the answer been "normal function" as you sit waiting for one to two seconds in traffic waiting for the car to accelerate? It seems to me that if you could put a tech rep in "harms way" so to speak, we might get some action, perhaps BMW taking the car for further analysis to find the real issue....

[nitramsen]

Originally Posted by aus

You give BMW WAY too much credit for doing the right thing on their own.
They REPEATEDLY DENIED there was a problem with the S54, saying it was operator error, saying people over reving and abusing the car... until the SMG equiped cars started blowing up too.
Then they REPEATEDLY DENIED there was a problem with the N54 AFTER a software update that introduced lag. It took the internet forums to force them to finally admit there was lag introduced into the software. They claim there's a fix now, but people still don't feel the car is the same as it was when they bought it.

BMW usually EVENTUALLY does the right thing, but they need to be pushed and sometimes forced to do the right thing.

My last question is do you OWN a E9x M3 with DCT? Or are you just another 6 speed guy who's trying to stir the pot, and should be IGNORED.

I never owned a turbo so I will not comment on that (Okay I had a GOLF GTI once in 1998). But the S54 issue is the way I think BMW normally handles bad mechanical implementations (Upgrade new versions and extend warranty on old once same with E39 M5 oil consumption and the piston rings). That does not help us because our cars will never brake because of the lag.

What makes things matters worse is that for the first time ever ( That I have personally experienced) BMW made advertisement that is not as the car performs.

Again my main opposition here is that some people here keeps clinging to false hope that BMW will fix this by postulating that this is only software based on nothing more than speculation. And I think that for people considering buying a DCT it should be made clear that it is very possible that the lag will never be resolved completely in old models because the likelihood of this being entirely software is low in my opinion.

And yes I own an E90 DCT search the forum I was the first to publish pictures on this forum with an E90 DCT.

[nitramsen]

Originally Posted by Montoya

First: please understand this post is in the interest of getting to the bottom of the issue!

Why is it that some of us have the lag and some don't? I have tried every method mentioned on this thread to duplicate the lag to no effect (believe me I have really tried multiple times!). Everytime the car either changes gear and accelerates or just accelerates in the current gear.

I know that I had a recent update to my software, but... the problem was not present before. A lot of speculation so far is that the problem is learned when you do a lot of drive in D modes, which I do not. But, reading through the most recent threads software vs. hardware, I'm beginning to think that the issue is hardware variation. In other words some vehicles have certain items like clutch and sensors that cause the software to cause a lag, while others like me have the same sensors etc, more "in spec" that allows normal operation.

My point is that it isn't universal and that some (many?) don't have the problem, so it makes the troubleshooting that much harder. I'm not trying to pick sides on the debate, but just saying maybe it is a little of both issues- hardware and software. Ultimately the fix might be either.

OK, now I'm lazy for not doing a search, but what has been the result when one of you with the lag has demonstrated the issue to the BMW regional tech rep? Has the answer been "normal function" as you sit waiting for one to two seconds in traffic waiting for the car to accelerate? It seems to me that if you could put a tech rep in "harms way" so to speak, we might get some action, perhaps BMW taking the car for further analysis to find the real issue....

I did test my tech shop man and he did clearly ackownlegde the issue until BMW told him otherwise. So did BMW tech rep until they failed to repair it.

So I am not the only one going for a not pure software bug......

Let me be clear software would have been fixed long ago by a company of BMW resources in my opinion.

[Projectile]

Originally Posted by swamp2

There are other cars with identical hardware with no lag.

Since it seems clear that the problem doesn't seem to affect all M3s, I don't see how this is a meaningful piece of information.

If you are referring to the Ferrari Califorina, I don't see how you could possibly know that none of them have this issue.

[nitramsen]

Originally Posted by mkoesel

I bought a BMW because I like their cars. I sure didn't buy a BMW because I believe that they have 100% bug-free software. How bout you?

I used to like their cars too but I think that my next car may have a Star.

Originally Posted by mkoesel

Why did you buy a BMW if you believe they would install a faulty transmission in a $70k performance car?.

Can you say $130K , and no I had hoped that the issue a would be limited to something covered by the usual extended warranty. By the way I would not have bought the car knowing what I know today.

Originally Posted by mkoesel

That is simple logic indeed. Simple enough, in fact, to cover the software possibility too.

No because with simple logic a simple software issue would have been fixed long ago.

Originally Posted by mkoesel

If you go back and actually read, the person who reported on the California owns an M3 M-DCT and knows exactly how to reproduce the lag. He tried to do so on the Ferrari but said that he could not get it to lag.

Good for him but common sense tells me that I cannot believe a single guy on the internet. And is it not oversimplifying the issue by saying same part therefore it must be software. Would it not be more correct to say same part but a lot of different factors (agree software is probably one of them but definitely not the only one)?

Originally Posted by mkoesel

I'm not making excuses, and I can't recall anyone else doing so.
What does this have to do with hardware vs. software? You're seem to be attacking me for no reason. I don't "know better" than anybody else. All I know is that the lag is an issue on this car, and I'd like it fixed.

I did not attack you but you decided to answer my comment to Swamp2/AUS. Please remember that my first comment was because Swamp2 replied the following to another peer.

“There is no way you are going to incite violence. However, if you want to be informed and speak intelligently on this topic you should read the entire thread. We know this is a software problem and we know it can be fixed 100% with no adverse affects on upshifting or other aspects of the performance of the system.

By the way, since you asked, my background is math, physics and mechanical engineering. I've done some programming as well. There are a lot of scientists, engineers and programmers here on this forum.”

I do not know who we are but it does not include me and I wanted to make that clear.

If Swamp2 is not excusing BMW what is “There are other cars with identical hardware with no lag.” With this he assumes that BMW did only f….. a simple software. To me they gave me false information that let me to buy the car and that Ferrari can make this work in its full complexity does not mean that BMW will be able to.

[Race_Doc11]

I couldn't agree more ...THERE IS NO EXCUSE FOR 80K CAR TO HAVE A SAFTY ISSUE LIKE THIS.

Originally Posted by nitramsen

To me they gave me false information that let me to buy the car

[swamp2]

In essence all of the discussion about the technical details of what is really causing the lag is only interesting speculation and in some ways some fun amateur detective work to pass the time as we anxiously wait for a fix. I personally believe the software angle myself and there is evidence for it. It could be hardware and it could be some combination of both. However, attempting to argue it is something other than software is nearly an impossible argument. It would require immensely detailed engineering knowledge about a very complex system and this will never be available to any of us. However, on the flip side if a software revision fixes the problem the conclusion is rock solid. Of course there is the small part of me that hopes it is software purely because it should mean an easier and more likely fix.

Let me also admit I completely over stated my position when I claimed we know 100% that this is a software issue. My frustration with folks chiming in who have not read enough pushed me to a very hasty post.

With that being said I will offer some direct replies to statements I find interesting or controversial.

Originally Posted by nitramsen

But how many of you have actually tried to trace the error via OBDII or CAN? I assume few so why don't you Swamp2 use your Math,Physics with limited development exprience to tried to get some hard facts for your postulates.

I only offered my background because a fellow poster was interested in the relevant background of those involved in this discussion. Seems like a great question to me; to whom would you lend more credibility to when discussing engineered systems an engineer or a random guy with a GED and really likes cars. It is no absolute, engineers are wrong all the time, it is just simply an issue of source quality.

Originally Posted by nitramsen

And yes I do have Accelerator pedal position, Absolute throttle position, RPM and speed read outs while a lag Attack. I took me less than 2 hours to make the program including chart readout for this and 58USD for the hardware.

I was unaware that there was a system available for recording vehicle functions. I have used a custom DAQ/accelerometer system to measure DCT shift times but have not done work like this. Again it is probably only an interesting way to pass the time, but you absolutely should share your information rather than taunting us with it. One reason I am interested in the data is because many believe the problem is with clutch or gear selection delay, whereas from watching the rpms and noting throttle pedal position my contention is that it is simply a flat/dead spot in the response of the actual throttle compared to the throttle pedal input. Now what causes this discrepancy is a whole other question.

Originally Posted by nitramsen

Again my main opposition here is that some people here keeps clinging to false hope that BMW will fix this by postulating that this is only software based on nothing more than speculation. And I think that for people considering buying a DCT it should be made clear that it is very possible that the lag will never be resolved completely in old models because the likelihood of this being entirely software is low in my opinion.

Let me provide the reasons why it is more than speculation that the issue is software related.

1. You can entirely avoid the lag by downshifting manually BEFORE the car enters its automatic downshift mode. If a human can push the shift paddle resulting in the cascading of gear fork selection/movement and clutch movements, such an identical sequence can be done by software, all of the systems software and hardware is so much faster than a human.

2. Other problems with the transmission have been fixed by pure software updates. Rather large problems like the no downshift under hard braking and the upshift delay. Both were certainly on the same order of complexity and severity as the current lag bug. From what I have heard there have been previous BMWs with many many revisions of ECU/transmission software. What are they doing - obviously both fixing bugs and making improvements.

3. Although it is speculative to believe the that internals of the Ferrari California dual clutch transmission are identical to ours they do share the same part number and from a manufacturing point of view it would make sense for Getrag to develop the transmission, get it working, leave it unchanged and sell the hell out of it. Although I've debated endlessly with member Footie who reported his experience with the F.C. I trust he appropriately tried to get the lag and found it absolutely not present. Of course since Footie has not driven every F.C. you can't jump to the conclusion that none of them have the issue. I have spent some time on google trying to find some evidence of lag in that car but have been unable to locate anything.

4. Despite what BMW claims about the issue being intended we know that is simply the liability legal team speaking, not engineering. Does it make sense that during development Getrag would release hardware with such a large flaw? It just doesn't.

Now, even though I know the answer, what evidence is there that it is hardware related? Being difficult or slow to fix does not allow the leap of faith to "it must be hardware".

Originally Posted by nitramsen

Let me be clear software would have been fixed long ago by a company of BMW resources in my opinion.

Now this is truly speculation at its best. I do see the appeal and logic of this belief but that is not evidence. I do think software for transmission issues has to make its way through EPA mileage testing and there are probably other governmental/regulatory/bureaucratic hurdles involved as well.

I too have worked in the software industry and I have seen seemingly simple bugs remain in release after release of software. There are a huge number of non technical reasons why things like this can happen, work place politics, personal decisions low in the chain of command, difficulty of the fix, concern for the effects of possible unintended consequences, etc., etc.

Again, please do post the data you collected. I am very interested in the TPS/rpm information. As well to understand the data we must know if the TPS is requested TPS from the pedal or actual TPS from the actual butterfly valve position.

[swamp2]

Originally Posted by Montoya

Why is it that some of us have the lag and some don't? I have tried every method mentioned on this thread to duplicate the lag to no effect (believe me I have really tried multiple times!). Everytime the car either changes gear and accelerates or just accelerates in the current gear.

I know that I had a recent update to my software, but... the problem was not present before.

This is very good information.

Although I do not want to challenge your powers of observation it would be fantastic if you could find someone with a DCT and with the lag to try to reproduce it in your car. You guys could trade cars and simply see what happens for both of you. Multiple observers can go a long way in increasing certainty and the reliability of observations.

-Did you really go out and try to reproduce the lag pre and post your software update?
-Why are you even interested in the lag if you never had it?
-Why did you need a software update?
-Which versions of software did you have pre and post update?

If it is the case that some cars are completely lag free this brings a whole other layer of complexity to the table. Understanding how systems that are very likely identical with regards to hardware and software can have such different behavior is a bit mind bending.

[garek]

Originally Posted by nitramsen

No because with simple logic a simple software issue would have been fixed long ago.

I've been reading your posts and your logic is flawed.

Your assumption is that because BMW released new software and that because the lag was not fixed or got worse, then it is a hardware issue. This simply just does not follow.

From my understanding, the lag was not a problem with the original software release. Then with a later release, the lag became a problem. Simple logic tells you there that since the hardware did not change, but the software DID change, then the change in the software is what introduced the lag issue.

Stated another way, you have item A and you have item B which gives you result C. Item A changed, item B remained the same, yet result C changed. It stands to reason that the change in item A caused the change in result C. However, by your statements you are saying that even though B remained the same, the new result C is caused by B remaining the same.

Also, this statement:

Originally Posted by nitramsen

No because with simple logic a simple software issue would have been fixed long ago.

This is just an assumption that has no basis in reality. How do you know that this is a "simple" software issue? The problem causing the lag may be the result of several complex factors combining to produce an unexpected result in the software. I write software for a living, so I know than unless you write logically perfect code, it is very easy to make a change in the code to fix one problem, but then create another unexpected problem.

Your assumption that each iteration of software will have fewer bugs and no new bugs will be introduced is simply wrong. If that were the case then Windows would have been bug free 10 years ago.

[nitramsen]

Originally Posted by swamp2

I was unaware that there was a system available for recording vehicle functions.

OBDII and a few lines of C# will make the system avaliable. I know that it sounds wierd but you will not get my data nor the software for personal reasons. I will happily assists anyone in building a system for loging the OBDII data but you have to do it on your own. (All you need is really wiki on OBDII).

Originally Posted by swamp2

1. You can entirely avoid the lag by downshifting manually BEFORE the car enters its automatic downshift mode. If a human can push the shift paddle resulting in the cascading of gear fork selection/movement and clutch movements, such an identical sequence can be done by software, all of the systems software and hardware is so much faster than a human.

Yes, but do you think EPA or the EU will like that solution? And is this not an argument for my position that software issue would have been fixed long ago?

Give my the code and a compiler and I will fix in 10 minutes. But I think that a lot of other BMW cliams will be voided by that fix.

Originally Posted by swamp2

2. Other problems with the transmission have been fixed by pure software updates. Rather large problems like the no downshift under hard braking and the upshift delay. Both were certainly on the same order of complexity and severity as the current lag bug. From what I have heard there have been previous BMWs with many many revisions of ECU/transmission software. What are they doing - obviously both fixing bugs and making improvements.

None of those requeired regulative(liter/100km) changes.

Originally Posted by swamp2

3. Although it is speculative to believe the that internals of the Ferrari California dual clutch transmission are identical to ours they do share the same part number and from a manufacturing point of view it would make sense for Getrag to develop the transmission, get it working, leave it unchanged and sell the hell out of it. Although I've debated endlessly with member Footie who reported his experience with the F.C. I trust he appropriately tried to get the lag and found it absolutely not present. Of course since Footie has not driven every F.C. you can't jump to the conclusion that none of them have the issue. I have spent some time on google trying to find some evidence of lag in that car but have been unable to locate anything.

Dependencies, Dependencies, Dependencies

Originally Posted by swamp2

4. Despite what BMW claims about the issue being intended we know that is simply the liability legal team speaking, not engineering. Does it make sense that during development Getrag would release hardware with such a large flaw? It just doesn't.

I never said the gearbox is the main reason, but Dependencies that is another issue.

Originally Posted by swamp2

Now, even though I know the answer, what evidence is there that it is hardware related?

After eliminating the impossible, whatever remains, no matter how improbable, is the truth. Software is the only close impossible i know of (As you said the fix would be very easy so that fact that is was not done must mean that it is close to impossible for some reason).

Originally Posted by swamp2

Again, please do post the data you collected. I am very interested in the TPS/rpm information. As well to understand the data we must know if the TPS is requested TPS from the pedal or actual TPS from the actual butterfly valve position.

The data is for both. But again I cannot supply that data nor software.
HEre is a simple meta for the software.

Open serial (Bluetooth device worked for me).
set speed
close serial
Open Serial fullspeed
Set disable echo

while not stopped
read sensor1
Store sensor1 + ticks or datetime
read sensor2
Store sensor2 + ticks or datetime
key pressed store lag is progress
close store
close serial
App 50 lines of code.
Simple excel can do the rest.

[nitramsen]

Originally Posted by garek

I've been reading your posts and your logic is flawed.

Your assumption is that because BMW released new software and that because the lag was not fixed or got worse, then it is a hardware issue. This simply just does not follow.

From my understanding, the lag was not a problem with the original software release. Then with a later release, the lag became a problem. Simple logic tells you there that since the hardware did not change, but the software DID change, then the change in the software is what introduced the lag issue.

Stated another way, you have item A and you have item B which gives you result C. Item A changed, item B remained the same, yet result C changed. It stands to reason that the change in item A caused the change in result C. However, by your statements you are saying that even though B remained the same, the new result C is caused by B remaining the same.

Also, this statement:

This is just an assumption that has no basis in reality. How do you know that this is a "simple" software issue? The problem causing the lag may be the result of several complex factors combining to produce an unexpected result in the software. I write software for a living, so I know than unless you write logically perfect code, it is very easy to make a change in the code to fix one problem, but then create another unexpected problem.

Your assumption that each iteration of software will have fewer bugs and no new bugs will be introduced is simply wrong. If that were the case then Windows would have been bug free 10 years ago.

Regarding windows your are correct but they keep adding new functionality (Did you forget that). Why do a have a stabilisation iteration?

And you logic is to simply for a programmer of real time systems. Requierement A changed which changed code B but physical Item C cannot change. Requierment A is now required Item C must change.

Regression tests will help you with your programming.

[garek]

Originally Posted by nitramsen

Regression tests will help you with your programming.

No need. I write logically perfect software.

[swamp2]

Originally Posted by nitramsen

OBDII and a few lines of C# will make the system avaliable. I know that it sounds wierd but you will not get my data nor the software for personal reasons. I will happily assists anyone in building a system for loging the OBDII data but you have to do it on your own. (All you need is really wiki on OBDII).

It does sound very weird. It seems a stretch for you be any sort of enthusiast with interesting data on a problem like this and be unwilling to share it. None of us want to deal with building a custom system and writing code. We simply want to be driving bug free cars. Surely you can understand that.

Originally Posted by nitramsen

Yes, but do you think EPA or the EU will like that solution? And is this not an argument for my position that software issue would have been fixed long ago?

No it isn't. If you didn't notice replicating human behavior vis computer remains a huge challenge. I also already addressed the point of view that "simple" always is equivalent to "will be fixed quickly". That is completely flawed logic and contradicted by multiple replies from other software folks herein.

Originally Posted by nitramsen

Give my the code and a compiler and I will fix in 10 minutes. But I think that a lot of other BMW cliams will be voided by that fix.

Wait you have me really confused now, in part because of poor/hurried/unclear writing. Are you now saying you can fix this problem yourself? You do not even believe it is software related in the first place. That is not only completely contradictory but an awfully bold statement. I guess it is consistent with your flawed view that everything is simple and everything is black and white. I would place any monetary wager that you could not accomplish this anyway, certainly not in 10 minutes.

Originally Posted by nitramsen

None of those requeired regulative(liter/100km) changes.

Pure speculation. Please prove that. You can't becuase BMW does not list publicly the content of software updates.

Originally Posted by nitramsen

After eliminating the impossible, whatever remains, no matter how improbable, is the truth. Software is the only close impossible i know of (As you said the fix would be very easy so that fact that is was not done must mean that it is close to impossible for some reason).

Absolutely not, see above. Calling it impossible to be software is just as premature as I was calling it 100% closed as being software. Either way you look at it there is more real hard evidence that it is software than that it is hardware.

Originally Posted by nitramsen

The data is for both. But again I cannot supply that data nor software.
HEre is a simple meta for the software.

Open serial (Bluetooth device worked for me).
set speed
close serial
Open Serial fullspeed
Set disable echo

while not stopped
read sensor1
Store sensor1 + ticks or datetime
read sensor2
Store sensor2 + ticks or datetime
key pressed store lag is progress
close store
close serial
App 50 lines of code.
Simple excel can do the rest.

Not interested in your code nor algorithm. As you said no particular magic there. I am interested in two things only. 1. What is the cause, 2. What/when will there be a fix. Perhaps you are interested in profiting from this issue with a software fix. I suppose that is fine but you should be upfront with your reasons for stating what you do and with holding what you do. There is an awfully quick way to settle any debate - post some data.

[alito]

wow... what a turn this thread has made... the aggression. Guys... we talk of hardware and software as if this was a computer. Has anyone actually taken a look at what the transmission looks like? It is a Dual Clutch Transmission. It is simply two transmissions side my side. One plate is ready on the adjacent transmission clutch to help anticipate and make the shift faster... not lag slower. The software simply tells these plates (clutch) when to be ready and when to engage. It is an granted elaborate algorithm that is meant to anticipate the next shift and shift pattern based on multiple factors. The initial algorithm (software) made the car have erratic shift patterns, have hard shifts, shift after the car was nearly stopped, shift too fast etc. In an effort to "smooth" things out, BMW designed an algorithm (software) that worked great until the very last shift when it literally gets "confused" as to where to go hence it "lags" in its decision making hence our problem. Hardware would mean the transmission itself was bad and it would have the same issues on manual because a plate is a plate and a bold is a bolt.
What I am surprised in is how little BMW keeps its customers in the loop and how they act like we are the crazy ones. No matter what your individual opinion of the cause, please keep it civil and just give your opinion... don't get into pissing contests. If everyone in this forum seems wrong, do not waste your time and go elsewhere. No more insulting please.
As previously stated by me, I have plenty of experience with dealing with BMW and I agree that they will eventually do the right thing but have to be pushed... hard. The recent DOT involvement I believe will push them to do it. I am truly surprised it has taken the engineers so long to do something.

Again, please be nice!

[gobruins]

i've been following this forum (and in particular, this thread) for quite some time, and i thought i should throw this idea out there. from what i understand, the general consensus has been that this issue is a software issue because some owners experience the lag and others don't (despite purported equality in hardware). personally, i experience the lag, and i have my reasons for believing it is software related.

just out of curiosity, for both members who experience the lag and those who don't, what happens when you hold down the "-" paddle shifter (i.e. the down paddle shift on the steering wheel) for an extended period of time? normally, i click the - paddle shifter to shift down, and 90% of the time, it downshifts. sometimes, it doesn't downshift. however, when i press and hold the paddle shifter (e.g., for 3 or more seconds), the transmission continues to shift down through gears sequentially and rapidly. For instance: if i am in S1 drivelogic (no rev matching) and in 3rd gear, i press and HOLD the down shift paddle on the steering wheel, the tranny will shift to second and then quickly to first gear, revs throw up, and rev matching occurs.

my intention with asking others to conduct the aforementioned test is to determine: 1) do other DCT owners experience this behavior? 2) do those without the lag also experience this behavior? if not, we may have pinpointed another symptom/abnormality in DCT behavior and in the process, provided further evidence that this is a software related issue and not one of hardware.

as a side note, i have neither heard nor read that this is a built-in feature of DCT, so please correct me if i'm wrong.

[Montoya]

Originally Posted by swamp2

This is very good information.

Although I do not want to challenge your powers of observation it would be fantastic if you could find someone with a DCT and with the lag to try to reproduce it in your car. You guys could trade cars and simply see what happens for both of you. Multiple observers can go a long way in increasing certainty and the reliability of observations.

-Did you really go out and try to reproduce the lag pre and post your software update?
-Why are you even interested in the lag if you never had it?
-Why did you need a software update?
-Which versions of software did you have pre and post update?

If it is the case that some cars are completely lag free this brings a whole other layer of complexity to the table. Understanding how systems that are very likely identical with regards to hardware and software can have such different behavior is a bit mind bending.

I will try to answer your questions. First, while I don't have the problem, of course I am interested in the issue since we are not positive what can cause it now or later.

I did try to reproduce the lag pre and post software update.

I had the software update because the lag on shifting, not this lag.

I wish I knew what version that was put in. I have asked, but not received an answer. I will keep trying to find out from my SA.

I would be happy to have someone in Portland, OR with the lag drive my car to see if they can reproduce it. Just PM me.