[beepbeep]

Originally Posted by scalbert

No digging but the below was a stretch.

Not to counter stretch too much, but he did say "mini". All in all, the humidity discussion was informative.

[Ilma]

I was thinking about this graph and got to wondering why the throttle pedal voltage is showing full throttle right at the start of the datalog?

Normally one sees the throttle pedal plot as an upward curve as one depresses it at the start of the run.

Originally Posted by Mike@N54Tuning.com

Sure, here is one with tps voltage added and scaled so that 2v = 1000. Boost targeting is a function of barometric pressure, IAT, and other factors. Getting targeting to work well under various conditions is tricky business. It may work well in the cold climate here, but when the ECU shifts around the boost targets for hotter weather be completely different. If you could include DME target and DME actual on your charts we could infer from the target curve your IAT and barometric pressure.

Mike

[TheTwinz]

Originally Posted by Ilma

I was thinking about this graph and got to wondering why the throttle pedal voltage is showing full throttle right at the start of the datalog?

Normally one sees the throttle pedal plot as an upward curve as one depresses it at the start of the run.

well the graph does start at 2500 rpms, so... i think he shifted into 3rd and had the rpms go down , then floored it ..

[Mike@N54Tuning.com]

Originally Posted by Ilma

I was thinking about this graph and got to wondering why the throttle pedal voltage is showing full throttle right at the start of the datalog?

Normally one sees the throttle pedal plot as an upward curve as one depresses it at the start of the run.

The logs were cropped to exclude any non-WOT activity, but here is the same log with more data going out on either end. Just makes things harder to analyze.

Mike

[Ilma]

Originally Posted by Mike@N54Tuning.com

The logs were cropped to exclude any non-WOT activity, but here is the same log with more data going out on either end. Just makes things harder to analyze.

Mike

Showing the complete data for the run is more transparent and informative and prevents people from jumping to incorrect conclusions.

Unfortunately, as others have pointed out, the lack of good graphics presentation is also making things hard to analyze.

If this is the same graph as before with simply more expansion of the data points, I don't understand why the colours for timing and throttle blade are suddenly reversed.

That is confusing to me when comparing the before and after graphs, and it requires menu manipulation of the colour choices chosen for the source data.....not necessary to simply extend existing data ranges in order to remove any cropping effects.

[TheTwinz]

Originally Posted by Ilma

If this is the same graph as before with simply more expansion of the data points, I don't understand why the colours for timing and throttle blade are suddenly reversed.

That is confusing to me when comparing the before and after graphs, and it requires menu manipulation of the colour choices chosen for the source data.....not necessary to simply extend existing data ranges in order to remove any cropping effects.

I think either you are blind, don't know how to read logs, or you are trying to find errors with the logs that aren't there. The colors are the same, and it basically IS the same log... why so critical? Maybe you should learn how to read graphs before making statements trying to discredit the OP.

[Ilma]

Originally Posted by TheTwinz

I think either you are blind, don't know how to read logs, or you are trying to find errors with the logs that aren't there. The colors are the same, and it basically IS the same log... why so critical? Maybe you should learn how to read graphs before making statements trying to discredit the OP.

Actually your words are autobiographical and you are wrong.

[munters]

Originally Posted by Ilma

Actually your words are autobiographical and you are wrong.

+1

[TheTwinz]

Originally Posted by Ilma

Actually your words are autobiographical and you are wrong.

ok, my bad, i'm blind... , but I do believe the logs are the same.

[Mike@N54Tuning.com]

Originally Posted by Ilma

Showing the complete data for the run is more transparent and informative and prevents people from jumping to incorrect conclusions.

Unfortunately, as others have pointed out, the lack of good graphics presentation is also making things hard to analyze.

If this is the same graph as before with simply more expansion of the data points, I don't understand why the colours for timing and throttle blade are suddenly reversed.

That is confusing to me when comparing the before and after graphs, and it requires menu manipulation of the colour choices chosen for the source data.....not necessary to simply extend existing data ranges in order to remove any cropping effects.

I had intended this thread for those familiar with analyzing logs. For example, looking at the DME boost target tells you the car was at WOT. Showing before and after run data throws off the scaling making it harder to analyze.

The colors are different because I didn't save the last file and had to remake the chart. Really I'm not sure where the confusion is coming from. :

Mike

[Former_Boosted_IS]

Originally Posted by Mike@N54Tuning.com

I had intended this thread for those familiar with analyzing logs. For example, looking at the DME boost target tells you the car was at WOT. Showing before and after run data throws off the scaling making it harder to analyze.

The colors are different because I didn't save the last file and had to remake the chart. Really I'm not sure where the confusion is coming from. :

Mike

Mike, just add a secondary axis scale on the right for timing and it will be much, much easier to quantify exactly what we are looking at. If you send me the file, I can do it for you.

[Ilma]

Originally Posted by Mike@N54Tuning.com

I had intended this thread for those familiar with analyzing logs. For example, looking at the DME boost target tells you the car was at WOT. Showing before and after run data throws off the scaling making it harder to analyze.

The colors are different because I didn't save the last file and had to remake the chart. Really I'm not sure where the confusion is coming from. :

Mike

Oh...that's right!

I forgot from our conversation at your condo that you mentioned you are colour blind - my bad.

I agree that bad scaling can be distracting. The scaling spikes can be controlled by limiting the upper and lower ranges of the y-axis. Just right-click on any of the numbers showing on the y-axis and choose the "Format axis" menu. Enter zero for the minimum value (which will cut off the negative timing values) and choose a maximum value that works with the majority of your data.

I recall pointing out to you in your parking lot that DME boost is not the actual boost when read by the BT tool. Piggybacks fool the DME as to those values. So how is anyone going to know anything about the boost on your graph other than it's relative position to target.

Besides, you scaled the y-axis as 1600 units of something for boost pressure and don't provide a legend or cross-reference, so it's left to interpretation. I know stock boost is around 9 psi and tuned boost around 15-16 psi.

But I see by your response that you infer I am not familiar enough with datalogs to understand

It looks suspect when you keep recreating your graph and swap colours around and omit data from the start and end of the run.

I'm skeptical of your excuse because you would have to save the file in order to be able to post it here. That's how excel works....the chart is imbedded in the worksheet itself.

If you are going to attack the credibility of the Procede data being posted then seems only fair that you should be subject to that same standard.

I don't claim to have any wisdom about timing control, but this topic has the same tone as the torque targeting smear campaign did, and I bought into that one and campaigned for JB3 until I tried a Procede for myself and started data logging on my own.

At that time, I recall letting you take my procede equipped car for a test drive and saw that you were impressed and that you commented on it's the bottom end driveability. And we both concurred that the JB had more top-end.

So in my experience, BMS has implemented and adopted some of the very same features that they criticized in the Procede as being smoke and mirrors.

If you can get me to recreate these results on my car, then I will have no choice but to acknowledge your claims.

But anyone can datalog in certain ways to emphasize or minimize the results they want to skew.

After all, isn't that the point of your thread? To reveal that Shiv is selecting choice datalogs to prove timing control?

Shouldn't a competing vendor who is attempting to prove claims of misinformation be subject to the same level of transparency ?

[brianhn1]

Originally Posted by Ilma

Oh...that's right!

I forgot from our conversation at your condo that you mentioned you are colour blind - my bad.

I agree that bad scaling can be distracting. The scaling spikes can be controlled by limiting the upper and lower ranges of the y-axis. Just right-click on any of the numbers showing on the y-axis and choose the "Format axis" menu. Enter zero for the minimum value (which will cut off the negative timing values) and choose a maximum value that works with the majority of your data.

As you know, with a piggyback DME boost is not the actual boost when read by the BT tool. Piggybacks fool the DME as to those values.

Besides, you scaled the y-axis as 1600 units of something for boost pressure and don't provide a legend or cross-reference, so it's left to interpretation. I know stock boost is around 9 psi and tuned boost around 15-16 psi.

But I see by your response that you infer I am not familiar enough with datalogs to understand

It looks suspect when you keep recreating your graph and swap colours around and omit data from the start and end of the run.

I'm skeptical of your excuse because you would have to save the file in order to be able to post it here. That's how excel works....the chart is imbedded in the worksheet itself.

If you are going to attack the credibility of the Procede data being posted then seems only fair that you should be subject to that same standard.

I don't claim to have any wisdom about timing control, but this topic has the same tone as the torque targeting smear campaign did, and I bought into that one and campaigned for JB3 until I tried a Procede for myself and started data logging on my own.

At that time, I recall letting you take my procede equipped car for a test drive and saw that you were impressed and that you commented on it's the bottom end driveability. And we both concurred that the JB had more top-end.

So in my experience, BMS has implemented and adopted some of the very same features that they criticized in the Procede as being smoke and mirrors.

If you can get me to recreate these results on my car, then I will have no choice but to acknowledge your claims.

But anyone can datalog in certain ways to emphasize or minimize the results they want to skew.

After all, isn't that the point of your thread? To reveal that Shiv is selecting choice datalogs to prove his timing control?

Shouldn't a competing vendor be subject to the same level of transparency?

Besides the color change between the throttle blade and timing graphs, the data looks the same to me. Obviously the DME receives boost dictated by the piggyback. About everything else you have said unrelated is too darn vague and makes no sense.

[Mike@N54Tuning.com]

Originally Posted by Ilma

I'm skeptical of your excuse because you would have to save the file in order to be able to post it here. That's how excel works....the chart is imbedded in the worksheet itself.

I made and cut/pasted the chart in to paint as a JPG, and closed excel without saving it. I'm sure you have good intentions towards transparency but like I said the DME boost target tells you the same information. The ECU only calls for 140+ KPA of boost at wide open throttle. With your charts and Shiv's it would be nice to see DME boost target and DME boost if we're going to have a transparent analysis.

Anyway the purpose of this thread is to discuss timing control and how one can establish whether or not CPS is effective.

Mike

[jpsimon]

Technically speaking... was your/terry's theory about it being "learned out" part of the "effectiveness" argument? Or did you decide that smear campaign was "learned out"?

[E82tt6]

I think the following would be helpful:
actual timing JB3 and procede overlayed for 4 4th gear dyno (or 3rd gear road) pulls, same car, same day.

That would end all of it. We'd get to see if adaptation changes anything. It should be done on street fuel.

[Mike@N54Tuning.com]

Originally Posted by E82tt6

I think the following would be helpful:
actual timing JB3 and procede overlayed for 4 4th gear dyno (or 3rd gear road) pulls, same car, same day.

That would end all of it. We'd get to see if adaptation changes anything. It should be done on street fuel.

The problem with that is boost levels, etc, may be different. Using the same system and enabling/disabling CPS is the best method I think. That way all other variables are kept even. There must be several runs on each map to allow for adaption time. It would also be interesting to test and overlay a JB3 log to see if it matches the CPS disabled overlay of the V3 and to what degree, and of course to compare the actual power differences. I'll arrange for this all to take place but it may take a week or two. Will practice my excel skilz before then as well.

Also there is a wealth of information available already just looking at the various logs presented here and in other threads. The problem is we have few people who really want to analyze data, and lots of people who want to jump to some conclusion and claim victory. Three times now I've seen logs posted claiming to prove something that they don't. It isn't enough to just post logs and hope they support your position, you need to analyze them.

Mike

[E82tt6]

If you do, I'd try to round up a local forum member so no one cries foul.

I suggested multiple runs to show adaptation. The problem with a lot of logs is that they're on the street, and the ecu is so quick to make timig adjustment that they aren't always consistent run to run.

[ktdw]

Originally Posted by Mike@N54Tuning.com

Three times now I've seen logs posted claiming to prove something that they don't. It isn't enough to just post logs and hope they support your position, you need to analyze them.

Mike

Just to be clear, the three logs you posted were actually the same log presented differently. But you're right in that you need to also provide some detailed analysis around them that supports your position.

[deamon32]

Originally Posted by ktdw

Just to be clear, the three logs you posted were actually the same log presented differently. But you're right in that you need to also provide some detailed analysis around them that supports your position.

+1

I own a JB3 and am very happy with it, but over a week ago Mike made some claims and promised to back up those claims with data. Those claims were that monitoring knock through just the total ignition time advance is not always sufficient and that the DME learns around the CPS offset that the procede uses.

So far the logs that have been posted really don't address either of those two points and as far as I am concerned the ball has been in your court for several weeks now and during that time Shiv and other procede users have posted countless logs while we are still looking at the same two logs from N54Tech.

Starting to get a little bit frustrated as a JB3 user, I really don't care about sides but if one side is going to make claims don't let everyone hang in the wind while it takes you several weeks to gather logs to back up those claims.

[jpsimon]

just posted some new logs from today:
http://www.e90post.com/forums/showthread.php?t=286341

I only changed the ign. correction value to clearly demonstrate the effects on ign. advance.

[Mike@N54Tuning.com]

Originally Posted by jpsimon

Technically speaking... was your/terry's theory about it being "learned out" part of the "effectiveness" argument? Or did you decide that smear campaign was "learned out"?

The debate on whether or not they are learned out has come down to whether or not the CPS offset applied is capable of maxing out the stock timing values. If it can't, then the timing values will float based on knock feedback as designed. Meaning your car will experience "knock" (not exactly but when in Rome do as the Romans) as the ECU attempts to raise its timing values. If the CPS applied can max out the factory values, then the ECU has no room to raise them further and you can make the case that the CPS offset reduced effective timing.

That is why these logs are so critical. They show CPS tuned timing values no where near being maxed out, as do all customer logs I've seen posted.

Taking your logs as an example, your charts did not reflect maxed out (stock like) timing values, therefore your CPS offset was learned out. Actually your logs with less offset resulted in lower (and equally smooth) timing values which is totally bizarre. While we're on the topic, you made the claim that your logs proved CPS offsetting was not learned out. Could you walk me through your analysis of that?

Mike