Scatter Plots for Beginners

[whittlebeast]

I very seldom (if ever) get before and after logs off the same Harley.  With most people that contact me, I do one TeamViewer session to set them up and never hear from them again, let alone ever see a big sample log off the finished tune.

Jamie told me that he has very few matching before and after samples.

Andy

[FLTRI]

I very seldom (if ever) get before and after logs off the same Harley.  With most people that contact me, I do one TeamViewer session to set them up and never hear from them again, let alone ever see a big sample log off the finished tune.
Andy

Why couldn't you bring yourself to admit to us you only had broken tune data and no follow up (post bust fix) data to show us?
This also brings up the question: Were you always able to fix a tune bust the first shot at it?
Bob

[98fxstc]

Well , as a learner tuner DIY person, I read all my PV data logs on MLVHD , a bit like the second lot of graphs Jamie posted.
It's great. Reasonably Easy to read and adjustable to your requirements. ( color, subject matter, time frames )
And INEXPENSIVE.

Might be OK for PV but at present wont load the the latest DM3 files (as stated on the site) without converting to csv files
Will be looked at , but not there yet

[whittlebeast]

Bob, I never claimed to be a "Pro Harley Specific Tuner"  I help people realize that motors are a very predictable air flow device and I help them find the patterns.  I got a log from a guy yesterday trying to turbo a sporty.  It is rather apparent that he is up against a tough project.  His "Pro Tuner" apparently gave up so he contacted me. 

I said all along that I would love to see a log off a well tuned Harley.  I would love to take a look at the data coming off the CAN bikes.  The only one I have is a tune that is a total mess.  It turns out that the  o2s were swapped and ran the tune into a mess.

Andy

[Mirrmu]

don't understand why you can't get a log from a well tuned bike,

[whittlebeast]

Simply to show people how all well tuned motors have patterns, here is a motor pulling a load for 3 hours of data at a 2 samples per sec data rate.

http://www.nbs-stl.com/HarleyTuning/Honda%20Tuning%20Logic.png

On this motor, Mass Air Flow Rate is essentially the same as MAPxRPM. 
Fuel Flow Rate Per Hour is essentially Duty Cycle.
Engine load is MAP

The three distinct lines that developed on the left plot is the VVT setting in and a Harley would not have that.  Still it is a very clear pattern in the scatter plot.

The color on both plots is Air Fuel Ratio Commanded or Target Lambda on both plots.

All we need is a big log off a well tuned Harley and I can do the same thing for here.

Have fun tuning.

Andy

[rbabos]

don't understand why you can't get a log from a well tuned bike,

Think about that for a while and it will come to you why. A log is a direct representation of someones work.
Ron

[PoorUB]

don't understand why you can't get a log from a well tuned bike,

Think about that for a while and it will come to you why. A log is a direct representation of someones work.
Ron

First you will need to find a well tuned bike, then post the log here and have everyone else rip the tune to pieces!

[Tsani]

I offered to do that but you blew it. You failed to agree to a set of observation controls. You failed to respond by clearly explaining what you were posting. Yet someone else actually did in one post what you have yet to do. In doing things your way, you have failed to establish a level of trust needed between a teacher and a learner.  We all know there are patterns to be shown and some of us have been using that to our advantage long before you started posting here. After all this time here and elsewhere, it is hard to believe you have not seen a file off a well tuned Harley big twin. In my opinion it is due to the fact that Mlvhd isn't or cant be used as the sole tool to tune a Harley especially when you consider we do not have all the high dollar ECMs and sensors added on like your racing buds. Scatter plots are an aid, not a means to an end. No more, no less. And you still fail to clearly explain what you post.  You make a lot of claims but have proved nothing. Even if you used your Toyota truck example and clearly explained it with before and after plots would have been helpful. But you didn't.  As a matter of fact, I don't think I have seen you post a set of plots side by side of before and after, much less during, using the same x y z criteria. Then you have what Ron just said and you can add to it that it is not likely to happen with the level of distrust that has been built.

[whittlebeast]

Here are two shots of a well tuned Harley motor.

http://www.nbs-stl.com/HarleyTuning/Fairly%20Well%20Tuned%20Harley.png

vs one with a tune in trouble

http://www.nbs-stl.com/HarleyTuning/Scrambled%20Tuned%20Harley.png

Andy

[azlou66]

Andy
What changes would you make to the tune on motor two so it performs more like motor one?

[whittlebeast]

The scrambled one turned out to be reversed 02 sensors and the tune was chasing itself.

[Mirrmu]

don't understand why you can't get a log from a well tuned bike,

Think about that for a while and it will come to you why. A log is a direct representation of someones work.
Ron

First you will need to find a well tuned bike, then post the log here and have everyone else rip the tune to pieces!

happens with every other thread, we are talking about a log not a tune file

[PoorUB]

don't understand why you can't get a log from a well tuned bike,

Think about that for a while and it will come to you why. A log is a direct representation of someones work.
Ron

First you will need to find a well tuned bike, then post the log here and have everyone else rip the tune to pieces!

happens with every other thread, we are talking about a log not a tune file

My point was is everyone is looking for a log file for a well tuned bike, but around here "well tuned" turns into a bunch of differing opinions.

[whittlebeast]

Here is a set of screen shots I put together for MLV HD.  If you unzip them into a folder, you can then view them in sequential order.

https://www.dropbox.com/s/c1vn27jhj8ytbjn/MLV%20HD%20Demo%2011.zip?dl=0

I will be doing a demo at the Megasquirt meet in about a month.

http://megameet2015.com/

Andy

[whittlebeast]

Here is a version that is a little easier to view.

http://www.nbs-stl.com/MLVDemo/MLVHD%20Demo.pdf

Feel free to ask any questions.

Andy

[Just Nick]

I thought this pretty picture post died once questions never got answered. Are you ready to show us a bust in tune and the fix of the bust ?

[whittlebeast]

Nick

A dyno does not tell you why a motor has a huge dip in the torque curve.  It simply documents that there is a dip.

This software does the same thing.  It simply helps you locate the issue.  The big difference is, you don't need the $60,000 dyno.

You would be shocked at the number of people that thank me off line as they are apparently too intimidated to post publicly.

Have fun tuning with whatever tools you want to learn.

Andy

[Tsani]

Nobody is trying to intimidate you Andy. You just don't answer the questions. You do a little side shuffle or ignore the question, or answer the question with a example that is not relevant. And you have given your share of whiney snotty shots as well. What ever your doing off the forum is not the same as what you are doing on this forum. As for the dyno bit, a lot of us can't afford the Motec crap either along with all the sensors needed etc that your racer buddies are using. And until very recently, none of us had a water cooled engine and really still don't. You know, like that Toyota truck that you claim is like a Harley engine. NOT, by the way. But the biggest thing is:

While most of us agree that graphically displayed data is helpful, you have yet to give one supported before and after, clearly explained example based on a Harley Davidson Big Twin even after having been asked for it many times over.

And that my friend says volumes.

And we won't even get into the idea that there may be a flaw in MLVHD.

[whittlebeast]

It is interesting how few people in the Harley world are willing to post real world data.  I don't own a CAN Harley.

Andy

[turboprop]

Andy, where are you located? I am in NC and have a couple extra bikes I could loan you for a few months.

[Tsani]

It is interesting how few people in the Harley world are willing to post real world data.  I don't own a CAN Harley.

Andy

I don't own a Can Buss Harley either. J1850 was offered. Real easy to blame the "Harley World".  You couldn't/wouldn't even explain clearly what you did post.

[whittlebeast]

Turbo

Join us at the Megasquirt Meet next month.

http://megameet2015.com/

I get there Thursday night and stay thru most of Sunday.  Most everything happens Saturday.  The people that attend are amazing.  They are all simply addicted to tuning.  The programmers and board designers are also available to answer anything.  Last year, we were doing tuning demos with a 1200 HP Supra.

Bruce is doing a session on how wide bands really work.

Andy

[glens]

We already know how broadband sensor systems "really work" so no need to get that "talk" reported on...

[turboprop]

Turbo

Join us at the Megasquirt Meet next month.

http://megameet2015.com/

I get there Thursday night and stay thru most of Sunday.  Most everything happens Saturday.  The people that attend are amazing.  They are all simply addicted to tuning.  The programmers and board designers are also available to answer anything.  Last year, we were doing tuning demos with a 1200 HP Supra.

Bruce is doing a session on how wide bands really work.

Andy

I was planning on attending, unfortunately another event popped up that has trump.