March 28, 2024, 02:27:26 AM

News:

For advertising inquiries or help with registration or other issues, you may contact us by email at help@harleytechtalk.com


Scatter Plots for Beginners

Started by whittlebeast, August 05, 2015, 07:45:07 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

whittlebeast

Quote from: turboprop on August 08, 2015, 02:59:42 PM
Andy, has the MegaSquirt hardware been packaged yet into a bolt-on plug and play system for Harley yet?

They have no interest in building a tuning device.  If Power Vision, TTS, or anyone else would want to do the effort, the other piece of software that they make could be utilized to work with any piece of interfacing hardware.  Knowing and keeping up with the ever changing stock ECUs is the part MLV management want nothing to do with.  Tuner Studio is good.

Andy
Dynos are great for getting the motor close enough to get on the data loggers.

FLTRI

August 08, 2015, 06:16:28 PM #26 Last Edit: August 08, 2015, 06:53:18 PM by FLTRI
Andy, can I assume you cannot or will not provide us with what I requested?
Nope! Don't need Harley data to do the following:

"NOW! PLEASE! Take ONE friggin scatter graph of yNope! Don't need Harley data to do the following:
"NOW! PLEASE! Take ONE friggin scatter graph of your choice and circle or point to where the bust you identify is in the graph and tell us what the bust is and how you see that. SIMPLE!
THEN! PLEASE! Show us the scatter graph with the fix to the bust and circle or point to where the picture changed after the bust was fixed. SIMPLE!"
Bob
The best we've experienced is the best we know
Always keep eyes and mind open

turboprop

Andy, arent there guys out there running some form of MegaSquirt or MiniSquirt on Harleys? Did they have to encase and connectorize the stuff themselves? 

I swear I have seen this stuff being used on some serious harley's at Bonneville or maybe it was Ohio.
'We' like this' - Said by the one man operation.

whittlebeast

They are running Microsquirts and those come with waterproof connectors.  I think they are running full sequential spark.  A few years ago, I talked the programmers into getting the MAP read to figure out the spark sequence.

Andy
Dynos are great for getting the motor close enough to get on the data loggers.

whittlebeast

Bob,

http://www.nbs-stl.com/HarleyTuning/TTS%20Tune%2001.png

The red stuff is a real mess.  On the plot, everything should be real close to a straight line.  Being that the the out of character part of the data, is above the straight line, it implies richer then you would expect.  Al  I would do at that point is hit any one point on the scatter plot and then go to the log viewer view and see what you were doing at the time.

That motor is knocking bad at the same point as you can see on the right plot.  Also the red dots.

Andy

Dynos are great for getting the motor close enough to get on the data loggers.

scott45acp

Why rpm x map?  What makes that useful?  Also what is the duty cycle of?  Injector on time?

Since this is scatter plots for dummy could i send a log file so i can follow along at home?   I could not duplicate your example since sportsters dont have a knock retard.

whittlebeast

MAP X RPM is something I found a few years ago when testing race cars the had MAF sensors mounted for the data purposes.  We found the if you created a new field that was simply the MAP times the RPM, it was almost a perfect representation of the air going into the motor.

Duty Cycle is the percentage of time the injectors are spraying.  In the Harley world it tends to max out at like 50%.  Way less than most motors.

Andy
Dynos are great for getting the motor close enough to get on the data loggers.

scott45acp

That make sense then. So air flow vs fuel delivery would be a straight line provided afr was the same thought out the range.   I ploted mine using hits as the Z axis.  Several outliners but on both sides of the line.   Do you kind of average those out and go by the trend?

FLTRI

Quote from: whittlebeast on August 08, 2015, 07:20:11 PM
Bob,

http://www.nbs-stl.com/HarleyTuning/TTS%20Tune%2001.png

The red stuff is a real mess.  On the plot, everything should be real close to a straight line.  Being that the the out of character part of the data, is above the straight line, it implies richer then you would expect.  Al  I would do at that point is hit any one point on the scatter plot and then go to the log viewer view and see what you were doing at the time.

That motor is knocking bad at the same point as you can see on the right plot.  Also the red dots.

Andy
Ok, we have picture of a shitty running engine, right? It's richer and leaner in places with bad knocking.
You fixed these conditions, right?
Now, if you will post the same engine same graph after the bust fiix along side the before fix we have what we can plainly see in contrast.
Thanks,
Bob
The best we've experienced is the best we know
Always keep eyes and mind open

98fxstc

Andy , your inbox is full.
must be popular  :teeth:

whittlebeast

August 09, 2015, 04:29:24 AM #35 Last Edit: August 09, 2015, 04:31:38 AM by whittlebeast
Quote from: FLTRI on August 08, 2015, 09:39:37 PM
Quote from: whittlebeast on August 08, 2015, 07:20:11 PM
Bob,

http://www.nbs-stl.com/HarleyTuning/TTS%20Tune%2001.png

The red stuff is a real mess.  On the plot, everything should be real close to a straight line.  Being that the the out of character part of the data, is above the straight line, it implies richer then you would expect.  Al  I would do at that point is hit any one point on the scatter plot and then go to the log viewer view and see what you were doing at the time.

That motor is knocking bad at the same point as you can see on the right plot.  Also the red dots.

Andy
Ok, we have picture of a shitty running engine, right? It's richer and leaner in places with bad knocking.
You fixed these conditions, right?
Now, if you will post the same engine same graph after the bust fiix along side the before fix we have what we can plainly see in contrast.
Thanks,
Bob

Jason was tuning that motor.  I never saw what he did to fix it or what the plots looked like when he was done.
Dynos are great for getting the motor close enough to get on the data loggers.

whittlebeast

August 09, 2015, 04:58:14 AM #36 Last Edit: August 09, 2015, 05:32:48 AM by whittlebeast
Here a plot from a guy that I got yesterday.  The first thing I will have him do is pull 8% from the VE tables from the upper right corner of the plots.  That works out to about 5250 RPM and above and 85 KPA and above.  He is going to retest today, weather permitting.

http://www.nbs-stl.com/motec/Motec%20Running%20Fat%20Up%20Top.png

I looked back at a previous log I had and the top end stuff looked way better...  I wonder what changed?  Weather, Fuel Pressure, closed loop misbehaving....
Dynos are great for getting the motor close enough to get on the data loggers.

whittlebeast

A member just sent me several TTS logs as he was on a long ride.  Here is a couple of screen shots of what lots of data looks like in scatter plots.  This is like 7 logs open all at the same time.  This worked out to 45000 total records.

http://www.nbs-stl.com/HarleyTuning/HarleySpeedDensityVsAlphaN.png

and

http://www.nbs-stl.com/HarleyTuning/HarleyScatterPlots01.png

The only thing that concerns me is on the MAP X RPM vs Duty Cycle1 (really front DC) gets messy up and to the right of the circle.

Andy
Dynos are great for getting the motor close enough to get on the data loggers.

whittlebeast

August 09, 2015, 07:27:01 AM #38 Last Edit: August 09, 2015, 08:23:37 AM by whittlebeast
For people with MLV HD, here is the data I plotted above.

http://www.nbs-stl.com/HarleyTuning/Lots%20of%20TTS%20Date.zip

Open the SW and then open all of these at the same time from File Open

Have fun tuning
Dynos are great for getting the motor close enough to get on the data loggers.

whittlebeast

Just for reference and complarison, here is what data looks like coming off a dyne.

http://www.nbs-stl.com/HarleyTuning/TTS%20Tune%2001.png
Dynos are great for getting the motor close enough to get on the data loggers.

FLTRI

Thanks for the pictures Andy.
When you post a picture with a statement how on earth do we know what and where to look?
You simply tell us what we are looking at. Don't you understand it is just a picture until you share your understanding of what is broken and how the picture changes from fixing the tune.

Show us the contrast of the before and after picture, SIDE BY SIDE from fixing a tune.

I appreciate you've been studying scatter graphs for years and can see a tune bust from the next room but we, at least I, need contrast to see the pictures of bust-to-fixed graphs.

I've also spent many years looking at data (in the racing arena), including scatter graphs BUT we looked at these graphs, not for the holy grail of tuning, but for trends, which is what I feel is their true value. You feel they have a lot more value so that's why I've asked NUMEROUS times, right here on this thread, for the before fix and after fix pictures. Nothing more, nothing less!

Why won't you do it? :scratch: :scratch:

Maybe because you don't have before and after pictures?  :nix: If so, they are nothing more to me but pretty pictures and I don't want to spend unlimited hours studying pictures looking for something I wouldn't see by tuning a bike without these picture graphs.
Come on Andy show us the direct comparison pictures so we can understand what you keep saying.
Bob
The best we've experienced is the best we know
Always keep eyes and mind open

turboprop

Andy, I too really appreciate the stuff you have been posting. Good stuff. Keep it coming.
'We' like this' - Said by the one man operation.

FLTRI

Quote from: turboprop on August 09, 2015, 10:47:01 AM
Andy, I too really appreciate the stuff you have been posting. Good stuff. Keep it coming.
Turbo,
I assume, based on your comment, you understand what you are looking at. :up:
I realize I may be a bit slow but I'm looking for before and after pictures so I can keep up.
Maybe you can explain how tuners can effectively utilize these graphs to do better tune a bike.
I do see it as a handy tool to identify trends (as we did in, especially, endurance racing) but that's about it and I'm not seeing the advantage to actually tuning a bike. :nix:

Bob
The best we've experienced is the best we know
Always keep eyes and mind open

whittlebeast

This is the sort of thing you find when you are willing to look.

http://www.nbs-stl.com/HarleyTuning/Harley_SampleFuelMap.png

Interesting where in the maps this rider spends most of his time.

http://www.nbs-stl.com/HarleyTuning/HarleySpeedDensityVsAlphaN.png
Dynos are great for getting the motor close enough to get on the data loggers.

turboprop

Not an expert, and I probably only understand a smidgen of this stuff compared to Andy. Not really sure these plots can be directly transfered over to a harley tuning device, ie; I see this happening so a will adjust this cell type of thing. Instead I see it as a way to see the effects of changes to the tune. And like Andy says, it pretty easy to see if something is wrong. The graphic representation of these scatter plots is for me much easier to visualize whats going on.

Perhaps if you were to collect some data on your dyno, look at it with megaviewer and then tune the bike and relook at the scatter plots, you would be able to see clearly the cause and effect in the scatter plot. Try it. What have you got to loose?
'We' like this' - Said by the one man operation.

Onthefence

How can we use MLV to fine tune to increase hp and or torque.   We always hear give it what it likes.  Those with a dyno have methods and tools to measure those changes.  Can i replicate with logs?

FLTRI

Quote from: turboprop on August 09, 2015, 11:49:41 AM
Perhaps if you were to collect some data on your dyno, look at it with megaviewer and then tune the bike and relook at the scatter plots, you would be able to see clearly the cause and effect in the scatter plot. Try it. What have you got to loose?
Turbo,
The purpose of this thread is to teach us the basics. Not the other way around. Look at the title of the threads.

After years of collecting, plotting, and studying scatter graphs I just assume it would be very easy for Andy to put 2 graphs side by side ...one before a tune bust fix, the one after that tune's bust fix.

If you would like to explain and show us some before and after scatter plots to illustrate the difference in appearances before and after tuning that would be great too.   :soda:

After all, we're the students here and would like the teacher to show us what the point of these graphs/plots are as applied to tuning...which is what Andy has indicated is an invaluable tuning tool. JUST SHOW US HOW TO USE IT BESIDES TELLING US "look at this".

Bob
The best we've experienced is the best we know
Always keep eyes and mind open

turboprop

August 09, 2015, 02:28:30 PM #47 Last Edit: August 09, 2015, 02:34:33 PM by turboprop
Never mind.
'We' like this' - Said by the one man operation.

whittlebeast

The proof is in the silky smooth power delivery and instant throttle response from any load.  That is all difficult to document but you know when you get it right.  I like to help people learn how to get it right.

Always remember, nobody wants your bike to run great more than you.  Nobody else cares to spend the effort.

Andy
Dynos are great for getting the motor close enough to get on the data loggers.

Onthefence


Quote from: whittlebeast on August 09, 2015, 04:00:18 PM
The proof is in the silky smooth power delivery and instant throttle response from any load.  That is all difficult to document but you know when you get it right.  I like to help people learn how to get it right.

Always remember, nobody wants your bike to run great more than you.  Nobody else cares to spend the effort.

Andy

There are already plenty providers in the tuning space,  who can put the magic touch to a tune.  Since it is a paid service, they are not going to give everything away.

How do we gain from what you are trying to teach if it's all about your secret sauce.

For example.  The latest scatter graph shows a big pile of hits in the rpm/map/ and tp for cruise.  What does that tells us about the state of tune?   

Wouldn't we get more valuable data by logging afr and viewing in a histogram view that is either map-rpm or tps-rpm.