Scatter Plots for Beginners

[whittlebeast]

Here is just another Sunday afternoon, collecting a little data.

https://www.youtube.com/watch?v=3iNdHm81lgs

Here is what 40 sec of data looks like off that sort of data collection

http://www.nbs-stl.com/CRX/CRX%20Autocross%20Trace.png

Have fun tuning.

Andy

[whittlebeast]

Here is the amount of data and map coverage I get in only 5 min of real world driving with my new setup.  And yes, I am thrashing hard on the motor..

http://www.nbs-stl.com/CRX/CRX%20Autocross%20Scatter%20Plot.png

When a Dyno guy does a WOT sweep and hand you a dyno sheet, here is that amount of the map he covers.

http://www.nbs-stl.com/CRX/Dyno%20Pull%20AFR.png

Andy

[Admiral Akbar]

You know Andy, this stuff all looks pretty cool but what does it mean?  Are we supposed to eat psychedelics before viewing?

I've asked the very same question in an engineering meeting and got a good answer.. So

[1FSTRK]

Here is the amount of data and map coverage I get in only 5 min of real world driving with my new setup.  And yes, I am thrashing hard on the motor..

http://www.nbs-stl.com/CRX/CRX%20Autocross%20Scatter%20Plot.png

When a Dyno guy does a WOT sweep and hand you a dyno sheet, here is that amount of the map he covers.

http://www.nbs-stl.com/CRX/Dyno%20Pull%20AFR.png

Andy

Ok, I see where you are going with the volume of data and while it does show there are changes in climate, it does not prove it is warming or that man is the definitive cause of it. I can only conclude it is not the volume of data that counts, it is the accuracy with which it is interpreted that yields useful results but maybe I am just in the wrong place at the wrong time here.

[whittlebeast]

Tuning a motor really is a little like your analogy.  You may not know exactly the cause but you can clearly see a pattern if you are willing to look.  If the Google earth pics of the forest used to be green and now they are brown, something is going on.

In the case of a motor, rich patterns and lean patterns can be AE/DE related or it can be VE related, it is a little difficult to be sure.  What you can see is a clear pattern where the color (AFR) is not what you expect.

Here is the same screen shot but I have highlighted the areas where I have concerns.  Keep in mind that this motor was tuned on a dyno before I ever raced it.

http://www.nbs-stl.com/CRX/CRX%20Autocross%20Scatter%20Plot%20Boxed.png

On the left is speed density with RPM and MAP and in the center is the Alpha-N plot with RPM and TPS.  The boxes on both sides match by color.  The red and purple boxes are rich (blue) and the gold and green boxes are quite lean (red).

Do I have everyone so far?

[harleytuner]

Here is the amount of data and map coverage I get in only 5 min of real world driving with my new setup.  And yes, I am thrashing hard on the motor..

http://www.nbs-stl.com/CRX/CRX%20Autocross%20Scatter%20Plot.png

When a Dyno guy does a WOT sweep and hand you a dyno sheet, here is that amount of the map he covers.

http://www.nbs-stl.com/CRX/Dyno%20Pull%20AFR.png

Andy

You never adress the fact that you are riding/driving out of tune while you are collecting large data.  I'll take a short accurate data run on an untuned engine on a dyno over a large data run on an untuned engine on the street any day. On other thing you clearly don't understand is we can do large data runs on a dyno in a controlled enviornment.  You seam to generalize all dyno operators as people that just do sweeps.  I realize that the products you sell/endorse are in direct competition with dyno's and you have to market against them.   

[whittlebeast]

Did you miss that this motor was tuned on a dyno with plenty of sweeps.  This is taking tuning to the next step once the dyno tuning is done.  At least for 90 percent of the riding time.  Sure, a dyno can get to places that a mild mannered rider seldom if ever gets to.  In the case of this motor. I am getting to places and conditions that simply can not be duplicated on a dyno.

[Admiral Akbar]

Did you miss that this motor was tuned on a dyno with plenty of sweeps.  This is taking tuning to the next step once the dyno tuning is done.  At least for 90 percent of the riding time.  Sure, a dyno can get to places that a mild mannered rider seldom if ever gets to.  In the case of this motor. I am getting to places and conditions that simply can not be duplicated on a dyno.

I think he's missing the fact that you don't drive motorcycles on a dyno.. The aren't any stop signs, corners, speed limits etc..

[Admiral Akbar]

Here is a similar plot for one of my bikes.. I cut adjusted the range to go from 12 to 15 instead off 11 to 15 mainly to increase contrast one the AFR..

[attach=0]

[whittlebeast]

Max, that is really good.  Post up MAPxRPM vs Duty Cycle front and rear.

Also post up the MAP vs RPM and AFF Front and Rear.

I would love to see how your tune is coming along.

[harleytuner]

Did you miss that this motor was tuned on a dyno with plenty of sweeps.  This is taking tuning to the next step once the dyno tuning is done.  At least for 90 percent of the riding time.  Sure, a dyno can get to places that a mild mannered rider seldom if ever gets to.  In the case of this motor. I am getting to places and conditions that simply can not be duplicated on a dyno.

I think he's missing the fact that you don't drive motorcycles on a dyno.. The aren't any stop signs, corners, speed limits etc..

Which one of these can not be done on a dyno?  I don't need a stop sign to stop, I can hit any speed an maintain it, and cornering has no effect on a tune at all (it's just acceleration and deceleration, and on a dyno I can do it at different loads which would simulate uphill and downhill). 

[whittlebeast]

What dyno can hold 100 mph wheel speed at 0 throttle?

[harleytuner]

Did you miss that this motor was tuned on a dyno with plenty of sweeps. 

Tuned on a dyno but by who and how and with what?  Were your pictures of the "dyno sweeps" before the tune or after the tune.  Anybody can show data to support their argument.  I've been tuning for over 15 years, averaged 364 full tunes a year from '09 to '14, find me the un-satisfied customers out their that I tuned.

[harleytuner]

What dyno can hold 100 mph wheel speed at 0 throttle?

That's one hell of a downhill to hit on the street.  The dyno I ran at Roeders could though to answer your question.  It would take 2 bikes to do it but it damn sure could

[whittlebeast]

Here is a sweep checking WOT fuel pressure.  That is a NA 1488cc motor built for torque so the low HP peak.  The same motor on a roller dyno would be right at 142 hp.  It has a fairly mild cam.  Right now we redline it at 8700.  We kept bouncing it off 8100 

https://www.youtube.com/watch?v=uABr5xN0uPE

http://www.nbs-stl.com/CRX/CRX%20Dyno%20Graph.jpg

http://www.nbs-stl.com/CRX/Dyno%20Pull%20Graph.png

Andy

[Admiral Akbar]

Did you miss that this motor was tuned on a dyno with plenty of sweeps.  This is taking tuning to the next step once the dyno tuning is done.  At least for 90 percent of the riding time.  Sure, a dyno can get to places that a mild mannered rider seldom if ever gets to.  In the case of this motor. I am getting to places and conditions that simply can not be duplicated on a dyno.

I think he's missing the fact that you don't drive motorcycles on a dyno.. The aren't any stop signs, corners, speed limits etc..

Which one of these can not be done on a dyno?  I don't need a stop sign to stop, I can hit any speed an maintain it, and cornering has no effect on a tune at all (it's just acceleration and deceleration, and on a dyno I can do it at different loads which would simulate uphill and downhill).

Can you program in a 6th gear to 3rd gear decel, then simulate an up hill grade through back to 5th, a long roll off down hill coast in 3rd gear where the bike is in 3rd with throttle shut off but the down hill is keeping the Rs at 3500?  Can you set up a constant load on the drum (not drum speed)? Simulate different weights of the bike?

If the answer is yes, then do you have a set of programmed load and throttle points that you use to validate you tune?  Or do you test ride?

Might want to try riding a bike on nice curvy road sometime. 

[Admiral Akbar]

Max, that is really good.  Post up MAPxRPM vs Duty Cycle front and rear.

Also post up the MAP vs RPM and AFF Front and Rear.

I would love to see how your tune is coming along.

This OK for the first one?

[attach=0]


What is AFF?

[whittlebeast]

AFF is your adaptive fuel factor, or long term fuel trim.  It is how hard the ECU has found it has to lean on the VE to get the AFR match the Target AFR.  If that comes up close to 100 everywhere, your fuel maps are golden.

Andy

[harleytuner]

Did you miss that this motor was tuned on a dyno with plenty of sweeps.  This is taking tuning to the next step once the dyno tuning is done.  At least for 90 percent of the riding time.  Sure, a dyno can get to places that a mild mannered rider seldom if ever gets to.  In the case of this motor. I am getting to places and conditions that simply can not be duplicated on a dyno.

I think he's missing the fact that you don't drive motorcycles on a dyno.. The aren't any stop signs, corners, speed limits etc..

Which one of these can not be done on a dyno?  I don't need a stop sign to stop, I can hit any speed an maintain it, and cornering has no effect on a tune at all (it's just acceleration and deceleration, and on a dyno I can do it at different loads which would simulate uphill and downhill).

Can you program in a 6th gear to 3rd gear decel, then simulate an up hill grade through back to 5th, a long roll off down hill coast in 3rd gear where the bike is in 3rd with throttle shut off but the down hill is keeping the Rs at 3500?  Can you set up a constant load on the drum (not drum speed)? Simulate different weights of the bike?

If the answer is yes, then do you have a set of programmed load and throttle points that you use to validate you tune?  Or do you test ride?

Might want to try riding a bike on nice curvy road sometime.

When you tune you are calibrating the ECM to hit target areas in cells, weather it's VE, spark, etc.  Once the cells are calibrated it doesn't matter weather or not you are accelerating or decelerating, the target is the same going through the areas, AI & DE come into play.

Can you hit 2,5,7,10,15,20,25% TP at 1250 Rpm's and hold it on the street long enough to get good data? 1500RPM? 1750RPM? Etc.  How long after the combustion event takes place does is data recorded on street tuning?  When using load on a dyno we can hold an area as long as it takes to get an accurate reading. 


My shop is 80' above sea level, I can be at the Blue Ridge Parkway and the Shenandoah mountain area in about 45 minutes.  I ride there often, as well as many of my costumers.  So yes, I hit the twisties quit often.

[whittlebeast]

Lot's people are shocked when they see data coming off their own bike as to where they really spend time running the motor.  What happens when they go up one of those mountain passes.  Or sit in heavy traffic and then roll up the highway entrance with a low fuel level.  I have even seen motors ridden with a 30 MPH tail wind wile climbing a mountain pass.  Strange things happen in the real world.

Andy

[Admiral Akbar]

Lot's people are shocked when they see data coming off their own bike as to where they really spend time running the motor.  What happens when they go up one of those mountain passes.  Or sit in heavy traffic and then roll up the highway entrance with a low fuel level.  I have even seen motors ridden with a 30 MPH tail wind wile climbing a mountain pass.  Strange things happen in the real world.

Andy

    While you can fill the fuel maps up with static loading, you can't see what happens under all conditions. Examples include engine temps, changes in barometric pressure different combinations of load.  After all the dyno can only tune under fixed conditions. Also, it's a good way to validate the tune and make sure important fringe cells and settings were not missed.

[whittlebeast]

This is really cool.  Did you ever wonder where you spend most of your time in the VE Map.  This happens to be 7 autocross runs with 3 different drivers.  Data was collected in about 5 min of driving.

http://www.nbs-stl.com/CRX/Autocross%20CRX%20RPM%20MAP%20and%20Time.png

The blue lines are where the motor passed thru the map.

Andy

[whittlebeast]

Have you ever wondered why all high HP NA motors run ITBs?  I am about to take everything I have shared with you guys and take it to an entirely new level of info of how motors work.  The motor in all of these screen shots is a relatively low reving motor with a torque peak of 5500 RPM and max power is at 7200 RPM.

Keep in mind that with a Harley, if you plot MAPxRPM compared to DutyCycle, you almost always get a straight line.  This rule works when the intake and exhaust never realy get into resonance.  They are really just pipes to contain the intake mixture and a pipe to channel the hot exhaust.  Double the RPM and the flow doubles.  Double the MAP and the flow doubles.  With ITBs all the rules change.

Let's look at what the MAPxRPM vs DC looks like if we limit data below 85 KPA and Limit the huge throttles to just 15% throttle.  Keep in mind that these are 45mm throttles feeding 375cc cylinders so 15% is a fair amount for throttle.

http://www.nbs-stl.com/CRX/ITB%20MAPxRPM%20vs%20DC%20at%20Low%20Power%20Levels.png

Relativity straight lines but RPM appears to have an effect on the straight line thing.

[whittlebeast]

So now let"s see what happens when you add the high MAP region all the way up thru WOT

http://www.nbs-stl.com/CRX/ITB%20MAPxRPM%20vs%20DC%20at%20All%20Power%20Levels.png

So much for a straight line.....   What are we looking at?

Let's take a look at what things look like when we look at any one RPM.

Here are four plots.  On the left is showing only 4000 RPM. Then 5000 RPM, 6000 RPM and on the right is 7000 RPM.  The color is showing TPS.  Notice each line has a kink at about 20% throttle where I left the cross-hairs.  To the left and below the the cross, things look normal or a straight line.  Above the 20% TPS, the motor wants way more fuel.  Keep in mind that MAPxRPM on most motors is essentially air flow.

http://www.nbs-stl.com/CRX/CRX%20ITB%20Mode%20Cal.png

So now, let's look at low HP mode on the left and High HP mode on the right.

http://www.nbs-stl.com/CRX/ITB%20MAPxMPM%20Low%20vs%20High%20Power.png

[whittlebeast]

We can start dissecting the data at at this high HP range and figure out what thing drive other things.

http://www.nbs-stl.com/CRX/ITB%20MAPxRPM%20vs%20DC%20at%20High%20Power.png

In all three shots, I have filtered out below 85 KPA and data below 15% throttle.  Check out how the DC is a function of RPM, MAP and TPS.

Here is how the Estimated HP works out.  Low power on the left and high power on the right.  This thing totally changes personality below about 18 to 22% throttle.

http://www.nbs-stl.com/CRX/ITB%20MAPxMPM%20Low%20vs%20High%20Power.png