Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Steve Cole]

An O2 sensor of any type is simply a measuring device that puts out a signal based on what it measures, no more no less. Every sensor has a tolerance of how accurately it measures things and most of the time that is a non-linear curve. So while it maybe more or less accurate at different points along the output curve it will still give a reading. That said, without knowing the accuracy curve and tolerance of a sensor you cannot tell what is what, for sure. This is true with ALL sensors and is just why when any sensor gets used, you must follow and know the manufactures specifications.

This is why I find it so funny that people seem to believe that the use of a Broad Band O2 sensor used out of the manufacture specifications is the best thing in the world and they blindly follow what it shows. Then when it's proven that the sensor is being used incorrectly all hell breaks out and they cry foul!


Andy

So what you are saying is your looking at scatter plots to identify something that looks out of the normal pattern to you. There is no measurement and no accuracy at all with it, so there is no scale to work with at all as there is none. Once you SEE something in the scatter plot that looks out of the normal you still have to go back to the data to find what may or maynot be the cause of it. How is this anything different than what has been done for the past 30+ years?

"To answer the original post, no you can not use the high o2 voltage to tune the WOT portion of the tune but you definitely can use what you learned at low and medium power to better predict what the motor will need for fueling at the top end.  Low voltage at near WOT is almost always a dead giveaway that you are way lean."

Just what Vtune has done since we introduced it back in 2008, glad you have come to the same conclusion we told you about 5 years ago. Now to add to it, we can also add in the O2 sensor output to give yet another test point to more accurately predict what WOT fueling needs to be. If you can also know (which we do ) the accuracy curve of the sensor it adds yet another test point. The more pieces of the puzzle you have the closer and closer you can get.

[whittlebeast]

It is true that you do not have a "Easy Button, fix this now" type thing.  It is more like an auditor having an easy way to locate the account that is the company slush fund.  The one that the boss uses for his personal fun money.  You home in real quick on the problem area where autotune may be falling apart and getting the wrong answer.  In the Harley world, the most common issue is super short headers or strange intakes that go into reversion under certain combinations.

The link I posted a few pages back gets into this stuff on the ITB stuff.  As it turns out the turbo stuff tends to be the easiest type motor to tune.  At least on the fueling side of the equation.  They are very predictable.

HFT

Andy

[Steve Cole]

Its frustrating when somebody doesn't just give you the answers but asks you to do test and to figure it out yourself.

Joe

Your data graphed from the test you did provide the answer in itself. It showed that the BroadBand O2 sensor measurement you took was no more accurate than the Narrow Band O2 showed under the same test condition. So if you want to say that using a Broad Band O2 sensor is good enough (as they are used today) for WOT mixture then the Narrow Band O2 sensor must also be good enough. If your going to say one is no good then you MUST also say the other is no good!

The whole point was to show that you can do just as good with one as the other during this test condition. Now, if you had a Broad Band O2 sensor that was used how the manufacture specifications say to use it, the answer may be different, but that is NOT how they are currently being used.

[whittlebeast]

This is a little off topic but shows how the location of the MAP read in the 720 degrees of crank rotation can mess with the basic fueling equation.  This is the pressure in the intake compared to the crank trigger teeth.  This was done on a Sportster motor running 2500 RPM.  The plot is showing about 900 degrees of crank rotation.  The long step in the tooth log is the missing tooth area as it sails past the crank sensor.  The map is swings from 48 - 93 KPA in the intake twice every 720 degrees of crank rotation.



http://www.ncs-stl.com/Ms3BetaTesting/Harley_MAP_Per_Tooth.png

This is what we are using to get engine phasing correct on the wasted spark to non wasted spark changeover.  Way cool stuff. 

Andy

[Steve Cole]

Andy

As you know the devil is in the details. How are you collecting the crank tooth count data? The sensor does not give a rising voltage signal as you show it, it is just pulses one after the other until you get to the missing teeth The XL crankshaft has 32 pulses per revolution. So you either have the wrong data or some is missing as you graph only shows 29 pulses plus the missing teeth.

[Coyote.]

His graph has a vertical step up for each tooth (30 of them). It's just how he has chosen to display the pulses. There are 30 teeth on a sportster flywheel followed by a gap of 2 teeth. Looks correct to me.

[whittlebeast]

The crank is a 32-2 so there are 60 teeth in 720 degrees of crank rotation.  You are looking at the tooth counter, internal in the MS code.

In this test mode, the code does a MAP read at every tooth.  The code does this randomly in this mode to allow enough time for SD card burns as I recall.  I don't code so.....

Andy

[FLTRI]

This is what we are using to get engine phasing correct on the wasted spark to non wasted spark changeover.  Way cool stuff. 

Andy

Andy,
Why would wasted spark ignition timing be different than non-wasted spark timing? Most aftermarket hot rod ignition modules offer single (non-wasted) or dual (wasted) fire systems.
Bob

[whittlebeast]

Things will get far more interesting when that motor comes up on the boost.  Right now we are still in the get it started and keep it synced stage of coding.  Tuning is when the coders get out of the way and I get involved.

I love to tune.  Any more, I only get involved in the strange ones.  They are the most fun.

Does anyone know at what boost, the heads start to lift?

Keeping the cylinder fueling balanced when you can't get the widebands to survive before the turbo is going to be a challenge.  There is talk about running two turbos to beat that one.  Wish us a little luck

Andy

[whittlebeast]

Andy,
Why would wasted spark ignition timing be different than non-wasted spark timing? Most aftermarket hot rod ignition modules offer single (non-wasted) or dual (wasted) fire systems.
Bob

This has everything to do with getting it to fire in as few revolutions of the crank as possible.  The coders sweat those things.

Andy

[hrdtail78]

This is a little off topic but shows how the location of the MAP read in the 720 degrees of crank rotation can mess with the basic fueling equation.  This is the pressure in the intake compared to the crank trigger teeth.  This was done on a Sportster motor running 2500 RPM.  The plot is showing about 900 degrees of crank rotation.  The long step in the tooth log is the missing tooth area as it sails past the crank sensor.  The map is swings from 48 - 93 KPA in the intake twice every 720 degrees of crank rotation.

 Andy

So, what you show is normal, the swing in KPA?  Because you are showing how it's polled in every position and not at a particular position?

[FLTRI]

This has everything to do with getting it to fire in as few revolutions of the crank as possible.  The coders sweat those things.

Andy

Got it. Starting from scratch.
What engine and mods are you asking about heads lifting?

What prep has been done to account for big boost?

Bob

[whittlebeast]

So, what you show is normal, the swing in KPA?  Because you are showing how it's polled in every position and not at a particular position?

It is just a guess on my part but my guess is that the programmers will elect to do every tooth MAP read till the motor gets to idle speed and then switch over to doing a map read for say three specific teeth and then resample about 180 degrees later depending on the motor phase.  Whether or not he uses the average of the two reads for the the fueling calc will be up to him.  It depends what things look like when the motor goes into boost for what we elect to do.  That is the fun part of coding.  The processor is plenty fast enough for anything a Harley motor can throw at it.

Andy

[whittlebeast]

Got it. Starting from scratch.
What engine and mods are you asking about heads lifting?

What prep has been done to account for big boost?

Bob

I normally do not get involved with the engine build. The thing that concerns me is the heads are only held on with 4 long bolts.  That sounds like trouble.

Keeping the heads cool at 130 MPH dragging a sidecar for 5 miles also sounds ugly.  I just look at data and do the typing....  others wrench on it.

Andy

[Steve Cole]

His graph has a vertical step up for each tooth (30 of them). It's just how he has chosen to display the pulses. There are 30 teeth on a sportster flywheel followed by a gap of 2 teeth. Looks correct to me.

Count the pulses on the graph. If you start with the wide gap as the end of the first cycle there are only 29 steps then the next wide pulse. The flywheel has 32 teeth with 2 missing so there should be 30 steps plus the wide one.