Advanced Level Tuning Strategies

AutoWiz

Well-Known Member
What if I said that the best way to tune was to learn to talk to our engines and let them tell us what they like for any given calibration? Now how about if we take it one step further and say this is the ONLY way we will find true and accurate calibrations for our builds? So much so that once we reach a point of a base calibration and are up and running and actually begin the tuning process we should toss out any scribbles we have on paper or beliefs of what a table or cell should look like. And just communicate with our powerplant much like a dr. would communicate with his patient to find out what is wrong. But with even more accuracy and absolutely no guesswork.

Datalogging is our path to listening to our vehicle's demands. And I say vehicle because datalogging can be used to tune more than the engine. It can also assist with shift points, shift times, shift pressures, torque converter lockup engagement and percentage control through PWM, and even trans temps. So powerful is having this ability to record data and time it to other logged data that it is the gold standard of tuning in the industry. And it didn't just become that way, either. Rather it has just matured and flourished into what we all hoped it could one day be. The end all to tuning solutions.

There is little hard calibration data that needs to be known. These are constants. Like injector flowrate. Also engine displacement and firing order. And surface area on the throttle body. A timing map would be good to have but like with the MAF calibration really we just need a starting point there as even that the car will tell us what it wants if we just learn how to ask. And I say ask not listen because I know a lot like to tune by ear and hear how the car sounds. And while to the skilled even this is better than math on paper when it comes to an actual functional real world calibration the simple truth is we designed the computers and the sensors to be far more accurate than we could ever be. These are the tools we made and how they function. And it would benefit some or most here to consider how these new processes might be applied to our older stuff in the name of the tuning perfection that we are all used to on todays cars and struggle so hard to attain on yesteryears stuff that we like to play with.
 
A data log is not a calibration and nor is it a direct telltale of what we need to do to any field or cell in any part of our tune. For that we need a special chart called a histogram. Or today we are just calling them charts. First we need to isolate what system or part of our tune we are working with. So let's start with something easy. Like a 2d table. The Mass Airflow Sensor, or MAF, is a pretty basic calibration. This sensor has a small heated wire that sits in the middle of the airstream and the PCM sends pulses of power down this wire to heat it up until a certain resistance is achieved. Then it measures how much it has to pulse to keep this wire at its resistance and in this way the controller logically knows EXACTLY how much air is entering the engine. So this calibration table is gonna be HZ(frequency of power sent) on one axis and airflow on the other. A simple 2d chart. Here is what the table looks like on a late model car..



The table actually extends all the way out to 14k hz or wherever I want to set the fail hz at. So to update this chart with the correct airflow for a given hz we will create a histogram which is a log over a chart. I will create this chart to mimic the maf table. I will make this chart to have hz on one axis with all the same breakpoints. But instead of airflow on the other axis, I will LOG long term fuel trim, or short term + long term fuel trim. OR if I am for whatever reason not using the factory narrowband o2 sensors, I will use a wideband that is plugged directly into my datalogging solution and log afr error. And have this plotted to my histogram. Then we just drive. On the roller or on the street. Making sure to log plenty of data in the areas we are trying to tune. Then we take the whole chart we just made and copy and paste it multiplying the values in the calibration by the values recorded in our logged chart/histogram. Then on the next log we should see all those number come down. And further and further as our car TELLS us what it wants that MAF calibration to look like.
 
yes , the more I learned, about engines the more I realized,
that the use of a vacuum gauge,
infrared temp gun
,reading spark plugs,
use of the timing light,
dwell meter,
fuel pressure gauge
and listening to the engine,
oil pressure, fuel pressure,
measuring exhaust back-pressure,
and plenum vacuum readings,alone,
was not getting me to where I wanted to be in tuning,
yes they are all helpful,
yes you can get very good using old school methods,

yet, the more I learned to depend on various sensors and temp and fuel/air ratio,
and exhaust temp meters,
and use the older tools available,
the closer to where I needed to be and the better the results were.
data logging with all the sensors on a lap top real time display:D
is and will continue to be more precise, allowing you an edge!

with ignition advance,
fuel air ratio,
exhaust temp,
plenum vacuum,
exhaust temp,
alternator voltage,
air and coolant and oil temp.
exhaust back pressure,
header exhaust temps,
and trans shift points,
etc.


and a code scanner are useful upgrades
 
Last edited:
since the days of LT1 back in 1993 GM has used 2x fueling systems over each other for better accuracy and reliability. Aside from the mass air flow, GM cars also have their fueling driven by VE or volumetric Efficiency tables. These are 3d tables that are MAP vs. RPM. vs VE. These are a big table that look a lot like this..



And we will calibrate this entire table and every individual cell in it with great precision by repeating the same above mentioned process. We will create a chart in our datalogger that will mimic the table we want to work with. Here our chart will be rpm and its breakpoints over MAP and its breakpoints and instead of ve we will plot fuel trim data or afr error into the cells to be plotted. And we drive. Fetch some data and we will do the copy and paste thing multiplying all cells of the table by the data filled in our charts. And the numbers in each cell should drop more and more on each subsequent drive. All the way down to -5, -3, or if you are the Wiz, -2%.
 
Last edited:
I use 2 different methods for dialing in ignition timing based on the vehicle I am working with. The late model controllers are good enough to provide a delivered engine torque PID. And as we all know horsepower is nothing more than a math of torque. So we can make our own PID in the datalogger based off of the delivered engine torque and some math. Here is a brief video on how to set that up in HPTuners detailed by the man himself, my tuning instructor from The Tuning School so many years ago, now..


His name is also on all the tuning books I keep investing in so heavily to stay ahead of the curve. Back to the task at hand, So now that we can log delivered hp and tq well that open the playing field considerably for other tables we can now modify to their exact correct position. No guessing. No feeling. No accepting second best. So now we will go to our main spark tables. Again we will make a chart to mimic the axes of the map we wish to alter. But instead of plotting fuel trims over our ignition chart we will plot hp over it. For this to work we need to mimic our drive each time as closely as possible. same throttle input across the same ranges. If on the road and not on a dyno we should have a designated 'track' layout to keep following and try to lap it the same way and speed each time. And we will just save the chart. Then we will pick an area idle or part throttle and make changes to the timing table. Just 2 degrees. And go drive again. Save the log and compare. We will see the hp stop climbing even before the detonation starts. This isn't just an accurate way to tune ignition timing, it also happens to be the only known safe way of doing so without bouncing off of detonation. Because by the time your ear-crometer hears the pinging or change rattling you are actually many degrees into knock.

For an older car without ECU control, a dyno is needed to emulate the hp readings. And that same dyno can also fill in with a wideband for the o2 readings. These processes are rock solid and have solidified my footing as a tuner in this industry. And even still I am awe stricken at the painful accuracy I am able to achieve on each car I tune with these processes. It is funny to me how some guys tuner lock their ecus so no one else can see what they setup. Because a tune, ANY TUNE, is not good because of the values and calibrations in it. No it is good because of the processes used to bring it about for that one very specific vehicle.
 
The professional strategy or approach to tuning is not to determine or find a value or number to enter or decide who to believe about what you should set where. The professional way is to flex that brain muscle in such a way that we can all deduce a logical way to understand how to let our machine tell us what it needs. Because no one else knows. No matter how good. Even the best mechanic in the world cant install head gaskets without the right tools for the job.

Even if you list out all them specs to the best professional on the planet. He has no idea HOW those parts were installed. He does not know if your alternator is holding your charging system at 13.5 or 14.5v and voltage offset really matters. And sadly even he is doing his math based of the specs provided by the manufacturer of each part which is a claim made to sell a product and not what 90% of the world will see with that part. Because if manufacturers claimed real world average numbers they wouldn't sell so much. Today the question of "has your car been on the dyno?" is really the smart way of asking someone "has your car been tuned?"
 
Last edited:
yes , the more I learned, about engines the more I realized,
that the use of a vacuum gauge,
infrared temp gun
,reading spark plugs,
use of the timing light,
dwell meter,
fuel pressure gauge
and listening to the engine,
oil pressure, fuel pressure,
measuring exhaust back-pressure,
and plenum vacuum readings,alone,
was not getting me to where I wanted to be in tuning,

Because we work on all generations of Corvette. 1953-current. We have this equipment for dialing in the late models and I long to be able to use it on the old. The datalogging solutions have external inputs. different amounts for different solutions. Ours has 5 external inputs. So for anything I want to do to accurately tune the old stuff with logging I just have to do it with 5 sensors or less and create a 0-5v analog signals from the above tools. This is a very easy thing to do today. A vacuum Tee and an earlier map sensor like for a C4. And now you can plot engine vacuum to engine rpm, or engine vacuum to an HP pull on the dyno. Most dataloggers have their own 5v supply for these sensors. If not then any cell phone charger is outputting 5vdc. Our only limitation today is our own knowledge and imagination. We need to log. we need to be able to see what something is doing when compared to what other stuff is doing at the same exact time. In a manner that can be recorded and saved with accuracy. And we don't have to have EFI to do it.
 
Last edited:
Even a datalogger is an option once we are fluent at what we are doing and understand that really we are recording many sensors together and they are all 0-5v things. I have a 4 channel lab scope for all the drivability stuff I have to do. My Snap-On Verus Edge:
https://www.digitalcorvettes.com/forums/showthread.php?t=257522

With nothing more than that I can log 4 inputs together and am only forced to do math to translate raw data into tuning data. But I never have to make up anything or take a guess at any correct value. Tuning today really is about understanding how to maintain control in our processes and methods. It truly is a discipline. The trick is logging things together in a manner that is precisely timed. That is to log that vacuum reading while you are just looking at the engine rpm is damned useless for finding the optimum timing off of vacuum levels. For that we want to log vacuum reading with engine rpm also being logged by the same solution at the same time. So using 4 multimeters sitting next to each other is not going to work unless you have some exotic high speed camera to record all of them. But even then their cheap displays will likely be a problem with accurately timing down to single digit ms. The 4 channel lab scope records 4 together in us when I need. So we need to be logical in our methods and how we make our solution if that is the path we so choose.
 
Last edited:
Back
Top