Car Modding, Car Tuning and the Science of Horsepower - Tuning Preparation and Clean Up
The first thing you're going to want to do before you apply any sort of tune is clean up the mess in the existing tune. Every car comes with the same "one size fits all" tune from the factory and off the cars go to various parts of the globe, to be used or abused by all sorts of people. That's why cars are equipped with all sorts of provisions for "compensating" for things. Knock sensors adjust timing. Your O2 sensors establish fuel trims, which alter the fuel tables. The transmission learns the load patterns that the car goes through and applies torque management when necessary to save the drivetrain. All of these compensatory mechanisms are meant to keep the car running no matter what you do to it... but you don't just want the car to run. You want it to run WELL and not have to rely on so many subsystems. The closer your tune is to perfect, the more reliably it will make power, the first time, every time.
Clean up is a multi-step process. The first step is to update the firmware to the latest version of the operating system available for that car. If you purchased a tuner, chances are you have access to some sort of repository where users can download the latest file for their car. You must be absolutely sure you do your homework to determine that the file is compatible however, or your may "brick" your car and then it won't work or work properly. Every year and model of car has specific OSIDs that are supported. Each OSID represents a new revision and you'll want the latest one. It's a good idea to belong to the forum for your car or for your tuner or both because everybody else will be able to tell you the latest one, as there are no "universal" websites I can find to determine the latest OSID. GM has one here: https://tis2web.service.gm.com/tis2web but again, always double check with your tuner staff and fellow forum members. To prep your file for the next step of cleanup do the following 2 things. Disable PE (Power Enrichment) mode by preventing one of its conditions from being met (like set the minimum throttle position to 100%). This is so that the fuel trims are consistent throughout the power band. Then copy your high octane or "good fuel" spark table over to the low octane or "bad fuel" spark table. This is so that the computer doesn't mess around with the spark while you're driving, which can throw off what you're currently tuning. When we tune we try to keep everything the same except for the 1 thing we're working on. Arguably impossible but that's the goal. Once you have your latest, prepped bin or hpt file or whatever file your tuner understands, do a FULL WRITE on your car, replacing the entire file with the new one. Make sure to turn the key off after the write, for several seconds and then start it up to make sure it worked. Then drive the car a few times so that all the sensors can dial themselves in again and the PCM will re-learn everything.
Preparing your wideband for scanning
By the way, make it a habit to ALWAYS do road scans when the car is at "steady state". This means all the fluids (engine/transmission) have to be at a constant temperature and you should avoid allowing the car to "heat soak" which is to idle it long enough that the intake air temperature starts to rise by 10 degrees or more. The latter is not always possible so just do your best. The car must be allowed to go into closed loop too but this will have long since happened if you wait for the coolants to heat up. In winter it can sometimes take a couple of minutes for the car to enter closed loop but if you wait for your engine coolant to warm up, you're almost certainly in closed loop. Also make sure to use the same gas during the tuning process. Decide which gas is the gas for you and stick with it. If you're too cheap to buy good gas, then your tune will be more of a de-tune and that's what you'll be zeroing in on during your road scans so I'd suggest going with a top tier gas, the highest octane you're willing to pay for consistently and fill up with only that gas, at least for the duration of your tune. Finally, try to pick a stable climate for your tunes. Don't do 1 tune in 90 degree heat and 80% RH and the next tune at night when it's 70 degrees and 30% RH.
Scan results showing fuel trims across the VE table
On to step 1. I consider updating the firmware more like step 0. So step 1 is cleaning up the fuel map. There's a high probability that your car utilizes a MAF sensor to determine the correct fuel enrichment across all operating conditions. If it doesn't, it relies on a VE table, which is a speed density tune. Read my hub on VE tables after you read this hub and employ the lessons learned there in place of the ones I'll refer to here for a MAF tune. A MAF sensor measures temperature, pressure and flow simultaneously and the sensor itself puts out a frequncy in Hz, which the PCM looks up on a table to determine what air mass that frequency corresponds to. The lookup table is already set from the factory and MAF sensors are notorious for fouling up and giving erroneous readings in different climates and when flow is particularly high. If you've had your car for any length of time, the lookup table will need to be re-calibrated because up until now your long-term fuel trims have been skewing more and more each day to compensate for how "off" the MAF table is and we want to fix that. Bring your laptop out to the car and hook it up for a road scan. Make sure to log STFT and LTFT along with the usual suspects (RPM, SPEED, AFR, O2, Fuel Trim Cell, MAF, MAP, IAT, ECT, IPW, SPARK, KNOCK, etc). Take the car out for a nice long drive... don't drive for less than an hour. Why? Because after doing hundreds of tunes myself I've learned that they're garbage unless you can get several data points for each cell of your fuel table. Cars are tempermental things and like to change their minds all the time. You may get an LTFT of 3 in a given cell and then the next time it will be 1. There are a lot of cells to hit and you have to obey traffic laws so you'll have difficulty hitting every operating point several times unless you drive for a while. Pick a route that has both city and highway driving. Find roads where you can really gun it from a stop, so you can hit those higher MAF values. Try to vary the load and the RPM on the car as much as humanly possible. Do all your gentle driving at the beginning and save the harsh driving for the very end too if you can help it. The reason for this is you may induce torque management or knock retard learn or some other safety feature that will reduce the performance of the car and throw off your readings. Better to do this at the end so that most of the data isn't contaminated by it. When you are finished your road scan, dump the data to an excel file and for each cell, take the average STFT, add to the average LTFT and then make it into a percentage (ie. value of 3.1 becomes 3.1%). Then apply that percentage correction to the MAF value for that cell (ie. 100.0 becomes 103.1). Do this for every cell in which you've collected more than 10 data points. You may wish to interpolate those cells for which you do not have anything, which I recommend because doing so will prevent abrupt blips in the MAF curve. Upload the new MAF table and write the new file to your car. Make sure to clear all the learned parameters and to drive your car for a while so the PCM can re-learn. Rinse and repeat this process of scanning, tuning and re-learning about 5 times and you should have all your LTFT values coming in much closer to 0. If all of your LTFT values are between -2 and +2, mission accomplished.
Scan results showing events of KR
Now that your fuel table is nice and clean you will need to move on to your spark table, which is step 2. Once again, go out for a drive and scan everything. Put the car through its paces. You should at this point, not notice any major fuel trim spikes. You will however notice some knock here and there... probably. As before, try to hit as many operating points as you can so you have plenty of redundancy and then make your excel file of the data and have a look at the knock retard column. By the way if your forum/tuner buddies have a tool so you can look at the 2D spark table populated with your road scan results, that is the best option. Working with individual data points in excel for 2D tables is a major pain. Luckily, there won't be THAT many knock events to deal with. Just filter out all the lines that didn't have knock and focus on the ones that did. You'll want to take note of the RPM and the MAP values at which the knock occurred. Spark tables are arranged with the X-axis of the table representing RPM and the Y axis representing either cylinder air mass or MAP. Try to log whichever your tables use or you may need a conversion tool to correlate MAP <--> air mass. In any case it's a speed vs load table and you need to find the cell where the knock occurred by looking at your scan, finding a knock event and then finding those 2 values at which it happened, RPM and load (MAP/air mass). If knock occurred multiple times in the same cell, try to average the knock values. The rule of thumb is, for every degree of knock detected, you want to pull (remove) 1 degree of spark. Go through your high octane (good fuel) spark table and start removing spark from the cells that knocked. Once you're done, copy the entire table over to your low octane (bad fuel) spark table once again. Upload this to the car, re-learn, re-scan, etc. and do this a couple more times until the knock is gone.
Now you've got the latest updates, good fueling and no knock. Well done. This isn't going to gain you any horsepower. What we did here was just clean-up and is a segue into the real thing... that is, a tune meant to increase power but this is a necessary step because unless you do this, you'll be changing parameters based on bad data. The goal of tuning is always to change 1 thing at a time so that the changes you make don't end up fighting each other. You have to eliminate co-dependency in a tune, not tune in spite of it. Many companies sell "canned" tunes by giving you a physical PCM module to install into your car in place of the existing one. Not only is this a waste of money since your hardware is perfectly capable of being re-written without a new module but it's also going to be a very poor quality tune because it's still a "one size fits all" tune that doesn't employ road scanning and iteration. Don't waste your money. Unless a human being visits you to tune your car for you, doing it yourself is the only viable option.