A Dose of Digital
Learning the basics of DIY tuning
Story Hib Halverson - July 11, 2014 02:28 PM
Part of learning HPTuners came from books like this.Photo: Motorbooks, Inc.
Also valuable was the Tuning School’s home study course and the DVD from Calibrated Success.
The key parts of Innovate Motorsports’ LC-1 are the wideband oxygen sensor (top) and the controller.
The MVPI Pro is the hardware part of HPTuners. It connects the car’s diagnostic link connector to your laptop. It has an optional input for devices such as the LC-1.
The LC-1 comes with a Velcro strap. With the Blazer, we strapped the controller to its frame then ran the wires underneath to the door.
We had Big John’s weld oxygen sensor bungs in the Blazer, Camaro and Corvette exhausts. Shown is the Blazer installation.
The goal of our Corvette calibration work was accommodating the installation of this Halltech Killer Bee 2 conversion kit.Photo: Sandt Rubel
In MAF programming, each frequency cell has a corresponding airflow in pounds of air per hour. If those numbers do not match the actual flow through the sensor, the engine will not run right.
An RC Engineering injector. They are the saturated type, have stainless steel internals and use the superior, disc metering system.
The injector change on the Corvette took us all of 30 minutes, not including posing for this photoPhoto: Sandy Rubel
We finished our Corvette calibration session back on Westech’s Superflow.
The Camaro’s high-octane spark table. Upper left, we fooled the ECM into advancing the spark at idle. At lower right, we retarded the spark slightly during wide-open throttle operation around peak torque.
Changing injector “size” is like changing the main jets in a carburetor, something carb-tuning gearheads can relate to.
Our Camaro on Westech’s Superflow AutoDyno.
“Cars with computers,” gearhead luddites whine, “Can’t work on them. Too complicated. Too expensive. Tuning – seriously? Only dealers or engineers can do that.”
Actually, if you understand how engines work and you have good computer skills, you probably can learn to tune – or “calibrate” as engineers say – engine control systems. While I’ve serviced cars with computers, I never learned calibration, so I was up for a new challenge. I ordered HPTuners’ VCM Suite from Summit Racing, which includes VCM Editor, an app for calibrating 1997 or later GM vehicles. I picked HPTuners because it supports a lot of GM vehicles and is also available for some Fords. A Chrysler version is in development.
You can learn tuning several ways. An internship with an experienced tuner is a fabulous idea, but a rare opportunity. Trial-and-error is a bad idea. Internet forums can help, but be cautious when reading calibration discussions. Sometimes the folks who post the most only know enough to be dangerous. Not that experienced calibrators aren’t on line, they’re just hard to find. You can also read books, watch DVDs or take classes. I opted for that.
I read Jeff Hartman’s How to Tune and Modify Automotive Engine Management Systems, from Motorbooks, Greg Banish’s Designing and Tuning High-Performance Fuel Injection Systems and Engine Management: Advanced Tuning along with Harold Bettes’ and Bill Hancock’s Dyno Testing and Tuning, all three from CarTech. I took a home-study course: A Beginner’s Guide to Tuning GM Vehicles with HP Tuners Software, offered by the “Tuning School”. I watched GM Beginner’s Tuning Guide, a DVD from Greg Banish’s Calibrated Success series and sold by Summit Racing along with Zip Products. All this training was incredibly useful.
When to face the dyno
The amount of work to be done on a chassis dyno evokes strong opinions among calibrators. Some insist the only way is with a load-bearing chassis dyno and road testing is for rank amateurs. Others make a case for some calibrating on a chassis dyno with the test track suitable for other tasks. Even our training material differed, with Calibrated Success’ DVD on the “all dyno” side and the Tuning School suggesting some work can be done on the road.
We mixed dyno and track testing because of the high cost of doing it all on a dyno. We believe our practical tuning needs were met, but we also agree that to get a perfect calibration requires a lot of dyno time.
Banish’s DVD taught us to think “lambda,” a measure of free oxygen in the exhaust, identified by the greek letter “λ,” rather than air/fuel ratio (AFR). Combustion is “stoichiometric”, or Stoich, when neither unburnt hydrocarbons (rich) nor excess oxygen (lean) are in the exhaust. If combustion is “stoich,” lambda = 1, and the engine has low emissions, high catalytic converter effectiveness and good fuel economy. Stoichiometric is 1.00 λ regardless of fuel, whereas AFR varies. Stoich for gasoline varies from 14.55 to 14.8:1 depending on humidity and the gasoline blend. Oxygenated gasolines, with up to 10 percent ethanol (E10), are 14.1 to 14.2:1. E85, popular with racers, is 9.87:1, but at 1.0 λ; all burn stoichiometrically.
An engine under high load needs a rich air/fuel ratio to prevent excessive combustion temperature and detonation. Calibrators call this “power enrichment” or “PE” and on straight gasoline, it’s about 12.5:1, AFR. With E10, it’s about 12.2:1 and with E85 it’s 8.4:1. If you’re using lambda, those are all 0.84 to 0.86 λ. Speaking in terms of lambda makes calibration, or “cal” as some say, easier.
Why Oxygen Sensors
In 1976, the Robert Bosch Corporation developed the automotive lambda or oxygen sensor to measure free oxygen in exhaust gases. Depending on the level of exhaust oxygen, they generate 100 to 1,000 millivolts. Most factory oxygen sensors are “narrow-band” with a linear output only in the narrow band of voltages generated when combustion is near stoichiometric. Outside a range of 0.97 to 1.06-lambda, the signal is non-linear, making useful measurement during PE impossible.
Wideband sensors, developed by Bosch in the mid-1990s, vary a reference voltage supplied by a controller. Their output is more linear so, given a low level of back pressure, a wideband is accurate from 0.34 to 1.51 λ, making widebands capable to calibrate PE.
We ordered Bosch sensors (PN 17014) from RockAuto.com. We used an Innovate Motorsports LC-1 controller (PN 3744) recommended by the Tuning School, which we ordered from Summit. It connects to HPTuners’ MVPI Pro interface. We connected the LC-1, MVPI Pro, an LED indicator, a push-button and a wiring harness we fabricated according to Innovate’s and HPTuners’ instructions. We configured VCM Scanner, HPT’s data logger, to record the LC-1.
Playing Our First Tune
I started with two easy tuning projects, an ’07 Chevy HHR and a ’99 Chevy Blazer. Both had K&N air boxes and Flowmaster exhaust. The Blazer also had ported and polished heads, 1.6 rockers and headers. The key goals were improved drivability via optimized fuel trims and eliminating the stock calibration’s power enrichment delay.
The Tuning School taught us to evaluate part throttle operation with VCM Scanner then use VCM Editor to produce a new calibration, which decreased the amount long-term fuel trim (LTFT) changed as the ECM selected different cells in the volumetric efficiency (VE) and mass airflow tables. The result was improved drivability and consistently low emissions.
A wideband oxygen sensor requires a bung in the exhaust ahead of any catalytic convertors. We used MSD bungs (PN 2335) and had Big John’s Performance in Valencia, California, install them. We temporarily connected the LC-1 to each vehicle.
To enhance fuel economy, some stock calibrations delay power enrichment. At high load in the mid-range, initially, fuel delivery is held at stoichiometric. This improves gas mileage, but it makes drivability annoying. The engine feels “flat” until about 4,000 rpm, when there’s a surge as PE cuts in. Using VCM Editor, I lowered the power enrichment threshold and the minimum TPS range required for PE. Finally, I doubled the rate to quicken the onset of PE. Mid-range drivability improved and the wide-open throttle surge was gone. The Blazer also needed a tweak of the air/fuel ratio at WOT.
Finally, to improve the HHR’s cooling in traffic, we changed the electric fan programming.
The instructional material took a week to digest. After that, I practiced with HPTuners’ VCM Editor. In couple of weeks, I was ready to “flash” the HHR’s computer with my first calibration and a few weeks after that, we did the Blazer. I couldn’t have begun to write calibrations in two weeks without the books, the Tuning School courses and Calibrated Success’ DVD. Indeed, we had more to learn, but we knew the basics. With two easy calibration jobs under our belt, we were ready to do a Corvette.
With our 2004 Z06, the goal was to calibrate for a different air intake system (AIS). We installed a Halltech Systems “Killer Bee II” conversion package on its LS6 in place of a Granatelli Motorsports AIS. The engine also had Crane Cams 1.8:1 roller rockers, dual valve springs and titanium retainers along with Denso IT22 plugs, MSD plug wires, a Corsa crosspipe and a custom tune from another source.
Our pals at Katech Engines suggested the KB2, which has the less-restrictive, Hitachi MAF sensor used on newer C6 Vettes and Halltech’s “Killer Bee” C6 Corvette air bridge, a cotton gauze air filter and a wiring adapter to connect the C5 harness to the C6 MAF sensor. The Killer Bee II requires a MAF table compatible with the LS7 sensor.
MAF sensors put out a five-volt square wave, the frequency of which varies with air flow. The calibration’s MAF table has frequency “cells” and each cell has an airflow value. From that value and the calibration’s commanded air/fuel ratio, the ECM computes the proper injector open time. This works well until one installs a different MAF sensor. If programmed values don’t match actual airflow though the sensor, the mixture will be incorrect and the engine will run poorly or not at all.
The Tuning School and Calibrated Success taught me to calibrate for the different sensor. The first challenge was a starting point. I used a stock LS3 table, but that was a problem, also. The scales were not the same. Factory LS6 programming starts at 1,500-Hz and has eight cells per 1,000-Hz while LS3 programming starts at zero and has 20 cells per 1,000-Hz. It took some math to put the LS3 data into the LS6 calibration, but I ended up with a useful starting point. From there, we fine-tuned the MAF table using data from VCM Scanner. Once that was complete, the LS6’s fuel delivery was much better, but we weren’t done, yet.
Our data showed high injector duty cycle. John Park at RC Engineering, a leading aftermarket injector source, told us when duty goes over 85 percent for extended periods, injectors can be difficult to control. RC Engineering recommended a 310 cc/minute injector (PN SL2-0310). RCs use disc-type fuel metering, have stainless steel internals and are renowned for minimal flow variance, ±1.5 percent, compared to the four percent RC measured from our stockers.
After installing the injectors, we used VCM Editor to put new flow and voltage offset numbers into the LS6 cal. Populating the offset table was more than entering new data because it offsets for manifold vacuum from 0 to 80 Kpa (0-11 psi). There was some figuring along with educated guessing necessary to change that table.
The ’Vette’s previous tuner copied the high-octane spark data into the low-octane table, a common mistake. Low octane spark tables are there for a reason: to prevent excessive knock retard during unintended or emergency use of low-octane gas. We replaced low-octane spark numbers with those from a stock calibration. We finished the Corvette by tweaking long-term fuel trim, changing the rpm and rate at which power enrichment enables and reducing the ECM’s fan-on temperatures.
We validated our work on Westech’s chassis dyno. We added enough Rockett Brand 100 unleaded racing gasoline to get 93 octane and made five dyno runs. The Z06, with the Halltech Killer Bee II, averaged 380.1 hp at the wheels, SAE-corrected, an increase of 11.1 hp over the Granatelli AIS, stock injectors and our previous tune and a 14 hp improvement over the stock air filter assembly.
Our final project, the V-6 Camaro, was more of a stretch because it was modified to a greater extent – Comp cam, DeGroff heads with bigger valves, Katech valve springs and titanium retainers, Yella Terra 1.8:1 rockers, Extrude Honed intake manifold, AIS, headers, Flowmaster exhaust, MSD coils, MSD wires and Denso IT-20 plugs.
The Camaro’s first tuner couldn’t fix an unstable idle. VCM Scanner showed spark timing fluctuating between zero and five degrees BTDC. That caused surging and made the engine lazy off-idle. Looking at the main spark table with VCM Editor, I saw more timing than was showing in scans. At idle, I suspected other tables overrode the main table, but access to them was not supported. E-mailing HPTuners confirmed that. My solution was to “fool” the ECM into advancing the spark by adding 13 degrees in cells where the engine is idling. That did not eliminate the surge, but did reduce it significantly and improved off-idle response from lazy to downright snappy.
The cam and head work altered volumetric efficiency and so changed the “main VE” table. For part of that, a chassis dyno is required so we headed for Westech Performance Group and its Superflow dyno. VE work requires temporarily disabling MAF sensing and forcing open loop. The VCM scanner is used to determine the amount of correction required. The home study material and the DVD showed us how to do all that.
Calibrating the low rpm part of the table requires a computer-controlled, load-bearing dyno which holds an engine speed as throttle opening is slowly increased. This minimizes acceleration enrichment or fuel film on the intake port walls near the injectors from affecting test data. This requires lots of cooling airflow and Westech had two giant fans for that.
For the high rpm/high load VE changes, even with fans, steady state tests are hard on the engine’s cooling. Errors from acceleration enrichment and wall film are less, so we used conventional “sweep” tests to revise that part of the table. We applied the changes to a normal calibration and deleted the temporary cal.
Next was wide-open throttle mixture. Wideband data showed power enrichment was out of whack. We saw 0.86 λ for several seconds, then it would go rich until about 5,000 rpm and then go lean, again. Part of this was the “Power Enrich Fuel Adder rpm vs. Time” table, which, once PE is enabled for a while, adds extra fuel for cat overtemp protection. We zeroed that table for WOT periods shorter than 19 seconds and reduced the rest by half.
Above 5,000 rpm, the VCM Scanner indicated high injector duty cycle – understandable considering we were making 270hp with injectors sized for 200. RC Engineering recommended a 240
cc/min. injector (PN SL2-0240), so we ordered them. They have an older style, “mini-timer” connector, whereas stock injectors use the “Multec II” connector. Six Racetronix adapter harnesses (IA-M2M) are required to connect to them to the stock injector wiring. Calibrating for the V-6’s RC injectors required the same techniques as did the Corvette. Testing showed PE was a tad rich so we took a little fuel out, then road tested again to find 0.85-0.86 λ and maximum injector duty-cycle about 85 to 90 percent.
We had knock retard around peak torque. In the main spark table, around peak torque at high manifold pressure, I retarded two degrees, dropped the bordering cells by one degree and tested again. Knock retard was reduced or eliminated. We finished our Camaro with fuel trim adjustments and revised fan-ons.
Back at Westech, as a knock retard hedge, I mixed Rockett Brand 100 with pump gas. In three runs, we saw 0.84 to 0.87 λ. We also made 244.9 hp at 5,750 rpm and 234.5 lbs-ft at 4,700 rpm at the wheels, SAE-corrected. That’s just under 17 rwhp and 14 rear wheel lbs-ft torque more than our last test of that car. Using a .82 driveline loss correction, that’s about 299hp and 280 lbs-ft, SAE, at the flywheel. Not bad for a 3.8-liter V-6 which started at 200hp. Lastly, we took the Camaro to a California Smog Check station and had it tested. The engine was way clean.
The Bottom Line
Considering the plethora of books on the subject, the idea we could teach you everything about tuning in a magazine article is crazy. Our intent was to give you an insight to training options, cover a few basics then discuss our experiences in becoming intermediate calibrators. Hopefully, you’re pointed in the right direction with your enthusiasm fired up and the understanding that learning to calibrate is not the fearsome task some believe it to be. In the meantime, we’re going to keep studying our training material and enhancing our tuning skills.
For Your Information:
Innovate Motorsports, (714) 372-5910, www.innovatemotorsports.com
Motorbooks, (800) 826-6600, www.motorbooks.com
RC Engineering, (310) 320-2277, www.rceng.com
Rock Auto, www.rockauto.com
Rockett Brand Racing Fuel, (847) 795-8400, www.rockettbrand.com
Summit Racing, (800) 230-3030, www.summitracing.com
The Tuning School, (727) 264-8875, www.thetuningschool.com
Westech Performance Group, (951) 685-4767, www.westechperformance.com
Zip Products, (800) 962-9632, www.zip-corvette.com