What the tune needs for cams, headers, intakes, superchargers and E85

The computer needs four things to deliver a good result. It needs a model of how much air enters the cylinder. It needs a fuel delivery model that matches the injectors, pressure and fuel in the tank. It needs a spark command that stays inside the knock and torque limits.

It needs a torque report that the transmission and the throttle can trust.

• A cam changes the first item at every rpm and at every cam angle. Headers change exhaust transport and can shift the oxygen reading. A throttle body or manifold changes the flow area and the runner behavior. A supercharger adds a pressure range the stock model may not cover. E85 changes the fuel's stoichiometry and the volume needed for the same lambda.
• The common mistake is to treat the part as a table. A cam kit arrives with a card, a limiter or a lock, and a spring requirement. None of those numbers are a calibration. A blower kit arrives with a pulley, a MAP sensor and injectors. The kit's boost number is not a spark value. An E85 conversation often starts with a stoich change and an injector rescale, and those two changes can double the same correction.

This page covers the 2005–2014 5.7, 6.1 and 6.4 Hemi in cars and trucks on HP Tuners. The 2005–2006 applications generally use NGC3. The 2007–2010 files are NGC4. The 2011–2014 window is mixed. A 2011–2012 Ram 5.7 is NGC4A.

Later cars and trucks include GPEC-family controllers, including GPEC2 and the 2014-truck GPEC2A variant. Identify the controller and OS from the read rather than assigning one controller to the whole year. This page does not cover the 2015 and later generation, which needs a different conversation and a different service boundary.

• The page does not choose a cam, a header, a throttle body, a blower or an injector. It does not cover a 2015+ write, a Hellcat, a TRX, a Demon, an 8HP procedure or nitrous. It does not print a stock cam map, a universal idle target, a boost ceiling, a spark recipe or an E85 timing adder. It does not tell a street car to delete a monitor. Those are different jobs with different evidence.
• A cam, a manifold or a power adder changes the air model, the fuel delivery or the spark limit. Those changes are inside the $199 Full Custom file when the engine is a supported 2005–2014 5.7, 6.1 or 6.4. The file has to model the hardware, the measured fuel and the torque report, so the checks below come before any value. Nitrous, a 2015+ write and an 8HP procedure stay outside the file.

Before a table is opened, write the combination down. The same engine can expose different labels, different axes and different limits in another operating system. A copied screen path is not a procedure. A copied value is worse, because it looks precise.

1. Save the stock read and the as-found file. Work from a copy. Record the controller and OS before a licence dialog or a write.
2. Write the mechanical combination down. Cam card, springs, phaser, MDS, injectors, intake, headers, blower, pulley and fuel.
3. Check the mechanical baseline. Timing marks, clearances, leaks, oiling, compression, contribution and any unresolved fault come before a calibration.
4. Confirm the writer and the definition. A 2007 procedure is not automatically a 2005 procedure. An aftermarket module may not open.
5. Take one warm baseline log. Label the fuel, the gear, the cam state and the ambient condition. A log without the file is only half the evidence.

A longer duration and greater overlap change when the valves open and how much exhaust stays in the cylinder. At idle, greater overlap can reduce manifold vacuum and raise MAP.

That example is COMP Stage 2 HRT part 201-303-17. Its card is useful because it shows the level of detail a tuner needs.

On the 2009–2014 Eagle 5.7 and the 2011–2014 production 6.4, the phaser moves the single in-block cam. Both intake and exhaust events move together.

A cammed engine often idles in park and dies in gear. The difference is load. The converter, the A/C compressor and the fan all add load at a lower engine speed.

• Log desired rpm and actual rpm, idle airflow or throttle command, spark advance, adaptive spark correction, MAP, lambda or trims, cam angle and gear. A large continuous spark correction in one direction can mean spark is compensating for an airflow error. Check the active spark limits as well as the airflow command. A lean lambda reading at idle with a loping cam can be fresh oxygen in the exhaust. Overlap can do that without a lean cylinder mixture. Adding fuel to fix that reading can make the engine richer and worse.
• The editor labels vary by operating system. Minimum idle air, startup air and airflow decay describe functions to investigate; they are not verified table names in every Chrysler definition. A GM throttle-follower path does not establish a matching Hemi control. Search the function, read the help and record the parameter ID. Do not copy a click path from a different OS.

1. Warm the engine and log park, drive, A/C and fan load separately. Do not use one neutral idle to sign off an in-gear stop.
2. Confirm the cam is tracking. If desired and actual angle do not follow, fix the phaser command before the idle fuel.
3. Check the airflow command first. If spark is pinned to hold rpm, the airflow model is probably asking for the wrong air.
4. Then check fuel at the same temperature. A lean exhaust reading with overlap needs a reversion and leak check before more fuel.
5. Test a return to idle from a stop. Log converter release, downshift and airflow decay together. A late converter release can look like an idle dip.
6. Stop after one small change. If the same symptom survives a small airflow change, do not keep adding air. Re-read the log and the mechanical baseline.

Hot idle that did not improve with MDS disabled

In one reported 2011 Ram cam case, MDS, VVT and the network were already disabled, yet hot idle still deteriorated and trims kept removing fuel.

A cammed engine that starts warm can still fail a cold start. Cold fuel vaporizes poorly, the battery and starter speed are lower, and overlap dilutes the charge at cranking speed. The first fire, the flare after first fire, the decay to idle and the warmup are four different phases.

A single "it starts then dies" symptom does not select the table.

1. Start from a genuine cold soak. Confirm ECT and IAT are plausible and the battery can crank at a normal speed.
2. Separate the phases. Cranking, first fire, flare, decay and warmup each get their own log window.
3. Decide whether the engine lacks air, fuel vapor or ignition. A richer command that does not improve stability is not evidence that fuel was the problem.
4. Change one startup control at a time. Keep the parameter name and units tied to this OS. No public cranking or afterstart value is universal for this platform.
5. Repeat the same cold condition. One cold start is one data point. A second morning is the first repeat.
6. Check hot restart separately. Fuel boil, injector heat soak and sensor heat can produce a different fault from a cold-start lean condition.

Do not fix a cold-start airflow deficit by dumping fuel. A rich start can foul plugs and wash the bores, and it hides the phase that actually failed. If throttle keeps it running cold, investigate airflow, fueling, misfire and learned trims together. Pedal input alone does not distinguish them.

Treat wideband readings as useful only after the sensor is ready; a cold sensor cannot measure the first firing cycles.

Long-tube headers move the primary oxygen sensors downstream and change the exhaust pulse timing. The sensor sees a different sample, and the closed-loop correction can arrive later than the stock calibration expects. Scavenging also changes, so the air model can shift at low and high rpm. A header is not one table.

It is a transport change, a bank-balance check and an airflow-model check.

1. Confirm the hardware and the wiring. Header, catalyst, sensor position and bank assignment come before a calibration edit.
2. Fix leaks and misfires. A fresh-air leak or a dead cylinder can look like a fueling error on one bank.
3. Compare both banks at warm closed loop. If only one bank moved, do not change the global air model first.
4. Check transport behavior after a transient. If the OS exposes a delay, read the units and make a measured change. Do not copy a delay map from another platform.
5. Correct the active air model where the data shows error. Inspect load, spark and torque effects as well as fuel trims.
6. Diagnose a catalyst code as a hardware and monitor question. Do not delete the monitor by habit.
7. Repeat cold start, warm cruise and a controlled transient. A header job is not finished at one warm idle.

Keep the stock diagnostic comparison in the header job. One cam-and-heads owner found misfire enable conditions moved outside normal operation and readiness that would not finish. A quiet dash did not establish a healthy engine.

Compare enable temperatures, rpm windows and DTC reporting with a matching stock file; repair the cause rather than making the test impossible to run. A separate long-tube owner reported that a transport-delay increase only slowed the lean codes. That was not a confirmed repair.

A cold-air intake on a speed-density Hemi does not need a mass-air rescale because there is no mass-air sensor. That does not make every intake calibration-neutral.

• A larger throttle body changes the flow area at the same blade angle. At an 80 mm stock bore, a 90 mm body has an area ratio of (90/80)², or 1.266. That is geometry, not a measured 26.6% airflow gain and not a Chrysler scalar. The stock size varies by application, so the 80 mm figure is only an illustration. One 90 mm unit is not recommended by its maker for a stock 5.7. Tip-in can become abrupt before any wide-open gain appears.
• The file may expose a throttle or airflow model with an effective area, a manifold volume or a desired-to-actual response. Find the supported parameter for this OS. Do not invent a GM-style area scalar in a Chrysler file. Test low-pedal drivability before a high-load claim.

The 6.4 manifold is a popular swap onto a 5.7 because it bolts to the square-port heads and has an active runner.

• One seller reports 15 whp from its own chassis dyno on its swap kit. That is one test, not a range for another build. Check PCV, MAP mounting, rail and injector fitment, and manifold volume. That seller's 2009+ VVT swap kit reuses the 5.7 injectors, rails and throttle body; a manifold swap does not automatically require larger injectors. "It bolts on" is not fitment validation. Log the runner state, lambda, load and spark through the crossover. A lean step after the runner switches can be a model error, a delivery issue or a transient. It is not proof of a particular gain.
• The 4,800 rpm crossover that appears in some guides is an unverified generic target. Log the actual transition and choose the switch region and hysteresis from the car. Do not type 4,800 because a guide repeats it.

A positive-displacement blower and a centrifugal compressor can make the same peak boost with different low-rpm load. The tune has to cover the pressure and airflow the engine actually sees, not one peak number on a gauge. Before any spark or fuel value, define the engine, blower, pulley and intercooling.

Record the MAP and IAT locations, the fuel system and the power goal.

A MAP sensor reports absolute pressure. A "2-bar" sensor measures a range up to about two bar absolute, not two bar of boost. At sea-level-like pressure, about half of that range is already used before positive pressure. Verify the sensor's rated range and voltage endpoints.

Derive the slope and intercept from the documented two pressure and voltage points, then translate them into the exact editor definition. The software may encode a range or span rather than a kPa-per-volt slope, so read the units and the help.

1. Define the combination. Engine condition, compression, cam, phaser, blower type, pulley, intercooling, MAP and IAT location, fuel and goal.
2. Validate the mechanical and fuel hardware. Belt drive, bypass, vacuum, intercooler pump, plugs, injectors, pump, filter, wiring and pickup.
3. Size the supply at the real operating pressure. Use pump flow at the actual voltage and pressure, not a free-flow name or a horsepower ceiling.
4. Calibrate the MAP sensor and injectors first. Check key-on MAP against barometric pressure and a known point. Confirm injector characterization and lambda instrumentation.
5. Represent the whole load range. Check every airflow, fuel, spark and torque axis the engine will visit. A sensor that reads boost without model coverage is not enough.
6. Establish the enrichment command before load. Confirm the command enters before the intended pressure region, and that measured lambda follows it.
7. Start at low load and add load in stages. Stop for unexplained pressure loss, injector saturation, misfire, a lean measurement or persistent knock response.
8. Diagnose command before delivery. If the command is wrong, inspect state, active table, axis and limits. If the command is right but lambda is wrong, inspect measurement, pressure, injector data and supply.
9. Repeat at heat-soaked conditions. A cool first run does not establish road heat margin. Check a hot intake, a repeated run and a hot restart.
10. Prove any failsafe separately. Document the input, threshold, delay, action and recovery, and test the response without destructive load.

An injector flows because there is a pressure difference across it. On a fixed-pressure returnless system, rail pressure is regulated relative to atmosphere. When manifold pressure rises, the pressure on the injector's discharge side rises with it. The differential falls even if the rail gauge looks steady.

Injector duty has its own arithmetic. For a verified single pulse each 720 degrees, duty fraction is pulse width in milliseconds multiplied by rpm and divided by 120,000. Multiply that fraction by 100 for percent duty.

At 6,000 rpm the cycle is 20 ms, so a 16 ms pulse is 80% and an 18 ms pulse is 90%.

U.S. retail E85 is specified at 51–83% ethanol by volume, and the blend changes with season and region. The pump label is not a measurement. A file built for 85% ethanol can run rich on a tank that measures 65%.

A file built for 65% can run lean on a tank that measures 82%. Test the fuel or read a content sensor before the calibration is chosen.

A sensor displayed in the scanner does not prove that automatic blend compensation exists. Measure where the ethanol is read, which controller changes fuel, spark and startup, and what happens when the sensor is disconnected or out of range.

A standalone's flex feature or a Hellcat product does not establish the same capability on an early Hemi PCM.

For air mass A and a chosen lambda, fuel mass is A divided by the product of stoichiometric ratio and lambda. Changing stoichiometric ratio from 14.7 to a hypothetical 9.8 at the same lambda increases the required fuel mass. The factor is 14.7/9.8, or 1.50.

That is 50% more mass, not 34%. Volume also depends on density.

A reported cammed 6.4 conversion initially made less torque on pump E85 than on its gasoline baseline. The tank measured near 65% while the file assumed near 85%, leaving it rich. The owner reported improvement after the tank measured near 82% and the calibration matched it.

That supports measuring content before changing spark; it does not establish a gain for another engine.

• Suppose the mass-based calibration already changes stoichiometric ratio and the fuel demand rises 1.50 times. Suppose the injector flow is then reduced to 1/1.4 of its real rate. The aim is to "add 40% for E85." Commanded delivery can move toward 2.10 times before feedback or a limit. The stoichiometric correction, the injector characterization and any external pulse-width controller are one control chain. Do not stack three independent compensations for the same fuel change.
• If the OS uses an additive fuel-to-air ratio, the adder for a target lambda L is stoichiometric F/A multiplied by (1/L minus 1). At a hypothetical 9.8 stoichiometric ratio and lambda 0.80, the F/A stoichiometric value is 0.102041 and the adder is 0.025510. Copying a gasoline absolute F/A adder after changing stoichiometric ratio does not necessarily preserve the same lambda enrichment.

When more power enrichment does not change lambda

• First compare the active request with the final fuel command. A later Whipple/E85 case had different bank stoichiometric settings and a reported maximum F/A below the requested enrichment. It illustrates a limit to investigate, not a menu path or a finished repair for an early PCM. Check both bank settings, enrichment enable, the active maximum and measured delivery before adding more to the request.
• For a worked example, assume stoichiometric F/A is 0.100 and target lambda is 0.80. Required total F/A is 0.100/0.80 = 0.125, so an additive strategy needs 0.025 above stoich. If a verified total-F/A ceiling clips the command at 0.100, increasing the adder cannot reach the target. The calculated command remains lambda 1.00. Confirm the limit's meaning in this OS before changing it. This arithmetic does not establish a safe ceiling.
• With a 14.7 gasoline display constant, displayed AFR is lambda multiplied by 14.7. A display of 11.5 on E85 means lambda 0.7823; it does not report the actual ethanol blend's mass AFR. Tune in lambda or equivalence ratio. The display number is a convention, not the fuel's stoichiometry.

Dedicated-blend workflow

1. Choose dedicated blend or a documented flex architecture first. Record the measured fuel content, injector and pump compatibility, and pressure capacity at that blend.
2. Establish real injector data and a stable baseline. Keep true characterization. Do not use an ethanol flow fudge as the fuel model.
3. Set the supported stoichiometric or conversion control to the measured blend. Confirm whether the PCM uses mass or another convention before recalculating an enrichment adder.
4. Verify low-load closed loop and measured lambda. Record the learned-trim state. Trims that mask a large error are not evidence of a correct model.
5. Add load with pressure, pulse width, lambda, knock and temperature logged. Keep supply headroom for the maximum ethanol content and power you intend to use.
6. Calibrate cold cranking, first fire and warmup separately. Ethanol vaporization and the seasonal blend change the start. A warm run does not validate a cold start.
7. Find spark through torque and knock response. There is no verified instruction to add a fixed number of degrees everywhere on E85.
8. Define the permitted content range and the out-of-range action. Label the file and the fuel requirement. Never assume a gasoline fill is safe on an aggressive dedicated-ethanol spark map.
9. Check the whole fuel path. Pumps, filters, lines and seals have to be compatible. Ethanol-compatible parts alone do not prove the system is suitable.

Flex-architecture workflow

1. Identify where ethanol is measured and which controller changes fuel, spark and startup. A scanner channel is not a control path.
2. For a native method, require the exact OS, pin, enabled tables, content PID, blending and sensor-fault behavior. Without those, the native claim stays unverified.
3. For an external controller, follow its exact injector, vehicle and firmware requirements. Document whether the logged PCM pulse width is before or after the external extension.
4. Avoid double compensation. Establish the base calibration and every controller assumption before a fuel edit.
5. Validate gasoline, an intermediate blend and the maximum intended blend at low load first. Then add load with content, lambda and pressure logged.
6. Test sensor disconnect, out-of-range and power loss safely. An external unit that modifies fuel does not automatically provide a matching spark blend or a safe gasoline fallback.
7. Keep a written recovery and refuel plan. Treat a vendor "any blend" claim as a component claim that still needs full-system validation.

Spark demand depends on compression, cam timing, charge temperature, boost, fuel octane and ethanol content. A cam that traps more air at low rpm can increase cylinder pressure there. A blower raises the charge temperature and the pressure at every point it makes boost. E85 can tolerate more spark in some conditions.

It does not force the optimum to rise, and it does not make mechanical torque capacity unlimited.

• Knock sensors are a feedback channel, not a calibration target. Zero knock retard on one run does not prove margin at a hotter intake, a different tank or a longer load. A log alone cannot find MBT without a valid torque measurement. The spark page covers the knock channels and the torque check. This page only says that no fixed adder from a cam, a blower or an E85 label is a procedure.
• Do not reduce knock sensitivity because the engine is noisy with a cam or headers. That lowers diagnostic reliability and does not prove the counts are false. Do not add a small timing step because it looks safe. A knock-limited engine can be hurt by a small increase.

The Hemi is torque-based. The pedal asks for a torque, the throttle and the airflow model produce it, and the computer reports an estimate to the transmission. A cam or a blower changes the real torque at a given air mass.

If the report is wrong, the shift pressure, the torque reduction and the converter behavior can be wrong even when the engine file looks clean.

• Do not zero every torque limit to make a combination feel stronger. Do not max every request table. Find the active request and the active limit in the region where the complaint lives. A torque-based computer that closes the throttle is protecting something, and the reason has to be found before the limit is moved.
• On a NAG1 car, the shift schedule is in the TCM. The engine file does not contain it. On an RFE truck, the transmission tables live in the PCM, and the 545 and 65RFE families are numbered differently. Do not copy a 545 row onto a 65 file by row number. Match the physical ratio and the displayed number in that file. A 6,000 rpm shift is not a universal safe limit. Read the engine limiter and the completion rpm from the file and the log.
• The NAG1 guide and the 545RFE and 65RFE guide cover the shift schedules. This page only says that the engine combination has to report a torque the transmission can use.

Validation is not one full-throttle run. It is a set of events compared before and after, at the same fuel, the same tire and the same ambient condition where possible. Log the file identity, the fuel content, the gear, the cam state and the converter state with every event. The logging guide covers the channel setup behind these checks.

1. Use the same route and the same fuel. A different road, tire pressure or load is a different test.
2. Log the before and the after with the same channels. A channel that was missing before cannot prove a change.
3. Compare the command and the measurement separately. A command that moved and a measurement that did not is a delivery or sensor problem.
4. Check the transition, not only the steady state. Tip-in, runner crossover, boost onset, shift and converter apply all matter.
5. Repeat after a heat soak. A cool first run is not road heat margin.
6. Send the whole package for review. Stock and current reads, logs, channel config, IDs, parts, fuel test and the exact changes. No VIN or customer name on a public example.

Full Custom is $199 during the launch special, $249 regular, for one engine file. That covers a supported 2005–2014 5.7, 6.1 or 6.4. It covers stock, bolt-ons, a cam and phaser, headers, an intake or manifold, boost and E85.

Each of those is part of the same calibration, so the scope is the work the file needs, not a separate product.

The free Tune Report checks a file and a log you already have. The Full Custom service covers the combinations above on a supported 2005–2014 5.7, 6.1 or 6.4. A cam, boost or E85 file still needs the parts, the fuel test and the log that show what the engine is doing.