Tuning

Tuning Experience

David has locally or remotely road-tuned the following engine setup:
• stock Nissan Skyline RB25DET with airflow meter.
• stock Toyota 1JZ-GTE (14psi).
• Toyota Supra 2JZ-GTE vvti automatic gearbox with twin-scroll quick-spool single turbo, ID1000 injectors.
• Subaru EJ20* WRX STI (MY99), 550cc injectors, 7-to-15 psi.
• Mitsubishi EVO 9, ID2000 injectors, 1.7 bar boost.
• Mitsubishi EVO 9, HKS 1000cc injectors, 1.85 bar boost.
• Nissan Stagea NM35 (VQ25DET), 19 psi, 725cc injectors.

David has also prepared initial base tunes (idle and vacuum/cruise region) for the following:
• Toyota Supra 2JZ-GTE non-vvti with 269 cams, 1000cc injectors.

Tuning style

There are different styles of tuning for the HKS ECUs.

The main ones are:
• MAP-RPM - formally called speed-density.
• AFM-RPM (volumetric efficiency).

Other combination tuning styles involve:
• MAP-RPM with TPS-RPM correction - formally called speed-density / alpha-N hybrid or 4d-table.
• AFM-RPM with MAP-RPM correction.
• TPS-RPM with MAP-RPM correction.

Problems: Airflow Meter (AFM) removal and variable cam control (vvti / MIVEC / AVCS)

HKS promote the idea of removing the airflow meter (AFM) as it is a "restriction" (read: on air flow measurement, not diameter!). Unfortunately, the idea of removing the AFM and replacing it with a pseudo signal does NOT really work in practice in the case of variable valve timing engines. The reason for this is that the OEM ECU still controls the intake cam, however it does it by AFM, even if a MAP sensor exists on the engine (eg: Mitsubishi EVO9). Unfortunately, most authorised HKS tuners are NOT aware of this limitation, let alone the conversion table (Option Output => Voltage Output 1 or 3 on Toyota/Nissan, Option Output => Frequency Output 1 on Mitsubishi), which results in an engine that responds poorly on the low end, and knocks without explanation under high boost.

Yes it can be made to work, but ultimately it requires retaining the AFM, either during tuning (utilising "Air Flow Study/Learn Function"), or permanently. For consistent results, permanent installation of AFM is required and setting:
• Toyota/Nissan: Parameter 2 => Option Voltage Output 1 or 3 => AirFlow Input 1.
• Mitsubishi: Parameter 2 => Option Frequency Output 1 => Air Flow Input.

There is one reasonable solution for the Mitsubishi EVO9 - flashing the OEM firmware to TephraMod firmware with speed-density support. The trick is to use 1:1 AFM:MAP conversion and 100% RPM conversion, then define the MIVEC table similar to Haltech's default.

Retaining AFM (for OEM ECU cam control purposes only) has successfully worked for:
• Toyota Supra JZA80 vvti automatic gearbox.
• Nissan Stagea VQ25DET NM35 eVCS automatic gearbox.

You could argue that HKS VALCON would solve cam control issue, but it's cheaper to retain the AFM (or if possible, flash the firmware).

Examples of really bad tunes

Toyota Supra (JZA80) 2JZ-GTE VVTI automatic

This tune is a very good example of an extremely conservative ignition tune, so much so, that it risked melting a piston running too little ignition and too much fuel. It probably made more power at 10psi than at 16psi, and 16psi made more power than at 20psi (confirmed on dyno)! Running anything less than 15 degrees under boost (with correct AFRs) is just asking for trouble. Yes, this was tuned (in late 2011) by an authorised HKS pro-tuner with 15 years experience, who is now bankrupt (since mid-2012)!

This ignition table was not the only problem. The complete list of wrong tables and parameters include:
• main IGNITION.
• main INJECTION.
• target AFR (table and mask)
• MAP-to-airflow table.
• acceleration (throttle change) - injection.
• individual cylinder correction - injection.
• ECT (hi temp) - injection
• ECT (hi temp) - ignition
• IAT (lo and hi temp)- ignition
• knock trim (4500 RPM+) - ignition
• auto compensation (and detection) - ignition.
• Injection timing.
• Parameters: AFR learn (and of course, the amount of learn change it does per cycle!)
• Parameters: speed limiter was set incorrectly.
• throttle tangent calculation time was set incorrectly.
• RPM and pressure scale were completely fubar'd.
Ignition table

Subaru WRX STI v5 (MY99)

Another example of a tune where the opposite has happened - too much ignition! Running this type of ignition is only advisable with ethanol fuel, not street fuel! This was tuned by a HKS authorised tuner in the UK.

List of things wrong:
• Main IGNITION table, running way too much ignition, and ATDC in 100% vacuum.
• Main INJECTION table, running ~12.5 AFR in vacuum.
• injector flow rate off by 10%.
• injector dead time off by 15%.
• original tuner utilised absolute throttle correction (injection) table, for who knows what reason.
• acceleration (throttle change) (injection) table completely wrong.
• acceleration (throttle change) (ignition) table utilised for no reason.
• individual cylinder correction (injection) table was done on incorrect cylinders.
• individual cylinder correction (ignition) table was done on incorrect cylinders.
• ignition dwell time vs RPM completely wrong.
• ignition dwell time vs battery correction completely wrong.
• decel fuel cut RPM set too high.
• start ignition timing was not set.
Ignition table

Nissan Skyline R34 GTR

Another example of too much ignition. It seems like a common trend for tuners to put too much ignition so they destroy the engine, take zero responsibility of such event happening, then get extra work for rebuilding the engine. The engine is not the problem, it's the tuner that's the problem. This was tuned by a now-bankrupt HKS authorised tuner. It is no wonder they are bankrupt, they destroyed a few too many engines before word got out on the forums to avoid these tuners at all cost.

Ignition table

Mitsubishi EVO9

This is a special example - 2.2L stroked with Cosworth head, 272 cams, and 1000cc HKS injectors.

There were quite a few things eventually found to be wrong:
• HKS MAP->AFM table was left as default. Big mistake. Solution: OEM ECU flashed with TephraMod-SD, and HKS MAP sensor signal sent back to OEM ECU to provide pseudo 3 bar Mitsubishi MAP sensor signal.
• Throttle / Acceleration tables were too lean.
• Cold start was not tuned.
• Idle tune at 1250 RPM was not done, so it was hunting until 50C.
• Too much ignition at idle. Reduced so it could idle at 900 +/- 50 RPM.
• Oil pump was off by 180 degrees.
• HKS clutch eventually failed. It was randomly slipping at 0-2500 RPM.
• HKS FPR eventually failed.
• Fuel pump was still wired via resistor and original thin wiring.