ECU Tunes and Installs

[Gammaboy]

Awesome, cheers!

[Jason Broadhurst]

your desktop pic needs more tits.

I feel that way about wind turbines.

[Jason Broadhurst]

Last week, Burns and I did the final cutover from the PowerFC to the adaptronic. The Adaptronic is actually a plugin for a R34 but because the engine harness has pins blanked off, it retricts additional non oem wires being added. To get around this, we bought a 150mm plug to plug extension which we can modify and repin to suit.

We bit the bullet and cut away all the usual stupid modules attached to these things because japanes. The look was unpinned from the plug and repaired with heat shrink or new pins.

We then installed the 150mm long patch loom which is almost straight through. The spare wires will be taken to additional equipment like flex, boost etc etc. It also allows us to fit pull up resistors and circuitry for the adaprronic to modify to keep the dash and LCD operating perfectly and remaining plug in.

The attesa system and diff control is done in the ecu also, the programming can be seen in the logic diagram. This isn't development mode, just the normal software anyone has access to. It comes with a great simulator too, to test code offline.

We then hit the road for a 6k and 8psi tune. This will be it until the 10,000km run in ordeal is over haha.

[Jason Broadhurst]

Not very a technical video, but this is Tim's 510 1600 on the dyno.

There were some very minor teething problems. 10 year old tires egged, a power problem to a fuel pump and the internal wastegate actuator needed a little more pretension. Once all those were sorted (2 new tires at the 11th hour Saturday evening and lots of wastegate fiddling) we got the rig to 315HP with a very healthy power distribution. Plenty of start stop, with another tune being completed during the tire excursion but a very solid outcome on a very well prepared car. It was a joy to tune.

[garvice]

Sounds like a weapon

[Jason Broadhurst]

Nearly got to the M1-300 club except for short gearing

[Jason Broadhurst]

I had some triggering issues above 5500rpm and loaded, so I set the limiter to 5000 and it was making 360HP on 15psi. Looks like it will reach 500HP easy once triggering issues are sorted.

[Jason Broadhurst]

JZK25 asked me nicely to look as this turbo conversion they completed had some problems with the VVT and triggering.

It all came down to home cam angle sensors, and the additional trigger input required for the second VVT cam. The Elite 2500 has a single REF wire, which when used with the single REF input gives you your engine position really nicely. That is until you have a V engine, and need to detect the opposite banks cam position because of VVT.

VVT works by the ECU controlling a solenoid which pumps oil in to the advance mechanism in the head. When the oil is pumped, the cam advances. For VTEC you just use a switch at a certain RPM and it changes very abruptly. For VVT now days, it adds it continually variably to have peak VE at every given engine position. It does this by PWM'ing the solenoid and using the cam angle sensor as feed back and a PID control loop. The cam angle sensor is usually static to the crank angle sensor, so and deviation from it's defined position is assumed cam actuation.

What was catching us out, is that the HOME wire has specific voltage levels for HALL sensors, lets say 3V for on and 2V for off. That means that on the below waveform, if it goes over 3V it is active and when it falls below 2V is the falling edge trigger. The HOME input is fine, but the multipurpose SPI input for the elite uses different trigger voltages. These voltages are 2V on and 1V off (Confirmed by Haltech ). Now if you look at the standard waveform of CH1 on this scope pic, it easily activates above 2V, but never falls below 1v to trigger the falling edge, therefor the cam angle sensor and VVT on that bank do not work.

Looking at this picture, the 1-> at the bottom of the screen represents CH1 0V. The line for CH1 starts at 1.8V and rises to 4.5V for on and then back to 1.8V for off. The top squigly line is CH2 and the crank angle sensor reluctor. You can see the two missing teeth (gaps in wave) and 34 actual teeth. When the VVT solenoid actuates, the bottom line, CH1, moves to the right or left. At 10 degrees advance, it would be shifted 1 tooth across compared to the top wave which stays constant.


The fix was to switch off the internal pull up, and add a variable potentiometer resistor as a pull up, and find the resistance required to meet both the arming voltage and the falling edge voltage. This can be seen in this video. You can also see that the target VVT angle and actual follow nicely due to some closed loop PID tuning. Without this tuning, the target and actual would not follow so closely and engine VE would not be reliable.

[BigMuz]

Very good.

Left cam has pid overshoot though needs more damping.

[I'm DJ!]

Clever cunt. Bravo

[10sec rx7]

Have put in the request to get the on/off voltages user adjustable added to the next firmware

[cracka]

Interesting the response is so good at idle. What's oil pressure like?

[Jason Broadhurst]

That was hot, so maybe 20psi? At RPM it will be interesting to see the VVT PID tune because of the increase in oil pressure changing the control behavior. You also won't see a 25 degree step change, it will be more like a ramp which will dampen itself.

[BigMuz]

Ahh, fair cop.

What fpr does it use? Boost referenced or rising rate?

[elfturbomax]

In the vid you appear to be discussing things as if the scaling is 1v/div however the scaling looks to be 2v/div.
The initial problem is probably also a rookie mistake, it took me a while to find when I did one. It is not a problem with ECU channel thresholds.