Supercharge Mitsubishi Pajero/Montero 6G72

[Tracy Macgrady]

The Mitsubishi has various models of their full size SUV are using 6G72 engine. This model of motor has been in the market for more than 20 years, way back to the early Dakar racing. There are still large amount of pajeros on road running this engine, v6 3000cc displacement with 12-valve and 24-valve version. The engine is durable but in general lack of power. The 3.0L engine yields about 147 horse power by the spec.

My Montero has ran 400K millage, it still behaves as it should. Along the engine rebuild, I decide to supercharge the engine with a Toyota SC14 supercharger, let’s see how it goes.

6 steps in general,

1. Do the math, it gives you a theoretical validation of what I try to achieve.
2. Mounting, to put the supercharger mount on the engine, and put on the right pullet and bells.
3. Piping put on all kind of hose to connect the supercharger to the intake and air filter.
4. Build the supercharger controller circuit, which includes a microcontroller with a step motor driver, supercharger clutch control, and all kind of input signals (RPM, TPS and etc).
5. Write a small piece of software to control the air bypass valve (ABV) and supercharger clutch when the accelating pedal is pressed at the right RPM.
6. Tuning, with different parameters of opening the ABV and engage the supercharger at different RPM and TPS angles.


I bought a SC14 supercharger from a junk yard, it’s easy to aquire and with a reasonable price.





the SC14 has 1420cc on 1 cycle, the 3000cc engine would suck in 1500cc with 1 cycle of cranking， therefore the SC14 would have to rotate slightly faster than the crank to fill in the same amount of air as it is running as a natural aspirated engine. When we try to boost the engine, we would definitely want to pump in more air with the supercharger. As I want to make a quick test, I used the original pullet, driving the supercharger with the wheel that drives the power steering. I measure the diameter of the driving pullet has 1.2 timers bigger than the pullet on the supercharger, so it should work, I decided to start from there.

[Tracy Macgrady]

Attachment 19501
Attachment 19503
Attachment 19502
Attachment 19504

[Tracy Macgrady]

[Tracy Macgrady]

The mounting was not so difficult, but still take quite a while, as you need to it the mounting on and off to manipulate the position and make the supercharger fit in the engine bay

[Tracy Macgrady]

one the mounting is done, I started to piping and connect the hose in between the intake ,supercharge, air filter and the ABV. I bought a air filter case of the jeep Cherokee, and it’s cheap and narrow enough to fit in the engine bay. Of course, I still need to modify the tube to fit in the airflow sensor.

[Tracy Macgrady]

eventually it looks okay…


I bought and trimmed a 135-degree aluminum tube into 2 pieces, one on the intake side, one on the airfilter size, fully utilized.


I disassembly the supercharger, it’s pretty simple indeed. Two fly wheels with two air blades, 4 bearings and a magnetic clutch, that’s the SC14 all about.

[Tracy Macgrady]

Building the circuitry

The SC14 has equipped with a step motor controlled ABV. Compare with the earlier model, it can control the ABV to open or close with more precision. The problem is, once I got the SC14 from the junk yard, I don’t have the ECU of the engine that runs the supercharger. Even I got that, it just won’t work as a separate independent module with my pajero ECU. In that sense, if I want to control the ABV, I would have to build my own supercharger controller to control the ABV and supercharger clutch. At the same time, I would still need to read the engine data from my parjeo 6g72 engine, and feed the run time data to my supercharger controller.

I bought a microcontroller, and step motor driver and relay modules from a website, and start to build the circuit, just like to old days when I was in school.

As I need to read in the RPM data, I tried to capture the PRM signal from the crank sensor.

the 6g72 crank sensor outputs a sequence of pulse, as I tried to read it from an oscilloscope, it shows 3 times faster than the RPM that I read from the dash. Probably the crank has 3 trigger points, so that the crank sensor would be triggered 3 times on 1 rotation.



as I read 44Hz from the oscilloscope, that means 44*60 sec./min = 2640, therefore I would have to divide the reading into 3, 2640/3=880 RPM, then it make sense to me.

now is about the TPS, there are 4 lines connects to the TPS, VCC, GND, TPS signal and a TPS switch signal. I was initially connects the TPS signal only at the beginning, but I lately found the controller would not able to make the right decision when I am pressing the accelerating pedal but the RPM is high (down hill, engine brake), as I would like to have the supercharge clutch to remain engage on this situation, therefore I take the RPM into the calculation.





When the intake shutter is closed, the supercharge drives by the engine, the RPM won’t suddenly stop, pressure in between the intake and the pipe connects to the supercharger. If the pressure cannot be release properly, it would damage the hose, intake or even the engine, that’s why the ABV comes into the play. The supercharger I bought has a ABV drives by a 6-wire step motor. The motor is drive by a sequence of pulse to control the rotation steps in both directions. By using this, I have to find out what are these wires, and what sequence to drive it open or close.

[Tracy Macgrady]

it’s not difficult to determine the wires, you could search on the web, there are a few way to identify the wires there. Encode the sequence to drive the motor, here is the first driver program I wrote, it works pretty well, just make it on-and-off, here is the video of the test



I soldered a mother board, put the microcontroller, drivers and relays on two layers.



the step motor driver is basically a Darlington driver(on the left), and the relay is on the right.(to control the supercharger clutch)

[Tracy Macgrady]

[Tracy Macgrady]

5. Software

It’s been more than 10 yrs since last time I wrote microcontroller programs when I was in school, but eventually it proves my poor memory still works somehow.

The nowadays solution is so advance, so convenient. A USB line can link to my macbook and complete the download process, without taking the chip out from the circuit.



Maybe I need a bigger pullet to push the boundary to 10PSI, and change the fuel injectors with a larger volume, maybe next week.


so far everything works pretty well