Pavlo

I have finally got round to implementing the John banks special boost mod, 5k pot and zener diode.

This appears to work, now I am reading 11 psi when it's actually 15psi.
Trouble is I am having trouble getting the boost up, I initially tried a needle valve in the boost solenoid to actuator pipe, but this had a marginally affect. My reasoning was at a given duty cycle, more air should be bled to atmosphere, oh well.

Then I bit the bullet, got the car on the ramps and removed the heatshield to get at the actuator. I tried to turn the arm in 2 turns, but this is a IHI VF12 turbo on a 94 Legacy engine, and I don't think it did what I expected.

So now I have put back in the valve, but soldered a small piece of capillary pipe (0.8mm ID) to bleed air off between the valve and the actuator. This gives me the ability to get serious boost, but I don't think the ECU has as much control over it.

I am now seeing about 15-16psi in 2nd gear, upto 17-18psi in 3rd and 4th. Boost starts to tail off gently after 5000 rpm, leaving me with about 11-12psi at 6500rpm in all gears (I assume 4th and 5th too). Then I turned the brakes blue and limped home.

I am little worried as the peak boost seems to vary depending on how I drive, when and how I apply the throttle.

I was thinking about putting a kink in the actuator arm,any better ideas?

Paul

john banks

You ask difficult questions - my bit was the electronics

In my car there is a 1.2mm restrictor in the line off the turbo outlet before the T-piece that splits to the solenoid and the actuator.

Replacing this restrictor with a variable valve or smaller restrictor has worked well on several cars.

A Dawes is another option.

Alternatives include tightening the actuator if a disconnected air feed does not give you the boost you require. Or you can bleed off on any of the hoses connected to the T-piece. I don't like the idea of excessive boost leaks to atmosphere though as it could potential increase the turbo outlet temperature and pressure or make the MAF overread for the amount of air in the system but I am probably being paranoid.

The other thing is that the MAP sensor adjustment circuit benefits from some calibration so that the closed loop boost control can use a new target - so if it aims for 11 PSI and you want to run 15 PSI then you need it to show 11 PSI when it runs 15 so that there is some control.

A lot of it is trial and error finding a site that does not result in peaky behaviour. Stick at it - I found some very funny/unstable combinations along the way.

Pavlo

"Stick at it - I found some very funny/unstable combinations along the way."

That is good to hear. I originally mapped 15psi to 10psi, but when the valve&bleed was in, the boost looked like it would go 18+psi in 4th gear. So now I think 14psi MAP equates to 10psi ECU.

The bleed to atm is very small, 0.8mm diameter, about 6mm long. Without it the boost would spike, returning to about 11psi at 4000rpm, I assume because the duty cycle wasn't enough to get it any higher. The problem seems to be that the pre 97 ecu, the duty cycle map represents the highest duty cycle at a given load/speed. So it won't increase it, even if the measured boost is lower than the target.

I will be mapping the ECU shortly, but this is very useful stuff.

Incidenty, and I did mention this to Stephen Done, I think the boost and duty cycle maps must use throttle position as the load points, rather than air flow. Does anyone know for this for sure?

Paul

john banks

Regarding the duty cycle maps - the AE800 and 801 use the same values across all loads which is why they are perkier on part throttle. Not sure what the load input is - it is certainly a different scale to the load inputs on the fuel and timing maps (which go up to about $2A) which seem to use MAF, so it could well just be throttle position. The scale used if using 5.05V 8 bit ADC would have the last column ($C0) at about 3.8V which is not far off full throttle.

Pavlo

96 ecu has similar boost/duty cycle load point values. The adc in the old ecus is abit 2v full scale. I suspect however, it takes 5v externally via an instrumentation amp. C0h equates to 76% from FFh, and the difference airflow wise over the last 25% of throttle movement is very small. I think this makes more sense.

I also suspect the the load for the ignition/fuel tables is actually load/rpm, because there is now way on earth that you are going to get as much MAF at 800rpm as 6600rpm, but the maps are full of seeming useful values at low speed/high load.

Paul

john banks

There are redundant values at high load below about 3500 RPM on my car and a few in the bottom corner from 6000-7000 RPM. You can see the map being accessed when you drive the car, and it is as you say.

Thinking about it, I know the boost target hits the last column even at low revs, so it must be TPS not MAF.