Adapting the Garrett AVNT turbo 1

[JustinJones]

My current project is adapting a Garrett AVNT turbo to a gasoline passenger car application with performance in mind. The Garrett AVNT turbo is an electro-hydraulically activated VNT turbo, the pressurized oil supply from the engine is used to change the position of the vanes in the turbine housing. The spool valves directing the oil supply are controlled electronically by the ECU based on calculated engine load.

Here's what I have found out about the signal that actuates the variable turbine: "A pulse width modulated proportional solenoid provides infinitely variable control. Further, the design is completely self-contained with only a single electrical plug-in to the engine's ECU. As a result, all variable forces are removed from the control loop; aerodynamic loading, friction and oil pressure." -- cut/pasted from this article:

http://www.findarticles.com/p/articles/mi_m0FZX/is_10_69/ai_110471904

The Garrett engineer I spoke to today told me to ask a Ford or GM dealer (they both use the same AVNT center section on their respective turbos) what kind of signal they should get going to the turbo when they put the truck on the scope.

Maybe I can end up running it the same way I would run a fuel injector off my ECU, the ECU will think it's sending the signal to a fuel injector but really it will be going to the turbo...

Does anyone have any experience or comments about this subject?

 

[NickE]

Generally the important parmaeters with a PWM signal are voltage, current (load), and duty cycle. Frequency also comes into play, though that comes more from the duty cycle.

From my understanding the signal should be a square wave with the duty cycle representing the on time or (pulse width). So that if the on time is 10ms and the off time is 90ms, the duty cycle is 10% and the frequency is 10hz. The signal has to have engough power (current & voltage) to actuate the solenoid reliably without burning out its drive circuit, or cooking the solenoid.

Disclaimer: I am not an EE, I have a very basic understanding of PWM signals.

It seems to me that you would want a control system similar in action to a wastegate. IE: at low RPM, WOT the VNT system should extract the maximum amount of nrg (energy) from the exaust. As the RPM's increase the VNT system should back off the blades to improve flow and keep the compressor wheel from running out of its efficient range.

This sounds really cool, imagine a boost curve that is flat from 1500-6500rpm!

 

[Fabrico]

That may or may not work. On the upside, most automotive PWM outputs use steady voltage but vary the pulse width or duty cycle. On the downside, most PWM outputs are coupled with a return signal which lets the computer know that what is supposed to be happening is indeed happening. This signal is usually a range of voltage.

Power requirement should be about the same as that for PWM solenoids used in transmissions.

Where to find this output could be difficult. You may have to resort to using an aftermarket ECU or ignition control computer, along with an analog to PWM converter/driver. Some cars have separate idle air control computers with PWM output that may also work.

As an alternative, there are non-PWM solenoids that vary flow based on changes to voltage or amperage.

 

[JustinJones]

thanks for the replies guys, very helpful. I am completely out of my league on this project, lol. I am behind the 8 ball and have lots of reading/learning to catch up on.

 

[Majik]

The 2004+ Ford Powerstrokes run these. They are a PITA due to the fact that the solenoid or the vane system fails and usually locks the truck into like a .5AR housing... egt's skyrocket. engine goes weak. It's neat but now they just are waiting to break. Good for me I guess. Keeps business up.

 

[ivanheow]

how about junking the valve ,and running a wastegate actuator on the control arm ? regards robert

 

[JustinJones]

Robert I think that you are thinking of an old-style vacuum actuated non-AVNT garrett turbo like the Shelby-Chrysler VNT turbo.

Here's a picture of the AVNT Garrett I have:

http://vgturbos.50megs.com/photo.html

Mine is the same as the 88mm Garrett in the pictures.

 

[unclesyd]

Why adapt, take a look at this. Saw it on TV this AM.

http://bankspower.com/twin-turbo-products.cfm

 

[JustinJones]

the intent of my project is to utilize an electronically controlled Garrett AVNT turbo in a custom application, not to twin turbocharge a SBC.

 

[Warpspeed]

I ran a Garrett VNT on my car for around three years, and was never at all happy with the results.

There are several issues, the most important one for me was total lack of top end horsepower due to excessive exhaust back pressure. Sure, the vanes will control boost, but ALL the exhaust volume still has to pass through the turbine. With a wastegate much exhaust is bypassed, and exhaust manifold back pressure remains fairly constant once full boost is reached around the torque peak.

With a VNT, exhaust back pressure CONTINUES TO RISE right up to redline Rpm. The vanes work fine, but that is not the problem.

Garrett are pushing the VNT for diesel applications only, the GT range of ball bearing turbos offer much better performance for gasoline engines. Only Porsche have been able to get a variable vane turbo to work effectively. Many others have tried and failed.

These VNT turbos have been around for twenty years, don't ever think that the hot rodders and drag racers have not tried them. These guys are not all fools either. I wish you luck.

 

[NickE]

What about combining the VNT with a wastegate?

Nick
I love materials science!

 

[Warpspeed]

I suspect that Porsche may be combining variable vane technology with a conventional wastegate, I cannot see how else they can make it work. But I don't know that for sure.

It was something I was going to try myself, but never did. In fact, I have seen a picture of a Garrett VNT exhaust housing fitted with internal wastegate on the internet somewhere.

One thing is for sure, controlling both the vanes and wastegate position would require some fairly complex electronic mapping and control. It is not the sort of thing that could be done successfully with just a few diaphragm actuators operating directly off boost/manifold pressure.

Combine that with variable valve timing, and the results could be quite good. But I doubt if it is the the sort of thing an enthusiastic amateur could duplicate at home in his garage.

 

[JustinJones]

Warpspeed was your VNT turbo an AVNT electro-hydraulically actuated turbo or an old style VNT?

also what size of turbo did you have? what kind of engine was it mounted on?

 

[Warpspeed]

My turbo was the old original style mechanically actuated VNT, and I purchased it brand new from a Nissan dealer as a spare part.

This turbo is fitted standard to the 4.0 Litre Nissan Patrol 4WD diesel engine in Australia. That diesel engine is (if I remember?) rated at 160Kw. It was and still is the largest VNT turbo readily available here.

I fitted it to a 1.6 Litre DOHC B6 turbo Mazda engine. In Australia this engine is the Ford Laser engine. In America it is better known as the Miata engine.

As I was looking for very approximately 130 Kw, I would have thought it would have been a reasonably sized turbo as far as airflow capacity.