Mechanical VNT Control - PeachParts Mercedes-Benz Forum

libbybapa · #1 09-02-2011, 11:25 PM

For several years I have worked on various mechanical VNT vane control systems. Most of what I have done has been on VW's, but I also installed a gt2559v along with an intercooler on my wife's '82 300TDT a few years back. My mechanical vane control has gone through several variations and at this point is refined to the point of IMO functioning ideally under all the various driving situations. My basic initial design goal was to design a system that would keep the vanes open as much as possible in order to minimize back pressure and so assist in overall efficiency, yet would close the vanes proportional to the demand for power and finally would override the pedal position and open the vanes the required amount in order to not exceed the desired max boost. Secondary design goals included preventing the vanes from sticking and maintaining, as closely as possible, the stock feel of the accelerator pedal. This current designs meets or exceeds all of those expectations. Very fast boost response and yet proportional to pedal position along with accurate max boost control. The feel of the accelerator pedal is entirely unaffected. The overall effect is entirely positive and a fantastic transformation of the driving experience. Here are some pics:

Basically, the vanes are closed proportional to accelerator position and then opened by the inline boost can. The boost signal to the can is regulated by an adjustable ball/spring valve (manual boost controller) and pressure is bled back out of the boost can via a check valve. The little spring mechanism in the linkage rod allows the accelerator to return from full pedal if the boost can is fully expanded when the vanes are fully open. Otherwise the accelerator would be locked in the full pedal position. Cheers.

OM616 · #2 09-02-2011, 11:34 PM

Very interesting!!

I am setting up a VNT on my 240D and have been thinking of different ways to get it to work as I want. You have spawned some insporation.

Thank you for posting.

libbybapa · #3 09-03-2011, 12:15 AM

A VNT turbo has a "V"ariable "N"ozzle "T"urbine. They are also known as VGT (variable geometry turbine). They consist of a series of movable vanes around the perimeter of the turbine wheel. Because the angle of the vanes is adjustable, they can be positioned to spool the turbo quickly and moved to flow freely. They allow the compressor to operate in its efficiency range throughout a much wider range of rpms. Max boost is controlled by vane position rather than a wastegate which allows exhaust gases to bypass the turbine. The net result is much faster spooling, better high end flow and increased efficiency. If your '98 TDI has the ALH engine then it has a VNT turbo fitted.

Dieselboy740 · #4 09-03-2011, 01:42 AM

I stared at it for a bit kinda confuses then I figured it out. Nicely done.

scottmcphee · #5 09-03-2011, 05:19 AM

Are you doing anything to add more fuel at the lower RPM where more boost is being made. Or is this mostly an efficiency (anti-smoke) device that helps make the best use of burning stock fuel flows?

Talk more about "fantastic transformation of the driving experience", if you can.

libbybapa · #6 09-03-2011, 08:54 AM

The ALDA is not rpm dependent and already automatically increases fueling according to boost pressure.

I live at 7,000 ft where stock lag was extended and off-boost power was very sad. The VNT turbo doesn't change the maximum power from what it is stock provided you keep the same max boost setting, but now if floored, boost is almost immediate and jumps very quickly up to 16-17 psi where I currently have it regulated. The power delivery is much smoother. Smoke is also reduced, but I did adjust the fueling so that at full pedal there is a slight haze.

R Leo · #7 09-03-2011, 09:12 AM

Nice work!

Sort of along the same lines, I've wondered about using the VNT off of a Sprinter and controlling the vane actuator with some sort of OTC process automation software. You'd need some transducers for intake pressure, air temperature and throttle position but other than that it seems pretty straighforward. It would have the added benefit of being easily changed as you learned what worked and didn't (provided you didn't blast something to kingdom come).

scottmcphee · #8 09-03-2011, 10:55 AM

Neat stuff. Doing some quick reading on VNT's suggests that geometry controlled by engine RPM alone optimizes the boost potential. Yet you've chosen to control vanes via throttle position (with a manual boost controller to limit it).

What do you think the effect would be if crank rotational speed (sensor) was used as the sole input for an electronically controlled actuator for controlling vanes? A mapping or table lookup could be used to create an arbitrary relationship between RPM and vane angle. The idea being to dial in the vane angle that is optimal for that RPM at all times, and it's your right foot that ultimately decides how much fuel goes in to take advantage of the boost. But there would be no direct linkage between pedal and vanes.

scottmcphee · #9 09-03-2011, 11:05 AM

"ALDA is not rpm dependent and already automatically increases fueling"

Some might argue the ALDA controls a fuel limiter and by itself cannot increase fuel.

Suppose the ALDA is removed completely from the car (as it is in my case) and now only the other devices in the IP (rack limiter, element size, et al) and your right foot control fuel flow. What kind of impact might this have on your VNT design?

sixto · #10 09-03-2011, 12:27 PM

Can a VCV be reworked to provide a VNT signal? How about the EGR vacuum signal in a 617? Heck, might as well put the EGR or ARV logic in the EDS to use in a 602/3.

Sixto
87 300D

bigblockchev · #11 09-03-2011, 03:00 PM

The designation TDT is in common enough usage that R-ckauto uses it to distinguish the wagon from the sedan. It seems that your approach to a mechanical control of the VNT is more likely to be successful for the majority of people. Electronics are great for engineers but problematic for the average Joe. I can build a linkage based system and have it work long before I got an electronic version to function. Perhaps a slightly more efficient setup could be produced electronicaly but how would you know what factors to optimize and how reliable could it be made. KISS is one way to look at it. Or my favorite "When in doubt, make it stout, of things you know all about" Cheers Dan

netboy · #12 09-03-2011, 03:38 PM

MEANWHILE,BACK AT THE FARM............ So are you saying that you removed the stock turbo and this V-DUB turbo bolts right up. And do you have intercooler pics? By the way it is an aviation concept. All new axial flow turbine engines use variable vanes for more efficient air flow across the compressor blades.

libbybapa · #13 09-03-2011, 04:31 PM

My initial point in posting was to outline the mechanical vane control for others who are contemplating or even already undergoing a VNT install. My focus was not on the extent of the whole project. Most of the project was completed several years ago and the initial vane control was fairly different in design.

It is not a v-Dub turbo. It is from an "other market" 3.0L Opel/Renault. It is not anywhere near a direct bolt-on. I welded an adapter plate to the Merz exhaust manifold, made a coupler to fit the Merz exhaust, custom oil supply fitting, custom oil return line, custom turbo support bracket, custom air cleaner to make room for it all. The turbo I used is water-cooled and so I even added an aux coolant pump and changed the plumbing somewhat. The orientation of the turbo center section is different from the stock application, so it needed to be clocked which w/ the VNT turbo requires precise drilling and tapping of three holes in the turbine housing. For the intercooler install I moved the radiator/condenser toward the engine 1", modified the fan shroud to fit properly, modified the lower radiator hose, etc. I'm sure I'm forgetting plenty... Definitely the whole project was a significant undertaking. I would probably opt for a water to air intercooler system if doing it over. My wife is driving the car right now so I can't take any pics ATM. This is the best I could find right now:

The pipes actually affect the approach angle less than they seem in that pic. I've always planned on adding a valence to hide them.

Marvelicious · #14 09-03-2011, 04:45 PM

Quote:

Originally Posted by libbybapa

...Some might say all of this is moot but they would be the ones thinking the common misconception that moot means irrelevant whereas it really means arguable.

1. open to discussion or debate; debatable; doubtful: a moot point.
2. of little or no practical value or meaning; purely academic.

More that the argument is irrelevant because it is impossible to for the argument to be concluded... But thats a moot point...

Nice work anyway. How does the turbo compare to the mercedes exhaust size and shape wise - what did your adapter look like?