The Nissan VG30DET engine comes with a variable intake manifold with 2 throttle bodies. This page covers the installation of individual throttle bodies while keeping the variable manifold.
6 Cylinder engines have a problem that is unique to them. An intake manifold is basically a resonance chamber where the pressure waves from the intake runners ‘meet’. However, on a 6 cylinder engine the pressure waves tend to benefit, or counteract each other, varying with engine speed. This interference occurs between opposing cylinders, which means that on the VG30DET (having a 1-2-3-4-5-6 ignition sequence) cylinders 1-3-5 and 2-4-6 need to be joined. But in order to allow the manifold to switch between single and double (divorced) manifold, Nissan installed a valve between the 2 chambers that once activated by an actuator, allows all cylinders to share their chambers. This valve is installed in the slanted section on the top of the manifold;
Many other manufacturers use variable 6 cylinder manifolds such as Toyota and Porsche. I noticed their designs, but what really inspired me came from reading ‘Design techniques for engine manifolds’ by DE Winterbone and RJ Pearson (ISBN 1 86058 179 X), a book I purchased in 2008. Their data (from both simulations and testing) shows how the volumetric efficiency can be improved by altering the manifold shape. A couple of pages from this book;
Manufacturing and design of the individual throttle bodies
The intake runners are made longer by the thickness of the ITB’s, and a larger manifold collector chamber is fabricated for more power. This in turn led to a completely new manifold. The individual throttle bodies are supplied by Jenvey and are 30mm thick with a 45mm diameter. This is the same diameter as the factory manifold runners. The combination of the throttle bodies and taller velocity stacks increases the runner length 55mm from 205 to 260mm. The VG30DET ‘lower’ manifold (part that attaches to the heads) is welded up with additional material so that the bolt profiles for the Jenvey ITB’s can be tapped. After these are performed the manifold looks like this;
The installation of ITB’s requires a more complex idle air adjustment system. Each cylinder needs it’s own take-off point for the idle control unit. On Nissan’s that use ITB’s from factory the idle control system for each runner is made from large hoses that allow some interference between cylinders, undesirable in use with the variable manifold. The use of such large hoses seems to come from expected containment from crankcase venting into the intake system, a worry I don’t have. Instead, I installed hose nipples, fitting each to equal length hose into a small manifold. That manifold leads a single larger hose into the idle air chamber. The single hose is used to allow for easier removal of the lower manifold, pulling the soft hoses each time would damage them.
The size of the manifold chamber was limited by the width of the valve covers. I designed the manifold for a volume of almost 3 liters. The smooth entries merge into 70mm pipes that improve the flow in and around the manifold substantially. The chamber is made from 3mm thick aluminium and the edges are completed by sections cut from cast bends.
In order to prevent the flange from warping when the chamber is welded to it, I cut a steel plate to size and bolted the flange to it. The manifold is then welded along this edge from the inside leaving more space for the 52 M5 bolts holding it down. I went with so many bolts since it offers a perfect seal with just liquid gasket, and during testing later not a single leak was found. The other welds were done externally to keep a smooth internal flow.
In order to link both banks and synchronize them, Jenvey supplies round levers with quadrants linked by a short throttle cable. These levers are offset from the throttle bodies by shafts fitted into housings with bronze bearings. Within, compensators made by Huco allowing for a little eccentric movement avoid seize. It’s an effective system though I noticed the steel lever and quadrants slightly rust. Being in the face of my engine, I manufactured new aluminium wheels and quadrants.
In addition to the linkage above the throttle position sensor has to be mounted offset as well, in similar fashion. I was unable to source the Huco compensators so I milled them myself. As I’ve learned from working with other ITB systems the large surface area of the ITB’s allows for many small air leaks, increasing the idle speed, or disturbing it. It is very difficult to adjust for this with the idle stop screws. (Individual) throttle bodies used on factory engine use moly coatings around the edges of the throttle plates and housings, making them leak proof when fully shut. Up until about 2014, manufacturers such as Tomei sold these coatings for DIY purpose, but environmental concerns banned them. Instead, I use a spray can of dry moly lube. The idle dropped about 250 rpm’s from doing this and stabilized.
The finished manifold prior to adding the variable chamber device, with the older, steel Jenvey levers and quadrants;
Variable manifold design
What I’ve learned from studying various variable 6 cylinder manifolds is that the distance between the 2 intake banks must be as close as possible. If the distance is too great, the manifold will still behave as a double 3 cylinder, as the strength of the waves decreases over distance. The first design I came up with appears as 2 throttle bodies on the outer sides of the manifold chamber;
A downside to this design is that cylinder runner 3 and 4 have a much longer distance between them than the other cylinders. In fact, this design may lead to unequal pressure waves for all cylinders, as the travel time throughout the manifold is effected. For example, a wave from cyl. 1 will quickly meet that of cyl. 2, but a part of the wave will go all the way around, and meet again at the other plate. The causes all kinds or various wave action inside the manifold which is highly unlikely to be beneficial.
Instead of the double valve design, I considered a triple valve design, with smaller plates in between 1-2, 3-4, and 5-6. However, this would also introduce some difference in wave action and it would be very difficult to link them without a complex system that disturbs the waves inside the chamber. It is important to keep a smooth inside to the manifold chamber and reduce volume as little as possible.
Eventually, I drew a design that uses a long, single plate in the middle of all cylinders. It uses bronze bearings on each side, larger roller bearings being ruled out as they take up too much space. The linkage is done by 1/4″ socket extension adjusted to fit, allowing compensation for misalignment. The shaft exits the chamber through a 3rd bronze bearing, that is 100% airtight by an oil seal.
The upper part where the plate rests against has a wide surface area that will mate it to the manifold chamber ‘roof’. This is done with JB weld, and it does not only allow for smooth airflow along the rounded edge, but it also reinforces the roof of the manifold which is not very well supported from high boost pressure blowing it up.
The variable and other stationairy plates are all 5mm thick, but cut down in the middle to save weight. The complete system is aligned and ‘JB welded’ into place. After drying, the extending sections are milled down carefully for a perfect flat fit against the manifold baseplate.
As the car went to the dyno for testing I used a small wheel to manually close and open the valve. Prior to the test I couldn’t predict wether the valve had to be open or shut, so it was pointless at this time to install an actuator.
The results were obtained on the Moritz tuning dyno at the same time the VVT system was tested. Since the engine ran best on the 22° intake cam advance setting, I used this setting to test the manifold valve. I had expected the manifold to run best with the valve open up to around 5700rpm and from there on with the valve closed. The results were clear, showing even more than I had expected.
With the valve open, a substantial amount of power was lost up to 132km/h (4400rpm) where the 2 gradually meet, at 145km/h (4800rpm) the opened valve shows a gradual increase that clearly climbs, showing how the difference would increase had the operator run the engine to the rev limiter. One thing I found, was the strong difference in wave action, is the fact that whenever the graph in one would raise a bit, the other would see drop. This is best visible in between 95-135km/h.
What the data shows is that the valve should be set as follows
-Open up to 75km/h (0-2500rpm)
-Closed to 145km/h (2500-4800rpm)
-Open from 145km/h (4800-????)
It is possible that close to 8000rpm the manifold might need to switch again, but this goes beyond the rpm limits of this test.
The actuator is a former turbocharger actuator. Though there are lighter, plastic actuators in use on engine with variable manifolds these tend to break down so I opted for the safest option. The actuator reponds to boost pressure and can be switched through a simple solenoid.