SV3Power.com - Exhaust systems and components for the Toyota MR2

7A-GE DIARY - THE MIDPORT CONVERSION
or
FITTING A SMALLPORT INTAKE MANIFOLD TO A BIGPORT HEAD

A smallport manifold won't work on a bigport head. It will bolt on, but the difference between the port widths is so great that the flange on the smallport manifold doesn’t even cover the big intake port. You could use it, but you'd have to spend $100 on a manifold adapter like the AE86 kids use to match bigport manifolds to smallport heads. Besides which, the smallport manifold won't flow enough for the bigport head, right?

Wrong, right, wrong, and wrong. It can be done, and sometimes it should be done. I will show you how to do it, but first, here's why it works.

For comparison’s sake, here are some measurements of the various components involved. I measured the dimensions of the ports at the face of the manifold and head as well as the dimensions of the gasket windows.

The oval scribed outside the port is the smallport gasket opening. Other lines show where I planned to remove material.
Smallport manifold, as cast: 26.5 x 46.5mm; 1081 square mm; equivalent to a 37mm ID pipe.

The #4 port, left, has been opened up to just inside the smallport gasket's window.
Smallport gasket: 31.5 x 51.5mm; 1409 square mm; equivalent to a 42mm ID pipe

The bigport head/gasket (no pic) is 30.5 x 65.5mm; 1798 square mm; equivalent to a 48mm ID pipe

Obviously the smallport manifold is totally wrong for the oversized bigport head, right? Well, maybe. The bigport manifold is divided into two circular runners, each 30mm in diameter.

The bigport intake manifold (no pic) is 30mm ID x 2; 1414 square mm; equivalent to 42mm ID pipe

The bigport manifold has about the same area as a gasket-matched smallport manifold! So if you were to gasket-match the smallport manifold - and make sure that it maintains that cross-section up into the runners - it should move as much air as the bigport manifold, right? But there’s more to it than that. There’s also the bigport’s TVIS plate. On cylinders 1 and 4, one of the two intake runners has the TVIS butterfly shaft through it, but cylinders 2 and 3 have the TVIS shaft running through both of their runners. That shaft is 5mm thick, meaning that there is a 5mm x 30mm obstruction in one or both of each cylinder’s runners. That is a 150 or 300 square mm obstruction, respectively.

The TVIS plate (no pic):
Cylinders 1 & 4: 1414 – 150 = 1264 square mm; equivalent to 40mm ID pipe
Cylinders 2 & 3: 1414 – 300 = 1114 square mm; equivalent to 38mm ID pipe

So suddenly the “bigger” bigport intake manifold with TVIS doesn’t look so hot. Even if I gut the TVIS plate and get back to 1414 square mm, each cylinder is still breathing through two small runners. That means extra surface area relative to cross-sectional area.

Bigport manifold, gutted TVIS plate (no pic):
30mm ID runners x 2: 1414 square mm, 188mm perimeter

Smallport manifold, gasket matched within smallport and bigport dimensions:
30.5 x 51.5mm oval; 1371 square mm, 138mm perimeter

The gasket matched smallport manifold has 97% of the bigport’s cross sectional area (i.e. size of the port), but only 73% of its surface area (i.e. amount of wall surface contacting the air within the runner). It has 32% more cross sectional area, relative to the surface area of the manifold runner, than the bigport manifold. That means less surface friction for the air traveling down the manifold - so, in theory, the smallport manifold can move more air even though it doesn't have more total port area. There is less boundary layer (air stuck to the port wall) and, theoretically, more air actually moving, inside the smallport manifold.

Even if we gasket match it, the smallport manifold is still misaligned with one side of the bigport head’s intake port by 14mm (though it lines up perfectly with the opposite wall of the port). It might flow enough air but it (probably) won't distribute it evenly to both valves. Unfortunately, there isn’t enough meat in the manifold to make it match the bigport head. About the best we can do is to port the manifold out almost to the edge of its flange, making the port about 55mm wide. The adapter plate can continue the resulting taper towards the “wide side” of the port, making the opening about 58mm wide where it meets the head. That is still 7.5mm short of the port wall in the head, but it is much tighter than the 14mm gap we started with (or 16.5mm if the smallport manifold was not gasket matched), close enough for me for the purposes of this experiment. I suspect the air will flow over that edge like it wasn’t even there. It's pretty far away from the valves, and pushing wall of the manifold back will tend to pull the airflow in that direction anyway.

In the process of porting the manifold and adapter plate to better match the bigport head.
The scribed ovals around the ports show the size and shape of the bigport gasket windows.

After a lot of pondering, I went ahead and ported the smallport manifold to the very edge of the flange, and decided to epoxy the valley between the 1-2 and 3-4 cylinder runners in the manifold with JB weld to support the thin edge of the manifold at the flange. It also provides a better sealing surface for the gasket between the manifold and the adapter plate. It’s very possible that I could even throw the adapter plate away and just bolt the manifold to the head with a regular bigport gasket, but I’m kind of afraid to try it. I don’t know how the JB Weld would work being exposed to the port. Theoretically, it would be in dry air, and it’s supposed to be impervious to gasoline anyway. As long as it doesn’t vibrate loose and break away, it should be fine. A length of Allthread threaded into the web supports it internally, and I roughed up and cleaned the surface thoroughly before pouring in the epoxy, so it certainly ought to be safe enough. The stock intake manifold is only set up for the stock 55mm throttle body, but I took a few minutes to hog it to accommodate a 60mm throttle body if necessary. On a near-stock 4AGE, the 55mm TB is more than enough.

Here you can see how close to the edge of the casting we got.
There isn't more than about one millimeter of aluminum on those inner edges.

As above, but with the adapter plate in place; much closer now to the shape and size of the bigport ports and gaskets.

Notice the threaded rod screwed into the reinforcing web to help anchor the epoxy.
Clear packing tape was used to form a dam to keep the epoxy in place while it cured.

Both webs now filled with epoxy.

The smallport manifold is longer than the bigport by the width of the TVIS plate, so they are now nearly interchangeable. It's hard to see, but note the slightly longer runner length in the bigport manifold as a result of the outward curve under the plenum. The smallport's shorter runners are better for high-rpm cylinder fill, while the longer bigport runners favor (slightly) lower engine speeds.

This is not a great photo, but the smallport manifold won't sit up straight for pictures like the bigport one will.

I hate cars sometimes. This was more hassle than it should have been.

More details are available in the other sections, accessible from the Articles page.