Interesting, my manifold has ~7.5" runners. I wonder is there is something about it being tuned to 75% of the wavelength the stocker is that makes it work for me. I seem to have lost no torque from purely subjective analysis.

 

Hi Bob, I have a VICS manifold here and be glad to measure it. If you could me a picture of exactly where to start the measurement point so my numbers represent something accurate.
This would also help me as I have made extensive modifications to mine and heard that when measuring plenum volume you don't include the runners. Maybe someone can confirm this and how to get an accurate measuremen.
-Jeff

 

3.85 is longer on the head side than I would have thought based on craming a tape measure up against the valve I was guessing more like 3.

 

With the head port there is some room for error since the port opening is at an angle. I put the depth gauge on my calipers down the side in what looked like the middle range of measurements until I hit the top of the valve.

 

I measured ports at centerline from flange to valve seat. They are all within a few mm of each other. The runners are the variable. They also have different number and direction of bends. So VE curve of each cylinder will be different. Thus why doing individual cylinder trim with OEM manifolds finds several hp with no other changes. The small ITB's we made 200whp and good torque with with were about 1/2" shorter than OEM plenum to valve seat.

 

When you have a ported head and cams like mine I think having enough airflow for the engine is 5x more important that runner length. Most if the race guys here use extrusion honingto open up the runners when forced to run stock intakes... basically getting 180 - 190whp from a NA8 engine with stock manifold and throttle body.

Basically my engine with stock NA8 intake = 175whp
Extrusion honed NA8 intake = 185whp
Skunk2 Street = 190 whp

My skunk2 manifold vs stock na8 picked up the most power @ 6500 rpm. That goes against what harmonics and runner length theory says. I'm guessing the power gain was mostly due to the runner area than anything else. Either way there is far more at play than just runner length.

 

Nope.

They are equally important. The best cams and porting possible will not reach their max potential on a bad manifold. The best manifold will not reach its max potential with the wrong cams and porting.

Either you have the ideally sized runners and plenum for your cams, or you don't. There is no magical point where manifold configuration stops being important.

With good enough cams and porting, you can make good power with a WAG manifold, as you have demonstrated.

 

Which OEM manifold? I sort think the 95 manifold is about a half inch shorter than the others, maybe more. The runners are all different lengths though 1 being shortest 4 being the longest. 1 through 3 are probably within about 1/4" or so of each other and 4 is just over an inch longer. was your ITB test done with stock cams?

 

Emilio, do you have any information on how the size of the plenum affects power relating to BPs? I've always imagined the plenum like a balloon. As the air pulse moves up the runner from valve close it pressurises the plenum then moves down the other runners. The optimal airflow is if that times with the next valve open. The larger the plenum the longer the delay so it suits a lower rpm. That doesn't explain why the dyno queen hondas run much larger plenums.

Even if how I think it works is wrong, plenum sizing not only helps total power but can shift it up and down a little. This applies to the Honda skunk2 as the plenum can be increased easily with spacers. Maybe a slightly larger plenum might help.

 

I only measured the three NB variants, no the NA8 although it is asymmetric too. The B6 manifold is actually quite nice if you take a close look at it, not as constrained by TB hitting hood and no EGR to distort runner positions.
The ultimate mean dimensions (runner dia, runner length, plenum volume, port angle) of the squaretop are almost perfect for sub 8000rpm power. It's just asymmetric and has obstructions everywhere. We are starting to fiddle with a manifold design I have had in my head for years (printing makes it easy). It won't look like a squaretop but it will have most of the same dimensions.

Starting point in plenum volume is roughly 1.1x engine capacity. Plenum volume isn't as critical as runner length though. In any case, if you have the luxury of being able to perform tests with different volumes, it's probably worth experimenting with. Pay note to throttle response as it will fade if plenum volume gets too big.

 

 

It is its freekin beautiful. just scan it in to cad and scale it up in every dimension a tad to fit the 1.8l

kind of thought about doing a shitload of cutting/ welding/ die grinding, shaping to do that but that would not be easy. and it would take 4 or 5 manifolds worth of cut up parts to achieve I think.

FWIW I rough measured the primary runners on a VCTS BP Protege/Escort GT manifold and they are somewhere between 14.5 and 15.5" long not including the port length in the head. Apparently Mazda thought there was something out there in that land to chase at one point.

Keep in mind the BP motor in a Miata in 1994 was rated at 128hp and 110 ft lbs with an NA8 manifold while the same time the motor in the protege/escort GT with the same pistons head and cams was rated at 140 hp and 118 ft-lbs. The biggest difference being this manifold.

 

If you want to look at runner lengths I was messing around with a spreadsheet for wave tuning. I think its only good for a general idea of what is happening at different lengths since there are so many variables that are hard to account for:

 

Here it is!



Unfortunately we changed a little too much to know what the manifold did, but despite many VICS/VTCS turbo installs, I've only ever seen the above blue curve. The red curve is with the honda manifold and has ~90 more hp up top despite only 4 extra PSI.

We changed to a better sealed Fab9 baby intercooler setup and upgraded compressor wheel turbo at the same time.

 

lulz. Certainly if we are going to take the time to build something in CAD, print and dyno test it, we're not going to copy any OEM manifold. We are just getting around to testing protos for our Supermiata Sport Cams. They are mild and what we'll optimize the manifold for. So no short runner Honda B series type geometry. I don't think we'll have a manifold OTS before the end of the year but the cams should be OTS by late Summer.

 

With tape measure pulled tight centerline is a little longer. Doing the same thing on the protege manifold it's ~twice the length and cylinder one is the one that's inch longer than the rest.

 

Nice... love the dead flat torque all the way to the limit. Have you got a pic of it all finished?

 

Post 932 is a finished pic pretty much

 

Results, Skunk2 Ultra B series intake manifold with miata specific Skunk2 64mm throttle body.
Car has BP5A engine swap with 35k miles, full exhaust and +4 degrees timing with a trigger wheel. BP5A ecu.

Red is VICS manifold, blue is honda manifold.

 

So VICS was better everywhere except the last 500rpm?

Wouldn't flat top take away that advantage?