Intake System Design        Back to Carb Section

This section will detail the how and why of intake system design. It's based on a simple formula for calculating what should be a "close to proper" intake tract length. The formula is 90/rpm (in 1000s) = intake tract length in inches. Don't ask me where the number 90 comes from, I don't know, it's a constant quoted in several engine tuning books. You divide 90 by the desired rpm in 1000s, for example 3000 rpm would be 3. For the 2001 race season it looks like the CD-1a (or perhaps I'll call it the CD-1b now) should be able to reach about 14,000 rpm reliably once the new clutch is installed. Using this formula would generate an intake tract length of  6.42 inches. Ok, lets back up, although I want the motor to run well at 14,000 rpm I do not want it to ONLY have power at that rpm. I also need some power at lower rpms. I came up with a figure of 11,000 rpm as the point where I want to produce the most reliable power. It's a compromise, if I tune the motor for 11,000 rpm it should run reasonably well at 14,000 and as low as 8,000. (we will see later that you can easily change this "power band").

So.. working with 11,000 rpm as my target, using the formula above I come up with about 8 inches. (note you can use the formula in reverse, pick the rpm you want, divide the rpm by 1000  and divide 90 by that number).  Let's see how we can come up with an intake system that is about 8 inches in length.

First of all the length of the intake tract is measured from the actual intake valve head, to the very end of the carb inlet. One aspect we cannot effectively change is the measurement from the valve to the outer edge of the intake port on the cylinder head. That measurement will be about 2.5 inches on almost all honda 50 and 70cc motors. Actually it's a little longer if meaured along the outside of the port's curve and a little shorter along the inside of the port's curve, the average is about 2.5 inches. So, there's nothing we can really do to change that, we must start with 2.5 inches.

Next we must consider the intake manifold itself, I am using a Takegawa 18-20mm intake, the short almost 90 degree type (monkey version I think). It measures 1.5 inches along the short side of the curve and 2.5 inches along the long side of the curve, we will take the average of that, 2 inches, this would be the length of the path along the center of the round manifold passage. Our total length is now up to 4.5 inches.

Next thing to consider is the carb itself, how long is it from the outer edge of the inlet, to the outer edge of the manifold flange. Our example is a 20mm Keihin carb that measures 3.5 inches. Our total is now 8 inches.  If we consider the small spacer between the intake manifold and the cylinder head add another 0.25 inches. This give us a total of about 8.25 inches. If I use this figure in the formula it give me an rpm of 10,900 rpm. very close to the 11,000 rpm target I wanted. I got pretty lucky this time, everything worked out to be the right length. However, if I wanted to drop the rpm down to around 10,000 I would need an intake tract length of 9 inches, what's a good way to do this? Easy, you just add a velocity stack (air funnel) of the proper length to the carb. Intake tract length is measure from the valve to the outer edge of the carb inlet, an air funnel adds length to the inlet of the carb so it "become" part of the intake tract length.

Ok, what if I wanted to tune the intake system to work best at 12,000 rpm instead of 11,000 rpm, I'd have to shorten the intake tract length to 7.5 inches. On my particular motor that would be a little more trouble. I would have to attempt to shorten the intake manifold in some way, perhaps machine off some of the thickness of the flanges where it bolts to the carb and cylinder head. I could try to use a thinner spacer between the manifold and the cylinder head (however this would add more heat to the intake, not good). I could perhaps cut some length off the inlet side of the carb, usually there is some extra length there that can be removed. I could try a different type of carb, like a "flat slide" that is not as long as a "round slide" type.

Remember, this method is designed to help you find a starting point for proper intake tract length. The exact length would have to be discovered by testing on a dyno, perhaps using different length air funnels or other methods to change the length. Also you must decide at what rpm you want to attempt to produce the most power, 14,000 rpm would not be good for me because the motor does not spend enough time at the rpm, 8,000 rpm would not be good because the motor doesn't spend much time going that slow. So I picked an rpm that serves as a "pivot" point, since the lowest rpm I will probably use is 8,000 and the highest is 14,000 I decided to choose an rpm right in the middle of that "range".  While racing next year I may discover that my rider spends most of her time at 9,000 rpm and above, then I might have to make some changes to tune the intake for around 11,500 rpm.

I will soon be adding a page or two on the modifications I made to install the new 20mm carb, and the modifications I made to get the intake length correct. I have some pictures explaining some of what is discussed above.