Technical Articles

Mikuni and General Carburetor Tuning

by Kevin Cameron

Experienced tuners employ many ‘vital signs’ to read an engine’s operating state. Listening to its sounds, hearing it accelerate, dabbing a finger inside the exhaust pipe, and examining evidence on its spark plugs and other parts all help to decide questions like these:

  • is the mixture correct, or is it rich or lean?
  • is the spark timing early, late or correct?
  • is compression ratio reasonable?
  • is fuel antiknock quality adequate?

A good starting point for any engine is the original tuning spec. published by the manufacturer. Very often, carburetor needles and needle jets are worn to the point that they supply an over-rich mixture, so it’s well to begin tuning with new brass. A particular sore point is the fit of the throttle slides; a poor idle is often traced to a worn slide-to-carb body fit, with the diagnosis made in the time-honored way by simply pushing in on the slide with your finger as the engine idles. If the idle improves, it may be time for fresh parts.

Often there is no published starting point for tuning as when Mikuni round-slide VM carbs are applied to engines that predate them. Because it’s stupid to re-invent the wheel, you’ll ask someone with a similar conversion for a beginning jetting combination, but the final tuning will be your own responsibility.

The first point is adequate fuel supply. Tank petcocks designed for 25 horsepower engine may become the fuel delivery chokepoint is asked to supply a 50 horsepower engine. The typical symptom of insufficient fuel delivery is that there’ll be good performance at all times except down a long straight; after running on the straight long enough to pull the fuel level down in the bowls, the engine will mysteriously cut off. It will then coast down as the bowls refill, then refire and run fine until the next long straight.

Four-stroke engines typically use 0.5 pound or so of fuel per horsepower, per hour, on steady full throttle. That means 100 horsepower engine will need a flow capacity of 50 pounds per hour. At 6 lb/gallon, that’s 8.3 gallons per hour, or at 32 ounces per gallon, about five ounces a minute. Adding 50% safety factor gives us a need for fuel delivery of 7.5 ounces (240cc) per minute, per 10 horsepower (ed. Kevin is using ounces, a measure of mass Pound – ounces, not fluid ounces which is a measure of volume – gallon-ounces. 10 pounds of water equals 1 Imperial gallon. Gasoline weighs 6.7 pounds per Imperial gallon (6.0 pounds U.S. gallon)). The right way to measure this is as follows; with the bike on the stand, empty the fuel tank and remove the carburetor bowl drain plug. Put a catch bottle under the bowl drain to catch the fuel that will flow during the test. With the tank petcock turned off, put 1/2 gallon of fuel into the tank .With stopwatch in hand, start the watch and open the petcock. Let the fuel run for a minute, and measure what has flowed. This method is good one because it includes all resistance’s in the system; petcock, lines, filters (if any), and carb float valve. Fuel starvation is, obviously, most likely when there’s almost no fuel left in the tank.

The one possibility that this method does not test for is an inadequate gas tank breather line. Air should flow to the tank from the free end of the line when you blow into it.

Mikuni float valves have their sizes stamped on them (older riders have magnifying glass at the ready), and you should be aware that there is such a thing as tiny float valves sized for fuel pump applications. For most purposes, a 3.0 or 3.3 float valve is big enough for any gravity-flow application, but some carbs set up for snowmobile applications come through with dinky 1.5 valve appropriate for use with a pump. Replace any such valve with a 3.0 or bigger.

Be aware that paper filters have been known to “prolapse” and block flow. Trust nothing to luck. Inspect everything. Filters, indeed, are good insurance against tiny flakes of tank rust, etc., that can easily stop up a main jet.

Make sure that carb or float bowl mounting is flexible enough to prevent engine vibration from inducing frothing and non-closure of the float valve. The standard Mikuni VM34-200 rubber mount is nice and flexible, but the much thicker snowmobile mounts are too stiff. Evaluate home-made rubber mounts made from hose and hose clamps for flexibility, being sure that the carb or bowl cannot bang against nearby chassis or tank parts.

Float level is the next trap. On most Mikuni stuff, setting the float arms to be parallel to the bowl gasket surface when the carb is inverted is a good starting point. If the level is too high, a tilted-towards-the-engine carb position, plus vibration, will make fuel dribble from the idle holes at the lower, engine-side edge of the slide, making the mixture rich and irregular. If it is too low, when the engine is at full throttle the float may not be able to drop enough to open the float valve sufficient to deliver the necessary fuel.

The next point to set is idle mixture. Fit street-type spark plugs of hot enough heat range to tolerate long idle running without sooting up. Be sure to synchronize multiple carburetors, noting that it’s more important that they lift together than they reach full throttle together.

Assuming that the engine starts and runs, warm it up and employ some means of holding the slides at a constant idle position. This may be throttle stop screws in the case of street-bike carbs, or a steady-handed helper in the case of carbs without stops. With the engine idling, take a small screwdriver and try other idle air screw positions, noting with each test the rpm raise or fall. Once you have found the position of fastest idle, note the airscrew position in turns from all-the-way-screwed-in. On Mikuni carbs, the idle screw controls the flow of air, bled into the idle fuel flow. Screwing the screws in enrichens idle, and vice versa. If the position of best idle is less than 1/2 turn open, the idle or pilot jet is too small. retest with a larger size. If the best position is as much as three turns open, the jet is to big.

Now follows roll-on testing, conducted with the machine standing still, engine running. The first test concerns slide cut-away. With the engine idling stable, slowly roll the throttle on to see if the engine picks up cleanly. If it does, well and good, proceed to the next test. If, however, you turn the throttle and get sluggish pick-up, or misfiring, or just have a sense that it’s taking a lot of throttle movement to get little engine response, retest with a richer (smaller cut-away number) or leaner slide until the engine does not pick up smoothly with the roll-on. Throttle slides are quite expensive now, so if you don’t have an assortment of them, you can use the “gas shut-off method” instead. With the engine running, turn off the fuel petcocks. If, as the engine gradually drains the fuel bowls, the roll-on response improves, you know the slide was too rich. Engines are often very sensitive to slide cut-away, down to 1/2 a cut-away number. If money is important, you can make a slide leaner by filling the cut-away (the raised entrance side of the slide) higher. If you try to make it richer, remember that this drops the needle by the same amount. Mark any modified slide so it doesn’t embarrass you later.

The needle jet is next. During the first ten millimeters of lift, most Mikuni needles are cylindrical – not tapered – and this means that mixture control in this part of the slide lift is performed only by the needle jet. If, as the slide rises into this region of lift the engine begins to run poorly, try a needle jet that is bigger or smaller, and follow the trend of improvement when you find it. Remember that this is still low throttle, and the gap between needle and jet is very sensitive to tiny size changes.

Beyond ten millimeters of slide lift, the needle taper takes over, and here the needle’s clip position controls mixture. Lowering the needle clip a notch enriches the mixture and vice versa – and this is a sensitive adjustment; even one clip position can make a big difference. As you will know if you’ve looked in the Mikuni book at the available needles, there are single-taper, double-, and triple-taper designs. The tapers start at various places and diameters. Is there any way to make sense of all this, any way to extract truth from the little numbers stamped on the needle shank? Alas, no. A 6DH4 on the shank tells us the needle is approximately 60mm long, and has two tapers (D and H) on it. The number 4 tells a Mikuni engineer which drawing to look at, but it does not conveniently encode such variables as the needle’s shank diameter or where the tapers begin or end. Mikuni has graphs showing needle area as a function of length; the best you can do is to overlay these graphs and try to see if there exists a needle that differs from what you are now running in the right way, in the right area. Pluck up your courage.

The above process will get a carb set-up out of left field and into rideable condition. Once you take to the track or highway, the settings achieved by the above method may turn out to be on the lean side. This because engines are always leaner when the throttle is moving than when it is sitting still.

Be an adult and keep a notebook with what works written down in it. Otherwise you’ll find yourself re-inventing your own personal wheels.

Most manuals on carburetion act as though each carb system had a fixed, non-overlapping range of throttle position. It isn’t so. The ranges overlap a lot… particularly slide cut-away and needle-jet size. Don’t look for an exact solution, look for improved running.

Any interested person can get good results using this rustic method, taught to me by ex-Kawasaki racer Hurley Wilvert, who in turn learned it from a grizzled Australian practical mechanic. You don’t need dynos, CO meters, and engineering degrees to make your bike go.