This lean condition at part throttle will cause the engine to miss or stumble, due to the lean air/fuel mixture. This problem is common on many carbs (a notable exception is the Demon line of carburetors, which have used the PerformanceGas to do exhaust gas analysis along with dyno, track, and drivability testing to establish the factory fuel curve that allows for a stumble-free guarantee). If the air/fuel mixture is too rich at idle and part throttle, the idle jet/restriction may be too big and may need to be replaced with a smaller one.

The next step is a track test using a portable infrared exhaust gas analyzer to check the cruise speed air/fuel mixture main jetting, followed by the power air/fuel mixture under load. During a road/track test, you are able to read and correct the jetting in order to have the proper mixtures at idle, cruise/light throttle, and power/wide-open throttle.

A carburetor has an accelerator pump, idle, main jets, and, in most cases, a power system designed to supply the correct air/fuel mixture for the engine's demands. An idle system has an idle jet/restriction that must be changed to supply the desired fuel mixture for idle and off-idle engine demands. For a carburetor that uses a power valve, the main jet size determines what air/fuel mixture is delivered to the engine at light load/cruise speeds (1,500 rpm and up). The power valve restriction (under the power valve) is the determining factor in the air/fuel mixture that the carburetor will supply when the power valve is open; a 6.5-inch power valve will be open and supplying the richer air/fuel mixture needed under high power demands anytime the vacuum is below its 6.5-inch opening point. A carburetor that uses metering rods in the primary jets (such as a Quadrajet) will use the metering rods to change the air-to-fuel ratio for the power and cruise mixture demands of the engine. Remember, the larger the metering rod diameter, the leaner the air/fuel mixture will be.

The accelerator pump adds fuel as the throttle valves are opened. Tuning the accelerator pump squirter volume and duration is mainly trial and error. For a Demon- or Holley-style carburetor, the combination used most often is a 0.031-inch squirter, along with a pink pump cam. We often shim the accelerator pump duration spring to make the pump more active. Many race-designed carburetors have changeable idle and high-speed air bleeds, but I will leave their affect on the fuel curve to be discussed at another time.

Readings provided by exhaust analyzers

* CO (carbon monoxide): We use the reading from an infrared gas analyzer to determine the air/fuel ratio. (Note: CO is partially burned fuel.)
* HC (hydrocarbons): The amount of unburned fuel or an indicator of an engine misfire. The best mixture gives you the lowest HC.
* CO2 (carbon dioxide): The product of complete combustion. The best mixture gives you the highest CO2 reading.
* O2 (oxygen): A high O2 reading indicates a lean mixture or an exhaust leak. (Note: if O2 is above 2-3 percent, CO readings will not be accurate.

The best power and cruise air/fuel mixtures (CO) will burn all the O2 in the cylinder and create the lowest HC reading (misfire). They will also cause the CO2 reading to be its highest.

A starting point for air/fuel mixtures for most race engines is:

Idle: 1-3 percent CO or a 14.1-13.4:1 air/fuel mixture
Cruise rpm: 1-3 percent CO or a 14.2-14.0:1 air/fuel mixture
Power mixture and acceleration: 6.6 percent CO or a 12.0:1 air/fuel mixture for a "normal" engine; a high performance engine with improved combustion chamber design such as a Winston Cup engine, in some cases, may use a slightly leaner power mixture of 4 percent CO or a 13.0:1 air/fuel ratio.

The Lambda meter method uses an extended range oxygen sensor to determine the fuel mixture by analyzing the unburned combustibles in the exhaust gas. An extended range oxygen sensor can read air/fuel mixtures as rich as 10 to 1. On the lean side, it can read air/fuel mixtures of 19 to 1 or leaner (a standard oxygen sensor is only accurate at air/fuel mixtures of around 14.7 to 1). This method has the advantage of extremely fast reaction times for the readings, but it can be less than accurate on an engine with a race cam at other than high load conditions because of the excessive oxygen in the exhaust created by the cam overlap. The use of an infrared exhaust gas analyzer, while slower in reaction time, has the advantages of reading the oxygen/unburned combustibles content of the exhaust and determining the air/fuel mixture by reading CO. The engine's rate of misfire can be determined by reading the HC, and the engine's efficiency can be determined by the CO2 reading.