cuppla simple things to check first: make sure the gas cap is venting OK. They can get plugged up and start to starve the engine of fuel. Ditto for the fuel filter and air filter. Engine may be overheating - have you cleaned out the engine fins under it's sheet metal lately?

what about richening up the the carb a bit - what condition is the plug in - if its whitish then you would be running a bit lean.... If that is the case - then you might need to clean the carb.

I have an ongoing similar situation. Kohler 23hp SV820. Runs fine for a few moments, dies, will not restart, splutters a few times if you choke it. I have disconnected from fuel pump, put a "squeeze bottle " on the carb inlet line. engine starts going dead, give the bottle a squeeze, no change. dribble fuel into the inlet air stream, engine picks up a few seconds, then repeat. I have had the fuel solenoid off, it works, & does not release before current is switched off. it does not get hot. I have had carb off engine, removed fuel bowl, float & needle valve.no visible damage or wear. I have blown through the inlet hose past the needle valve seat. I have blown out the jets. The orifice that the inlet solenoid needle seals in is not loose. I have verified spark. what next?

***"If it was burning oil I'd think I'd get a consistent black smoke?"*** No. Gasoline overfuel (or excessively rich) causes black smoke because of the very "complex" way gasoline burns, or "combusts" in the combustion chamber, because of the "complex chemical make up", of gasoline. Gasoline, Diesel "fuel oil", and engine lube oil, are all in the family of chemical compounds we call Hydrocarbons. Hydro for Hydrogen and carbon for Carbon. Hydrocarbon compounds are made by combining "elemental Hydrogen" and "elemental Carbon" together. Because hydrogen and carbon can "link up" in very complex combinations (compounds), these combinations are called "chains". How many atoms of each element, and the varying ways they alternate and hook up determines the "properties" of the final compound. Some of the resulting combinations are very "stable" and DO NOT want to evaporate (or even "melt") easily. Other combinations will "flash" at low temperatures and remain "a thin liquid" at extremely low temperatures. One thing all hydrocarbon compounds share is "combustibility". The ability to combine with oxygen (or "burn"). Some hydrocarbons are eager to combust and only need to get near a "molecule" of oxygen and they just "throw open their arms" to the oxygen (very "promiscuous"). Other compounds are so reluctant to "mess around" with oxygen that they must first be heated to a high temperature before they'll have anything to do with any oxygen molecules (or oxygen atoms). You can draw your own analogies there, but suffice it to say, these compounds are very "chaste". Ok, lets use this in a practical way now. Starting with "how come "burning oil" makes BLUE smoke. The short "answer" is: Because the oil "ain't actually burning". Some portion of the oil (in an engine that is "consuming oil") does actually burn, but the majority of THOSE hydrocarbon molecules are only "vaporized" by the heat of the combustion chamber environs. The hydrocarbon compounds that make good "lubes", are some of those "chaste", or stable ones that don't enjoy "a good combustion". The color of lube oil "vapor" in an exhaust plume is similar to to what you would see if you poured some oil on a hot exhaust manifold. Starting a Diesel engine in frigid temps will result in the following observances in the exhaust plume: First, you will see the fuel oil "vapor plume" as a white or blue/white color, because the initial heat of compression, is only hot enough to vaporize a portion of the Diesel fuel oil, but not hot enough to "ignite" (flashpoint temperature) the fuel. The next color change will be the addition of black as the combustion chamber temperature begins to reach "ignition" flashpoint. But, even though ignition has been achieved, the temperatures in the combustion chamber have not risen enough to support "complete ignition and combustion" of all the available hydrocarbon molecules. This "incomplete combustion" means that some of the hydrocarbon chains are combusting "partially" or incompletely. In this stage, some of the HYDROGEN atoms break out of the compound and unite with oxygen atoms. Some of the CARBON atoms get liberated and quickly "recombine" to form both "new hydrocarbons" and "plain carbon compounds" (soot.....black). It is the resulting carbon and new carbon compounds that give the black color to the plume. This black color only last briefly in this scenario, because the temperatures inside the engine rise rapidly now, and the combustion of the fuel is near enough to complete, that the exhaust plume will become almost clear in a "good engine". But that was in a Diesel engine having combustion chamber compression pressures of more than 500 PSI. Fuel oil can be processed (combusted) very well under those conditions, but engine lube oil would require even a greater magnitude of pressures and temperatures to "burn clean". As a result, a worn out Diesel engine throws out "blue oil smoke" too (though some of that is actually "uncombusted fuel" too, when the engine is in bad condition!). In a gasoline fueled engine, the gas is highly volatile (evaporates readily). Highly volatile compounds are also "highly combustible", and will therefore "burn" in the lower compression environs of a "gasoline engine" (though not without a well timed spark to "incite ignition and propagation"). Gasoline is a complex blend of different "complex hydrocarbons". The more volatile compounds are very eager to burn and create a fast moving flame front, or "propagation wave" that moves through the combustion chamber from the spark plug "spark gap" when the plug "fires". This combustion process is not a "homogeneous" or "simultaneous" process. The highly volatiles burn and consume oxygen first, followed by the "lesser volatile compounds" as the temperatures rise (remember it takes some heat before some of these "chaste" compounds will join in). The least volatile compounds (both of the "original hydrocarbons", the "new hydrocarbons", and the resultant carbon compounds) are faced with a problem. Most of the good looking oxygen atoms have already paired up with the more promiscuous hydrogen atoms and early carbon atoms, so there "are not enough" oxygen atoms around for everybody to find a partner. Not "everybody" will locate an oxygen partner and there will be "some wallflowers" at the end. If you make the fuel/air mix too rich with fuel, there will be even more leftover wallflowers "after the dance". These "wallflowers" end up going out the exhaust in various quantities (depending on specific criteria of fuel, air, spark, temperatures, compression, and a few etceteras). It is the carbon compound "wallflowers" (soot) that give the exhaust its black color when too much fuel is present. Lube oil, (in the same analogy) does not get heated up enough to burn, but does get hot enough to vaporize and so lends the characteristic "blue oil smoke" color to the exhaust. A term that has come up in recent years (in technical speak) is "stoichiometric combustion". That term essentially means that there are "just enough" molecules (atoms) of Oxygen to combine with the Hydrogen and Carbon atoms in the hydrocarbon fuel during combustion so that there are no atoms of any of the "original" elements remaining after combustion stops. As far as one can see into the "near future" stoichiometric combustion is unlikely to be attained in commercially available power plants (engines I mean). CLASS, DISMISSED! PS, check the "choke wizard" per tomplum's suggestion.