Okay, so now we are getting somewhere and we know what you want to do. I know diddly about modifying or tuning motorcycle engines, but it seems that most "improvements" fall into one or both of these categories: improve the aspiration of the fuel-air mixture during the intake cycle or improve the removal of combustion products during the exhaust cycle. On automobile hot rods both of these methods are well understood and work well if exhaust emission controls are not a consideration. For example, "she's ported and relieved and stroked and bored" refers to mechanically smoothing and polishing the path taken by the fuel-air mixture into the intake valves, and increasing the size of the piston cylinders and the the displacement of the crankshaft to increase the volume of air-fuel mixture taken in during the intake cycle. Both approaches will significantly increase the performance of an internal combustion engine.
Going further, one could add an engine-driven compressor (supercharger or turbocharger) to increase the effective volume of air admitted to the cylinders during the intake cycle, modify the exhaust path to reduce back pressure during the exhaust cycle, and introduce nitrous oxide into the cylinders during the intake cycle to increase and accelerate fuel combustion during the power cycle. The latter is an especially simple and effective way to boost motorcycle crankshaft power.
The downside to all of the common "improvements" is the defeat of mandated exhaust emission controls. Which is okay for off-road racing and closed-track racing because the "bad" emissions do not occur often, or in large enough quantities, to significantly affect the local environment. Imagine an Indianapolis 500 race occurring every day though: that's
Los Angeles traffic on a bad smog day without the advantage of a high-speed commute.
I could be wrong about this, but it appears that the engine oxygen sensors are designed to produce their median output when a stoichiometric mixture has been achieved. If an engine runs too lean, the sensor would "bottom out" while the ECU module would continuously call for a richer mixture, either by increasing fuel flow or restricting air flow or perhaps both. In any case, if the correction is effective, the air-fuel mixture will become richer in fuel and the sensor output will then increase and "peak out" at some level of fuel richness, overshooting the desired stoichiometric fuel-air mixture ratio. The ECU module would then call for a leaner mixture, either by decreasing fuel flow or increasing air flow or perhaps both. If the correction is effective, the air-fuel mixture will become leaner in fuel and the sensor output will then decrease and "bottom out" at an overly lean air-fuel mixture ratio.
What I fail to understand is how the sensor output can be "bottomed out" on the lean side if the ECU module is doing its job. If that does indeed happen, there can be no oscillation in the oxygen sensor output. It will remain stuck in a limit cycle, continuously calling on the ECU to increase the richness of the fuel-air mixture, but with the ECU apparently unable to comply. I don't see how changing the sensor output is going to do anything to correct this.
Since you have access to a dyno, it might pay to temporarily disconnect the oxygen sensor from the ECU module and substitute a DC signal derived from a low-voltage power supply. Adjust this substitute DC signal to a level that avoids the "check engine" condition. Using an exhaust gas analyzer (if one is available) tune the engine for a stoichiometric air-fuel mixture at various levels of engine loading. Measure the output of the oxygen sensor as you do this. Vary the air-fuel ratio from lean to rich and observe how the oxygen sensor output varies. The oxygen sensor output
should produce an output voltage somewhere between its minimum and maximum output voltage when a stoichiometric air-fuel ratio is achieved. If this does not occur, or if you desire to run the engine rich, determine what setting of the substitute DC signal will allow proper engine operation to occur. It is likely that the oxygen sensor will "top out" before reaching that level, in which case you simply cannot use the stock oxygen sensor at all, even if you offset and scale its output. Perhaps the "wideband" oxygen sensor you mentioned will allow operation at a richer mixture than the "stock" setting the ECU expects. If that is so, then we can advise you on how to build a simple, low source-impedance, device that will offset and scale its output to place its output in the "stock" range, even though your engine is running richer than a stoichiometric mixture ratio.
It is possible (programmers are sneaky) that the programming of the ECU module
requires the sensor output to limit-cycle between the rich and lean extremes, producing the roughly sinusoidal waveform, as the ECU attempts to control the air-fuel ratio. The only reason I can think of to program it this way is to verify that the oxygen sensor is actually working: if oscillations are observed it must be working; otherwise turn on "check engine" light and go into "limp home" mode. If that is the case, then you can "fake out" the oxygen sensor by providing just such an oscillation to the ECU and disconnecting whatever the ECU is controlling that varies the fuel-air mixture ratio. You probably want to do this anyway if tuning the engine for performance. Let the ECU handle the "tough stuff" like rpm-sensitive ignition timing and spark advance while the rider controls the fuel flow with the throttle.
An after-market thingy that the oxygen sensor plugs into, that has adjustable "offset" and "gain" controls, could be a viable product. If it really does need low-valued variable resistors to be practical, these can be manufactured to spec. Thick-film potentiometers, or variable (rheostat) resistors, deposited on ceramic substrates would be inexpensive to make in values ranging from about one ohm to ten ohms or so.
