To address a few of the issues raised and a bit more:
N2O ideally should be added down stream of an intercooler (IC), other wise some of the cooling effect from the boiling N2O will be lost in cooling the mass of IC metal (and ultimately the outside air) rather than the intake charge. It will be worse with a water-to-air IC (W2A), where the mass of the water will also be being cooled. If the IC is sandwiched between a */C and the LIM, and you want to use a single point of N2O injection (ie. Before the */C and IC), you will just have to live with this.
Some have used N2O (CO2 is a cheaper substitute with very similar properties) sprayed onto the outside of an IC to pre-cool it.
All the manufacturers of N2O equipment recommend using high octane fuel with N2O. Presumably for good reason.
N2O will never be run lean (well, not for long anyway
), nor even close to stoichiometric. Excess fuel is required because there is less "stuff" that doesn' t
burn in the cylinder; the proportion of Oxygen (which burns) is increased at the expense of Nitrogen (inert), so the temperature would be higher without some additional means of cooling. (The higher temperature is distinct from the amount of energy released, though that is also highter as you are burning more fuel - the whole point of the exercise). Excess fuel over what is required to combine with the extra O2 in the N2O is used to absorb some of that heat energy. This is why fuels containing alcohol are of particular interest for use with N2O; high latent heats of evaporation.
As a result, I believe that the best way to tune N2O is using exhaust gas temperature (EGT). This may or may not produce a tune that releases the maximum power from a given flow rate of N2O but will at least be safe.
Dry systems introduce the fuel through the exi sting fuel injectors. This can be achieved by increasing fuel pressure, done by bleeding some (pressure regulated - not f ul l bottle pressure) N2O from the bottle onto the fuel pressure regulator diaphragm (it can't work where the fuel pressure is regulated at the pump in the tank), or by increasing the injector pulse width electronically (if there is sufficient head room on the injector flows - which may necessitate an injector swap). I think that a combination of the two would be feasible, though I don' t know of anyone who has done so, and changing injector pulse widths allows much finer control of fuel flow than does swapping jets.
The former method is how the dry kits from NOS and ZEX (and others?) work. The NOS systems use a fuel pressure actuated switch and two solenoids for the N2O. The first N2O solenoid supplies N2O to the fuel pressure regulator and the second N2O solenoid, supplying N2O to the jet, is only actuate when the fuel pressure rises to a level sufficient to trip the fuel pressure switch.
The relationship between fuel pressure and injector flow is not linear; double (2X)fuel pressure gives (theoretically by Bernoulli'*) ~1.4X flow. This non-linear relation, and that the regulated N2O pressure being applied to the fuel pressure regulator is designed to work against ~1 atm of manifold pressure (~ the case at WOT), is why dry kits (but not dry systems) are not suitable for boosted applications without modification.
Wet systems supply the additional fuel via (a) jet(*) and solenoid(*) separate to the fuel system the car normally operates on (but typically using the same pump and lines from the tank).
Direct port injection can be either wet or dry but individual N2O jets are fitted into each intake port or manifold runner (as distinct from one centrally located jet before the intake manifold plenum supplying all cylinders).
Because N2O will (or should) only be activated at wide open throttle (WOT) - air is free, use as much of it as you can - the PCM won't try to correct what it would otherwise see as a rich signal from the O2 sensor. This is because the PCM ignores the O2 sensor signal under high load conditions ie. Operates in open loop mode. This is possible because high engine loads aren't reached as part of the government emission testing procedures, so the cat. doesn' t need to work, and excess fuel is provided in the factory fuel map for some more power and to ensure durability (cooling). (At least this is so on every ECM/PCM I have so far seen. I assume that this holds true on US ones) .
One manufacturer (I can't remember who) of N2O systems has tried to use the O2 sensor as a de facto EGT sensor, relying on the sensitivity of O2 sensor voltage to temperature, as distinct from O2 concentration (which is not good away from stoichiometric, esp. with a narrow band sensor), for closed loop N2O mixture control with richer than stoichiometric mixtures. I don't know how effective this could be but doesn' t seem to have much potential for great accuracy to me.
I think that a window switch - combined frequency switches working off an ignition signal, that control the switch on and off rpm of the N2O flow - is an essential part of any N2O system. The former to avoid an excessive torque increase associated with very low rpm N2O actuation, the latter to avoid an injector fuel cut rev limiter with N2O still flowing.
You guys are lucky with the U$50 (10lb?) bottles. Down here it '* ~ A$110 (~U$93, depending on the day) - it'* all imported, as are most of the ki ts and parts.
(the edits were to fix the formatting)