I experimented with N2O twenty years ago, but not much has changed except the controls & electronics. The wet plate (fogger) systems were prefered for their simplicity. A N2O solenoid and a fuel solenoid, and a metering device (jet) at the plate entry with distribution tubes in the plate opening. The only hardware change these days is additional sets of matched soleniods for each stage.
N2O is only progressive when fed in stages, and only beyond WOT power. If your normal engine WOT output is 250hp then a 25/50 N2O system will deliver and increase in power from 250 to 275 (25hp) and then 300hp (50hp). There is no gradual throttling between each step.
I used a brake pedal switch as a lock-out, and a WOT throttle switch to enable the system, with a separate enabling switch. I did not have the rpm based electronics available today to prevent low rpm triggering, I used a loose torque converter (3500 stall) to assure the rpm was well into the powerband. I would stage with the brakes on, enable the system (after the burnout tire cleanup ritual), and mash the throttle.
Adding an rpm based switch (MSD or other) to accomplish the same function is a better method.
My primary ride where I played with the N2O was a 70' GTO with a moderately built Pontiac 400. It would run mid-13 second ETs at 100 to 105 mph, without the N2O. Not too bad for a 4000# car with a full interior and street tires (~375hp). The N2O would drop the ET a second and a half (easy low 12's) with a fresh bottle. The system started as a Downey Service Center fogger system (a plate system advertised at 100hp), a design copied from the early NOS systems. I ran a Q-jet for a while, and then various square and spread bore 650/780/850 Holley vacuum secondary carbs, and finalized with a modified 700 cfm double pump Holley (modified to flow about 735 cfm with no choke horn and polishing). The vacuum secondary carbs were not safe (IMO) as the secondary transition to WOT was neither consistant or reliable once the N2O was triggered.
Things that helped power and consistancy:
Locate the solenoids as close to the plate and pill (jet) as possible. This prevents the fuel and N2O from premature phase change (liquid to vapor) before it is injected into the manifold. Vapor in the fuel lines is bad, fuel vapor does not have the density of liquid fuel, presenting a lean condition until the vapor clears (a boom if the line needing clearing is long).
Protect the solenoids and fuel/N2O lines from heat (for the same reason, to prevent vapor in the lines). I built an oversize aluminum heat reflector plate that bolted between the manifold and the N2O spacer. The solenoids and brackets screwed to this plate.
Use a return line fuel system to keep the fuel in liquid form. The pre 1996 XJ MPI uses a return line fuel system, so this should not prove too difficult. I ran separate fuel pumps for the N2O after too many fuel pressure related failures (from vapor in the lines). I regulated the fuel pressure for both the carb needle valves and N2O solenoid.
Use a cool can for the fuel line (a can with a spiral fuel line inside to precool the fuel before feeding the solenoid). I noted power and consistancy gains with chilled fuel. The cool can was downstream of the fuel regulator (to allow it's volume to work as a pressure tank).
Own a couple of N2O bottles, and fill them all at the same time (at the same pressure and temperature). I owned a range of cylinders: 3#, 5#, 10# & 20#, in both siphon and regular bottles (with mounts for both bottle styles). I owned a 120# cylinder to fill the smaller cylinders. Race day preparation including filling four 5# cylinders off a hose manifiold from the larger cylinder. I could only fit four 5# cylinders in my ice machine box to cool them while filling (it helped with the fill).
A 5# cylinder would produce about 20-30 seconds of consistant power (four runs max) before the N2O pressure would drop off. The 10# cylinders would produce about six or seven runs before the same drop in power. The 20# cylinders would produce consistant power for about twelve runs. For maximum power it was better to swap-in a fresh cylinder than trust an aging bottle.
I did not use the bottle warmers common today, or any electronics other than simple switches. While my goal was consistant power for bracket racing I was able to drop the maximum enhancement from the N2O three tenth's of a second using the above small modifications (the improvement over the original bolt-on performance was from ~80hp to ~120hp). The most effective aid to consistancy was multiple fresh bottles.
The main drawback to the N2O was WOT-only power, and the limited number of runs between fresh bottles (even with a 20# bottle). You could not use the N2O playing around on the street, not for burnouts or tail-out power slide turns (no throttling capability). You needed to be on good traction with a straight run to make good use of the enhancement. The on/off power delivery from the N2O was not easy to compensate for, and any lack of traction put the engine deep past the redline rpm. A slip out of the tire traction grooves at the drag strip (on one run) zinged the engine past 8000 rpm (in second gear at over 60mph), with two broken valve springs being the result. Not exactly a good attribute for a daily driver.
Power for passing on the road was also an easy way to use up a bottle (pass one line of cars and you quickly catch the next line, needing more power, just to catch the next group, etc.). The power was easy to use, but it got to be a chore to charge and swap bottles all the time.
If you have the need for power, and can afford the expense, supercharging is as effective and less trouble (so is buying
a big-block buggy or racer).
