Converting a 914 to aftermarket EFI:

Why do this? If you have D-Jet now, the MPS (manifold pressure sensor) is a relatively delicate part that is unobtainable new, and cannot be reliably rebuilt (the diagphram, which is the part that goes, is made of some material no one has successfully reverse engineered). The MPS is the heart of the D-Jet system, so when it goes, you stop. The rest of the system is pretty robust and more or less standard issue. Using an aftermarket ECU, you can replace the stock "brain" and the MPS with modern, solid-state parts that are readily available. You'll also gain tunability in the bargain. If you have L-Jet, you gain tunability with an aftermarket system, though there's no compelling reason to switch if your L-Jet system is working well. If you have carbs now, you'll get more accurate fuel delivery and a system that can be tuned to work even with an otherwise stock engine, which will restore good gas mileage. Carbs with the stock cam usually get no better than 20mpg in town, where an EFI engine can get 30mpg.

Available ECUs: a good number have been tried on the 914 with good results: Megasquirt, the DIY system; Perfect Power; SDS. The system from Link is currently being installed. I'm guessing both Haltec and Electromotive have been tried, but don't know any details. All of these systems are speed-density (manifold pressure/rpm as primary signals) with a throttle sensor for acceleration enrichment and decel cutoff. The SDS and MS units can also run in alpha-N mode (throttle only, no manifold pressure). The MS can also run in MAP only mode (no throttle sensor, MAP used for accel/decel enrich/cutoff).

The SDS system costs about US$1100 for the minimal system (ignition control through the distributor), with a harness and some sensors. This gives programmable fuel and ignition timing, using an MSD or other CDI box to fire the coil. There's a multi-coil waste spark option available for more money.

The Perfect Power system costs about $700 for the PRS2, which is their bottom of the line unit. That price includes a harness. It can do waste-spark using multiple coils.

The Megasquirt unit can be purchased as a kit for about $100, or bought pre-assembled for $250. Full schematics and code are freely available, so if you're particularly handy with electronics, you can build the unit yourself using whatever parts you can get at whatever prices you can manage. $100 seems to be a reasonable minimum, however. It can handle ignition control through an MSD or other CDI setup much like the SDS, plus many other ignition options, though they're all roll your own, at present. The price quoted above does not include a harness or sensors.

By far the easiest way to do the conversion is simply to use stock injection parts for the manifold, throttle bodies, injectors, and plumbing. If you're converting from carbs, it's cheaper to buy a D-Jet setup than it is to modify the carb bits. A complete D-Jet system can be had for about $100, and a pair of Weber IDFs with linkage and manifolds can be sold for $400 on Ebay.

CIS = continuous injection system. This is also known as K-Jet (K = Kontinuous). It's an electro-mechanical system, and was used on a wide variety of cars in the 70s and 80s, from Mercedes to VW, including all "non-special" 911s from '74-'83. The full name was K-Jetronic, even though there were no electronics involved (in most versions).

It uses a mechanical airflow sensor that has a balanced arm holding a plate at one end. *ALL* of the air coming into the engine has to pass through a venturi the plate blocks, and this airflow moves the plate. The arm is also connected to a fuel distributor that has fuel coming in under high pressure (100psi), and is regulated out to the injectors based on a plunger that the airflow arm is attached to. The more the plate moves, the more fuel is allowed to flow to the injectors, which are nothing more than sprung pop-off valves designed to open at 50-60psi.

There are some compensation devices to help cold-starts (very like L-Jet and something like D-Jet), and depending on the exact version of K-Jet used, there are some other devices to enrich the mixture based on engine temp, manifold pressure, and throttle position. These control devices all work by controlling fuel pressure resisting the movement of the plunger in the fuel distributor, thus leaning out the fuel for a given airflow if control pressure is high, and enriching it if control pressure is low. Some very late K-Jet systems used a servo valve to directly manipulate the control pressure under computer control, based on the input from an O2 sensor, so the computer could now control mixture strength.

Acceleration enrichment was controlled by airflow arm balance. When the airflow suddenly increased (like to stomp on the gas), the plate will swing too far from inertia, thus giving an extra bit of fuel, before it swings back. Some very early K-Jet systems had a throttle sensor to reduce control pressure when this happened, but later systems just used a fine balance factor, and just had a throttle closed switch to cut fuel on decel (by raising fuel pressure).

In pure engineering terms, it's a beautifully simple, elegant system.

It's major flaws were that it didn't work well with wild cams, as the intake pulses set up by such cams tend to make the airflow plate bounce. Really violent movement can also make the plate flap around, so fuel delivery over jumps and the like can cause problems, making the system generally unsuitable for off-road use. It's completely incapable of any kind of "sequential" injection, so fuel control at very small throttle openings isn't quite as good as it could be, so eventually it became impossible to meet emissions standards with the system, which is one of many reasons why it was abandoned.

Electricity and power:

Watts = Amps x Volts.

The basic charging voltage is 13.5-14v, so a "12v" system is actually a 13-14v system.

The stock alternator produces a max of 50-55A at 14v, or 700 to 770W. This is, btw, about 1HP (1HP = 746W). Alternators turn physical motion into electricity with about 50% efficiency, so the alternator uses about 2HP from the engine to generate its max power. The alternator only generates as much power as needed, so it doesn't use 2HP all the time.

Some common loads:

Ignition 3A
Injection 3A
Fuel Pump 4A
Head/tail/dash lights 10A
Brake lights 5A
Turn signals 4A (each side, so 8A for hazards)
Blower fans 20A (both)
Stereo 2-5A
Wipers 8A

So, with everything on, that's 44A, about 616W, or about 1.5HP taken from the engine to drive everything. With just the ignition/injection/fuel pump on, it's only 0.4HP. If you're running carbs, the fuel pump is likely only 2A (less pressure = less work), so it's more like 75W total, or 0.1HP.

Electricity and power:

Watts = Amps x Volts.

The basic charging voltage is 13.5-14v, so a "12v" system is actually a 13-14v system.

The stock alternator produces a max of 50-55A at 14v, or 700 to 770W. This is, btw, about 1HP (1HP = 746W). Alternators turn physical motion into electricity with about 50% efficiency, so the alternator uses about 2HP from the engine to generate its max power. The alternator only generates as much power as needed, so it doesn't use 2HP all the time.

Some common loads:

Ignition 3A
Injection 3A
Fuel Pump 4A
Head/tail/dash lights 10A
Brake lights 5A
Turn signals 4A (each side, so 8A for hazards)
Blower fans 20A (both)
Stereo 2-5A
Wipers 8A

So, with everything on, that's 44A, about 616W, or about 1.5HP taken from the engine to drive everything. With just the ignition/injection/fuel pump on, it's only 0.4HP. If you're running carbs, the fuel pump is likely only 2A (less pressure = less work), so it's more like 75W total, or 0.1HP.

If I were F1 dictator:

Reduce pit stops, which really don't improve the racing or the show. Do this by making pit stops very expensive, by simply reducing the number of people allowed to touch the car during a stop to, say, three (not including the driver). Tire changes could be made for safety reasons (puncture), but would be slow, so no one would make them unless they had to. Fewer crewmen to get hit during stops, and fewer to transport to each race. No one complained about the racing or the show in the 1960s, and no one made stops.

Go back to qualifying the way it used to be: one hour on Friday, another hour on Saturday, everyone out at once. Best single time counts. This reduces the chances of weather upsetting things, and provides more to see. Encourage everyone to use all of both hours by giving one point to the car that's turned the best lap time in each 30 min section. So, four points available during qualifying. No restrictions on tires, but the same engine and chassis used during qualifying has to be used during the race (no qually specials that only last for two laps). Blow up the engine or crash, and you start at the back. Increase the points to 15 for a win, 10 for second, 5 for third, then 4, 3, 2, 1. A team acing qualifying but not finishing the race still gets 4th place points. Acing qualifying and winning still isn't 2x second place.

Drop the requirement that a team make their own chassis. Privateers can be very healthy for the sport and the size of the grid, but the current system requires way too much infrastructure to have a team. Good privateers can still win.

Completely remove all current technical regulations and replace them with a set of crash protection regulations (driver cell must experience less than Xg for Nseconds in an impact of a given force, with a set of impacts defined from above, to the sides, front, and rear). No other regulations. Let physics determine the best power to weight v tire life v fuel use v aerodynamics. If the engines make 2000hp and still last the whole qualifying and race, and don't destroy their tires, and the drivers can keep them on the track, what's the problem? Racing at this level should be expensive. If it's too expensive, the problem will solve itself, as teams will simply expire from lack of money. One team dominating from spending more than anyone else won't last forever, and putting in a zillion rules won't stop it (viz. Ferrari over the past 6 seasons), so don't even bother trying. This also ties in well with the previous rule, since a few manufacturers could be more cars, and thus afford to pass even tougher crash requirements.

If I were F1 dictator:

Reduce pit stops, which really don't improve the racing or the show. Do this by making pit stops very expensive, by simply reducing the number of people allowed to touch the car during a stop to, say, three (not including the driver). Tire changes could be made for safety reasons (puncture), but would be slow, so no one would make them unless they had to. Fewer crewmen to get hit during stops, and fewer to transport to each race. No one complained about the racing or the show in the 1960s, and no one made stops.

Go back to qualifying the way it used to be: one hour on Friday, another hour on Saturday, everyone out at once. Best single time counts. This reduces the chances of weather upsetting things, and provides more to see. Encourage everyone to use all of both hours by giving one point to the car that's turned the best lap time in each 30 min section. So, four points available during qualifying. No restrictions on tires, but the same engine and chassis used during qualifying has to be used during the race (no qually specials that only last for two laps). Blow up the engine or crash, and you start at the back. Increase the points to 15 for a win, 10 for second, 5 for third, then 4, 3, 2, 1. A team acing qualifying but not finishing the race still gets 4th place points. Acing qualifying and winning still isn't 2x second place.

Drop the requirement that a team make their own chassis. Privateers can be very healthy for the sport and the size of the grid, but the current system requires way too much infrastructure to have a team. Good privateers can still win.

Completely remove all current technical regulations and replace them with a set of crash protection regulations (driver cell must experience less than Xg for Nseconds in an impact of a given force, with a set of impacts defined from above, to the sides, front, and rear). No other regulations. Let physics determine the best power to weight v tire life v fuel use v aerodynamics. If the engines make 2000hp and still last the whole qualifying and race, and don't destroy their tires, and the drivers can keep them on the track, what's the problem? Racing at this level should be expensive. If it's too expensive, the problem will solve itself, as teams will simply expire from lack of money. One team dominating from spending more than anyone else won't last forever, and putting in a zillion rules won't stop it (viz. Ferrari over the past 6 seasons), so don't even bother trying. This also ties in well with the previous rule, since a few manufacturers could be more cars, and thus afford to pass even tougher crash requirements.