In search of better fuel economy, I decided to install a capacitive discharge ignition (CDI) system in Pod.
The system I selected is the Mallory HyFire 6A. I purchased the system from aircooled.net, along with all of the various other parts I needed:
The whole kit came out to roughly $260. The tach adapter is necessary to trigger the L-Jetronic fuel injection properly. The non-resistor rotor is pretty important, as I came to find out; not only will the CDI burn out a resistor rotor, but the resistor limits current to the plugs to such an extent that poor ignition occurs when used with the CDI (which is high voltage, but low current). The Aurora wires are there primarily to suppress noise that might otherwise occur with the non-resistor rotor. The coil is there because my original coil is old and worn out. The plugs are there because the whole point of CDI is to open up your plug gap.
The only thing I was disappointed in was the Aurora wires. They are nice wires, but even though I ordered Type 4 engine wires from aircooled.net, they lacked the bakelite plug connector extensions necessary to reach the plug and allow the seal to do its job on the cooling tin. aircooled.net does say this is the case (to use your old ones; I picked up a spare set at the junkyard), but it is still a pain to modify the wires:
I took the opportunity to measure the resistance of the old and new wires:
OLD WIRE NEW WIRE DIFFERENCE
#1 970 ohm 1670 ohm +700 ohm
#2 1460 ohm 2220 ohm +760 ohm
#3 950 ohm 1650 ohm +700 ohm
#4 1550 ohm 2050 ohm +500 ohm
coil 7720 ohm 390 ohm -7330 ohm
rotor 4880 ohm 0 ohm -4480 ohm
I was a little surprised that the Aurora wires have more resistance than Bosch OEM wires. Except for the coil wire, of course; in that case, I was surprised how much resistance the Bosch OEM wire had. Note that the bakelite extensions add resistance (they are rated and labeled with that rating), and these measurements are taken with the extensions on both sets of wires.
I decided to reroute all of the ignition-related wiring to a set of terminal blocks located on the firewall. This simplifies going back to stock if I ever care to do so, and also actually simplified the CDI wiring process overall (can't connect stuff to coil + and coil - anymore when using a CDI).
The following is a diagram of the two terminal blocks I used to make the ignition system connections. In this diagram, vertical lines indicate shunted terminals, and horizontal lines indicate jumper wires. This diagram shows only the permanent jumpers. The terminal blocks are mounted on the upper-most edge of the firewall, above and to the right of the decel valve.
1---2 3---4 9--10 11 12 | | | | | | | | 5 6---7 8 13 14 15 16
The blocks are wired up in the following groups:
The following are permanent connections made to the terminal blocks:
3 - tach
5 - trigger output of Pertronix [black wire]
6 - FI trigger
7 - to Test Network terminal 13
9 - power for Pertronix [red wire]
10 - power for reverse lights / double-relay
13 - +12v from ignition switch
15 - to coil +
16 - to coil -
For the stock ignition system, we want the coil to sense the trigger and get power from switched +12v, so we make the following connections:
Jumper 8 -> 12
Jumper 14 -> 11
For the HyFire, the HyFire needs to sense the trigger, we must provide switched +12v to the HyFire, and the HyFire must control the coil. Thus we remove the stock-configuration jumpers and connect the following HyFire wires:
1 - white (points input + tach adapter lead)
14 - thin red (+12v switched input + tach adapter lead)
11 - orange (to coil +)
12 - thin black (to coil -)
NOTE: The yellow "to tach" wire of the HyFire is unused!
One other thing to note is that the Mallory tach adapter gets VERY hot. This is apparently normal. It did get hot enough to soften the plastic core the wiring is wound around, such that pressure from the top engine hatch warped it a little. Oh well, as long as it doesn't stop working, I don't particularly care what it looks like.
After the system was installed and run for a week or so, I installed the new plugs, gapped to .045" (instead of the stock .028"). So far, results seem good! The engine idles very smooth, and throttle response definitely seems improved. Within the next couple of weeks, I am hoping to borrow Richard Atwell's wide-band O2 sensor to tune my AFM, and really make a dent in my gas bill!
You can find photos of the project here.