Nice to see someone finally doing something about that mess that the factory harness is. I modified mine about a year and a half ago.. I provided an individual ground to each coilpack which also allowed me to nicely heat wrap the whole harness ( topmount problems lol ). Thing is Since then I have switched to the R8 coilpacks so now I will have to go over it againto eliminate a lot of useless extra wires. Thanks to anyone posting it's nice to see and great for ideas.
The longest factory wiring from injector to the first 12-pin junction connector is 32in and they are 0.75 mm2 (~.97722 mm diameter) metric wires according to the factory wiring diagrams. The electrical resistivity of copper is 1.72 x 10-8 Ω at 20c. That calculates out to the wire having a total resistance of 0.0187 Ω for the factory wiring. Doubling the length will increase the total resistance to 0.0374 Ω if you used the exact same wire. Because total resistance is inversely proportional to the area of the conductor, increasing wire size will lessen this effect. Even going to a true 18AWG conductor only adds 0.0170 Ω resistance. Likewise, using 16AWG would only add 0.0107 Ω resistance.
Let's step back and look at the driving circuit for the injectors, MOSFETS (Specifically the FDB14N30, a n-channel high-voltage style MOSFET in my MSD80). The electrical characteristics of this MOSFET has a maximum continuous Drain-Source Forward current of 14A. The piezo-style injectors peak at around 160V and ~8A current. If we take those numbers and put them into ohm's law, we get a resistance of 20 Ω for a single injector circuit. So adding 0.0170 Ω to 20 Ω would be your new total resistance. That is a 0.085% increase in resistance. Let's assume amperage is constant, using ohm's law again, we can calculate that it would take 0.136V extra to maintain 8A.
Considering the range of what the piezo injectors are designed to work within, I do not think there would be any measurable issues with extending the wires.
HOWEVER, with that said, typically the longer the wire in a PWM circuit, the more chance you have for EMI. Hence, the twisted pairs.
Would anyone doing this want to match the twist per inch of the factory harness as well? How is the number of twists per inch calculated?
Crosstalk is when a wire has current flowing through it, generating a magnetic field, which causes unwanted current to be generated in the wire next to it. Standard twist ratios vary wildly, but the factory wire has 49 twists in 31 inches, so roughly a 1.58:1 twist ratio. I would think wire diameter and insulation diameter play a big factor in twist ratio.
Additionally, we could shield the wires which definitely would help with more EMI rejection, but that also requires that the shield material be grounded somewhere so that it can act as a low-pass filter. Personally, I think twisting any/all the wires you can, wrap in some PET or PVC sheathing and have at it.
