Forum » Professional Motorsport Wiring Harness Construction » Determining current draw for ignition coil and selecting wire gauge

Determining current draw for ignition coil and selecting wire gauge

Professional Motorsport Wiring Harness Construction

Discussion and questions related to the course Professional Motorsport Wiring Harness Construction


Page 1
Author
744 Views

How was the current draw determined for the ignition coil power supply and ground circuits in the attached circuit sheets? I'd like to fabricate a motorsport grade engine harness for my B5 platform Audi A4 with a 1.8T big turbo engine. I'm not sure my ignition coils will draw a current similar to the FD35 RX-7 used as an example in the Circuit Design module. The stock engine wire harness for my vehicle uses 12 AWG wire for ignition coil power supply and ground. The ignition coil power supply circuit is protected with a 15A fuse from the factory. I'll be using VAG coils P/N: 06H 905 115B and running upwards of 30 psi of boost. Is there a general rule for determining ignition coil power supply current draw or is this more a function that is specific to each type of coil and engine design?

To measure this properly, you're going to need a 5Watt 0.1Ohm resistor, and an oscilloscope. Put the resistor in line with the power wire going to a single coil, and with your scope, measure the voltage drop across the resistor. Knowing that I = V/R, your current is then your measured voltage divided by the resistor value, 0.1Ohm, which is the same as multiplying your measured voltage by 10.

What you will see is that the current passing through the coil will rise linearly over the dwell period of the coil, then drop back to zero sharply, in a repeating fashion. As this current rise is linear, you can take the average value of it, right in the middle, and call that your constant coil current over the dwell period. This current will only be passing through the coil periodically though, and the rest of the time the wire is sitting there not doing anything, except radiating away any heat that might have built up. Working from your dwell time and max engine RPM you can determine the maximum duty cycle of coil current seen on the wire, and this gives you your average current along that single wire, and you can size from here.

In reality, I probably wouldn't do the math, but would run 20AWG M22759/32 wire for the coil powers and call it good.

The OEM fuse will be protecting all four coils, so the 15A rating makes sense there. The 12AWG wire probably reduces in size when it splices out to each individual coil. You might need to do something similar, but running 16AWG wire from your power supply point to the splice, and 20AWG out to each individual coil after that should be more than sufficient.

Thanks for that info Zac! Great work on everything presented in the learning modules!

Isn't it a bad idea to have 20awg wire protected with a 15a fuse?

I've attached amperage graph for 115 coils (source: https://www.sxoc.com/vbb/showthread.php?584905-A-quick-how-to-use-VAG-coil-on-SR20)

It looks like that 115 Coil will draw around 16 amps with 3 ms dwell time.

Blue is coil current, red is ecu signal voltage, green is coil voltage and brown is coil earth.

Attached Files

And you can see it only needed a 2ms dwell (and 10 amps) at that voltage. See where the slope of the current starts to rise -- that is becuase the coil is fully saturated, and extra energy is just going into making heat.

So, let's do the math on 3ms/16 amps. The average current when charging is 1/2 of the 16 amps, or 8 amps. Let's say you ran this coil on a 12,000 RPM 4-cycle engine (ie, so the coil fires 6000 times per minute, or 100 times per second) -- that is 10 ms per firing. So 3ms/10ms = 30% duty cycle of charging.

30% of the 8A average, is 2.4A, well within the capability of a 20 AWG wire. Multiply this times the number of coils to figure the current requirement for a feed wire that is spliced to to all the coils.

If you don't run 12,000 RPM, then this number would be proportionally smaller. If you only charged the coils at 2ms/10A, then the number would be 33% less.