Forum » EFI Wiring Fundamentals » Power supply

Power supply

EFI Wiring Fundamentals

Discussion and questions related to the course EFI Wiring Fundamentals

Page 1

Hello I am beginning contruction on my first high quality tefzel harness soon and had a couple questions.

What is the general rule for splitting 12v supply wires for things like injectors and coils? Can I use a 16ga supply wire split into 6 separate 18ga or 20ga feeds for the coils? Or is the norm to use more than one wire to supply the voltage? Ex, one 16ga wire to power two coilpacks?

I know a 5v reference signal wire is very low current , so can all my sensors share the same one or two 5v reference wires? Do I need to worry about stepping the gauge up where they conjoin to one wire?

From my research, there is just very little info on this subject. What are the norms in loom construction for these situations?

I appreciate the help!

G'day Austin. Brilliant questions, and I agree, very little info out there on what is generally done. I've addressed this in our updated wiring fundamentals course that will be released soon.

To help you out right now though, you need to work out how much current your injectors are going to draw, assuming they are being run at 100% duty cycle (this'll never be the case, but the extra safety factor this assumption gives us is worth it). If we look at something common like ID1000 high impedance injectors, we can say they'll draw about 1.5 Amps each. So for six of them, you'll have a max current draw of 9 Amps. You can run an 18AWG wire out to a location near the injectors as it'll handle this current, and then splice it out to six 22awg wires that lead to each individual injector.

Similar story for coils, but their peak current draw tends to be much more, in the realm of 5 to 10 amps, depending on what your setup is, so you'll need to bump up the wire sizes.

For sensors, I splice the supply and ground wires as close to the ECU as possible, and run individual 22AWG wires to each sensor. This way I know they're all sharing a common splice point, and there wont be any loading effects going on. I also like to have a single splice point because they're one of the most common points of failure, and reducing their number is a good thing :-).

This is what I mean, orange is the colour I like to use for sensor 5v, and green for sensor ground:

Attached Files

Hey Zac, thanks for the great info. This has helped out loads! I am looking forward to learning more in the new course!

Just to make sure I'm getting this down though, I need to make sure the wire for each coil can carry the full 10amps (I'm using the oem JZ coils) which would be approximately 18ga. I am still relatively confused on what wire size to use when all 6 of these are conjoined. From the charts online anything rated for 60amps requires 12 gauge and that seems huge! Even the oem harness does not go that big. Am I doing something wrong in my calculation or reasoning? Or do harness builders use say two power feeds, one for 3 coils each?

The total amperage in a series is equal to the sum of the currents in the 6 other wires correct?

Attached Files

G'day Mate.

I haven't got any experience with the JZ coils, but 10A peak sounds like quite a bit for factory coils? Totally plausible though, so we'll run with it :-).

Lets math it out.

If we figure you're going to be revving this JZ powered beast (btw, you should make a post about your project in the build thread section, I love me a good build-thread) out to around 8000rpm, thats 8000/60 = 133.3* revs per second, and 133.3*/2 = 66.6* complete engine cycles per second (4 stroke engine).

This means each cylinder has an ignition event 66.6* times per second, giving us 1/66.6* = 0.015seconds, or 15ms between those ignition events. 4ms is a pretty standard dwell setting, providing your alternator is giving you a good solid 14V (once again, I'm not super familiar with the JZ coils, so this would need to be confirmed). So, at 8000rpm, each ignition coil supply wire (after the splice) is passing current for 4ms out of every 15ms, or ~27% of the time.

Lets make the assumption that our peak coil current is 10A, and it builds linearly from 0A to this over the 4ms dwell period, as the primary winding is an inductive load. We can average this current, and say it's the same as the supply wire passing a constant 5A for the entire 4ms dwell period... Now, since its only passing this current for 27% of the time, we can play the averaging game again, and 27% of 5A is 1.35A.

Where does this get us? Each ignition coil supply wire can be thought of as drawing 1.35A of current continuously, giving us a figure of 1.35A x 6 = 8.1A passing through our main ignition coil supply wire before its splices out to each coil. This is also at 8000rpm! You're not keeping the engine at 8000rpm for minutes on end (if you are, I would like videos please), and as our dwell period will remain at ~4ms, but the period of time between ignition event increases the slower the engine is revving, the situation only gets better.

With all this in mind, I'd run a main 16AWG wire out to the coils, then splice it out to 20AWG wires to each individual coil. That gives you a pretty decent factor of safety, and should keep the wiring nice and compact :-).

Holy crap mate! You answered my question and 3 questions I was about to ask! At first I was only assuming the max 10amps because I figured to overkill everything for safety.

All good broseph, keep us up to date with your project!

Overkill is good, better the wire is too big than too small, but sometimes we really need to do the math to figure out if our overkill is to... over killed :-).

I will definitely create a build thread! I am starting my last two years of college this fall in mechanical engineering with a focus in motorsports. Does your knowledge on this stem from school or mainly experience? This has got me really excited for the next addition to the wiring course!

Both school and experience. I've got a degree in electrical engineering from the university of canterbury, and was heavily involved in formula SAE towards the end of my studies... Plus years of being a filthy bogan building cars for myself and mates :-).

I'm trying to make the course as complete as I can, and have a lot of demonstration material. The first one to be released will be mostly theory based, and the second one will be a complete practical walkthrough, to get the wiring side of our courses more inline with the tuning and engine building :-).

I'm excited for it! If it's anything like the information you just taught me, count me in! You've been super helpful!

I just wanted to confirm something here. Great read Zac, thanks! This example is dealing specifically with Tefzel wire correct? Would i be correct in assuming that with TXL i would be ok with the at least one gauge bigger in this particular example? Or is it still safe to do the math to be absolutely sure? You mention 16ga before the split and 20ga after for each coil, would 14ga before the split and 18ga after for each coil be correct for TXL or not necessarily?

G'day Claudio. You should be right using the same sizes of TXL wire as we've calculated above. While its true that 'Tefzel' (M22759) wire of the same gauge can handle a bit more current that then the equivalent TXL wire, in my experience its not a huge amount difference, and I chalk it down as added safety factor.