At this website, https://moretraction.com/2018/12/31/2416/, the claim is to run separate power and ground cables for a clean bus and a dirty bus to minimise voltage spikes and drops. I just about to tackle my first full harness and have not really seen this in any of the cars I have wielded spanners on. Comments?
This is common best practice. Not to say all OEMs do it, but dig through the wiring diagrams and you will find that most do.
OEM mostly yes. It complicates battery isolation a bit
The easier version is to just make sure you don't use the same grounding point for sensor grounds and power grounds.
The site referenced above says that because of what the box has to do, which is limit wheel speed or spin on a very marginal drag strip track surface with stooopid power, like 3500hp stooopid. You can't have variance or noise in the system or the box doesn't work worth a E36 M3.
In my previous years of car stereo work, I always supplied power and grounds together and my rcas were completely separate. I even went so for as to wrap the ground around the power feed to minimize the noise potential.
Isolating the electronics from devices that turn on/off with high current loads is definitely something OEMs do, but it's a bit more complicated than the picture above shows. In the cars I've looked at you won't see two sets of power leads coming off -- rather, you see one nice big wire going to a distribution point (fusebox), with it branching out further from there, and the different types of devices being on different branches. This means that the isolation is worked into the selection of which devices share a fuse.
It's also more subtle because even within the "digital electronics" realm, you find some devices that are noisy and can pollute others. As another OEM example, on my Miata the ECU has three connectors to it, and if you look at the pinouts on the wiring diagram it becomes apparent that the signals are not just randomly assigned to pins. All of the delicate analog sensor inputs (MAF, oxygen sensor, cam/crank signal, etc) are on one connector, with all of the "high" current outputs (which may 1 amp or less, things like idle air control, or the control side of the fan or fuel pump relays) on another.
Another example, this one aftermarket -- the Innovate LC-2 wideband sensor that I have in my Miata uses a common ground for both analog signal reference and for the heater power circuit. Unfortunately the square wave power going into the heater means that the analog wideband output signal is noisy, if you hook it up to an oscilloscope you can watch the wideband signal change by much as 0.5 (so from 12:1 to 12.5:1 and back, for example) in time with the heater turning on and off. (To fix this I had to switch to using a digital signal output from the LC-2 and feeding it into the Megasquirt over CAN).
Back in the Link ECU days, we used to actually add a signal ground to the early Miata ECU wiring to clean up the information coming from the various sensors. Mazda is quite clear about signal vs power grounds in the wiring diagram.
In reply to codrus :
A lot of that is also how the circuit board is laid out - high current transistors tend to be in one general area, makes current management and heat management easier.
I don't think much in the way of digital hardware will work if it's not isolated from solenoids and motors and whatnot somehow. Trying to make digital equipment coexist on a bus with that stuff would be like trying to practice the piano at an indoor gun range.
This topic is so complicated that it could be debated daily in the "Guild of EFI Tuners" fb group daily, forever, without coming to any consensus. For best practice, Bosch makes a wiring book (Blue Bible) that covers the stupids.. that German car engineers constantly don't do.
You'll need to log in to post.
