Recently, I decided to upgrade the carburation system on the 460 residing in my 1982 E350 motorhome, by installing a Fitech EFi system. I had no end of problems chasing down and eliminating EMI issues and sources. I the course of this there were several suggestions from Fitech Technical Support to check the grounding of the Fitech unit as well as the engine in particular. With my trusty DVM, I proceeded to measure resistances everywhere. I ran as many single point star grounds as possible and ran 8AWG cable everywhere. That is from battery ground to alternator case, engine block, ignition module, car mounting bolt, gauge common wire, body and frame. In all cases of equipment that had ground terminals, these were linked to the 8AWG cable. At the end, all resistances indicated zero ohms.

But then, I fired up the engine and proceeded to look at voltage differences from the ground terminal of the battery to the various points I had connected together. To my surprise, in nearly every point I measured I found a potential of from a few to more than 100 millivolts! Still heavier wire made no significant difference. Heavy cleaning of every contact point made no difference - and I had already established that there was zero resistance within meter resolution. I can only conclude that there are substantial currents flowing in the various ground wires, but what is the impact of that on EMI or the general function of various electronic components?

During these tests, the alternator was delivering about 11 amps, down from a startup value of nearly 35. after full warm up, the electric fans would kick on for an alternator response of as much as 55 amps.

Any comments or better understanding of this observation and whether it constitutes a problem would be greatly appreciated.

 

The rule of thumb for grounding sensitive equipment is ground it at one point only. You are correct, there are minor differences in ground points, and there can be currents flowing through these ground wires. If you happen to ground a sensor or some remotely mounted part of the system at the sensor itself, and then you have another grounding point at the main piece of equipment, you can get a current flowing through the ground wires of the sensor. One common rule of thumb are shielded wires. The shield is tied to ground at one end only. If you tie the shield of the wire at both ends, then you can get current flowing through the shield, and this can cause noise in the signal wires.

 

Franklin2, I have a full slate of Speedhut gauges and all bring back sensor ground to the gauge and all the gauges are grounded to a single point. The stereo system, head unit, amplifier and subwoofer are currently (sorry) grounded to different points because of separation distance.The Crane ignition and Duraspark II module are grounded to the same point on the body.The alternator case is bolted to the engine and the battery has separate ground cables to the engine, frame and body. Yet the Fitech suggestion was to add additional 8AWG cables from the battery ground to the alternator case, the intake mounting flange for the FiTech unit and the distributor. Its had to understand how competitive a moderate length of 8AWG wire can be to the metal mass of the engine, yet the voltages were there. Ground loops are scary things to resolve.

I was aware of the importance of a single ground contact for shielding wire, but didn't realize it came down to small currents again!

 

The problem only usually lies with the sensitive circuits. For instance on a stereo system that has a remote amplifier and uses shielded RCA cables. the stereo will have a ground and the amplifier will have it's own ground somewhere else in the vehicle. No way to really get around this unless you isolate the mounting for the amplifier and run a ground wire just as heavy as the power wire to the amplifier and then to the ground on the radio. This is not practical in most cases, so when noise is present on the RCA cables, you can buy ground loop isolators for them. The different grounding points can cause small currents to flow through the RCA shielding causing annoying noise in the stereo system.

A vehicle is a very noisy place electrically. Fittech should have already considered this in their design. Some designs are better than others.