0
\$\begingroup\$

A small (3.3V/10mA) uC circuit requires a ground. I designed the circuit’s PCB to use a nearby ground wire that is attached to the car’s frame in at least two places and used by the 2A windshield washer motor (other devices may use it as well). I created a break in the ground wire (denoted by A and B) and it is soldered to the PCB onto two pads nearly touching and connected by a 100 mil trace (2 oz), more than adequate for the washer motor. (The trace width calculator indicates 8 A conservatively.) Connecting the remaining ground wire C completes the uC circuit and everything works fine; the washer and the circuit. Both are on the same 12V supply but I’m curious if there is anything inherently wrong with this approach. Ideally, (from what I’ve read) I would add another separate ground wire for the uC circuit, but I would like to avoid adding another wire if at all possible. The uC is only controlling a relay. Should I isolate the A-B pads from the ground pour using a polygon cutout and then connect the uC circuit grounds to those pads with a single connection instead of having A and B directly connected to the ground pour? Does it matter?

There are numerous threads about connecting grounds, but the many I've read address noise and different voltage potentials; I have not seen anything that addresses sharing a ground with another higher-current circuit.

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
  • 1
    \$\begingroup\$ Related: electronics.stackexchange.com/questions/97198/… \$\endgroup\$ – Matt Young Mar 23 '16 at 12:13
  • \$\begingroup\$ I wonder why there is a wire running from the chassis to the 'metal dash'; I assume that it is not connected to the chassis in any other way? \$\endgroup\$ – JvO Mar 23 '16 at 12:16
  • \$\begingroup\$ @JvO The car is nearly 40 years old; there are very few non-conductive parts and the metal body is used extensively as a ground. Some items don't even have a ground wire, per se, like the AM radio. It has a 12V power wire and is bolted to the dash for the return path. (My license plate lights didn't work until I realized some powder coating had broken the lights' ground path.) \$\endgroup\$ – unix Mar 23 '16 at 14:32
  • \$\begingroup\$ @Matt Thanks for the link; very helpful. Had not found that one in my searches. \$\endgroup\$ – unix Mar 23 '16 at 14:35
  • \$\begingroup\$ Then the wire only seems redundant... Perhaps you were trying to save some wire but I'd recommend drawing a new ground wire from the closest metal car part to your PCB. \$\endgroup\$ – JvO Mar 23 '16 at 14:48
1
\$\begingroup\$

In this particular case, a single point ground from the microcontroller power return makes sense as it will tend to attenuate the very noisy motor power which could otherwise cause intermittent operation of the controller.

Grounding and power schemes in general have to be analysed on a case by case basis; there is no 'one size fits all'.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.