Is it necessary to "clean" an automotive ground connection to Arduino

(deleted)

schemeit-project.pdf (274 KB)

You can normally connect the chassis ground directly to your Arduino ground. The only reason not to is if it's already connected and you don't want to make a ground loop.

But all other wires entering or leaving your box should be protected against screwups. The most obvious screwup is connecting it to power with the wrong polarity. I've done that a hundred times. There's actually a long SAE standard that describes all the screwups possible with an automotive electrical system.

The car might get jump-started from a 24V battery, the occupants may create static electricity and then a spark jumps to a wire, or even more esoteric things. I've seen a 'load dump' in my own car many years ago and it blew out every single light bulb in the entire car.

So imagine what might happen if your device is connected to -12V or +24V for a long period. Imagine what might happen if it gets a static zap of 15kV for a nanosecond. Imagine what might happen if it's un-plugged in a strange way that breaks the ground connection first.

Almost everything can be dealt with by a series resistor on the inputs. That limits the fault current possible in most scenarios. For everything else, a TVS diode on each input is a great way to go. I like the SMAJ15CA a lot.

Also, the ignition system generates a lot of noise, as well as the alternator. A car is a brutal place for electronics, so the most protection as possible!

Nah. Accessories are off during ignition.

(deleted)

And an alternator with a functioning voltage regulator isn't substantially 'noisy' to be worth worrying about.
Just think about all the delicate electronics that are in your car. Even 12V accessories like GPS or simple phone chargers that are on the 12V accessory line, which is sometimes only fused at 10A. These cheap devices somehow manage to survive this supposedly atrocious environment. Either these cheap devices are not as combustible as doomsayers like to preach, or the car's accessory line isn't as hellish that those doomsayers also like to preach. Can't be both.

INTP:
Nah. Accessories are off during ignition.

No, I mean all that high voltage stuff – you know that ignites the air/gas mixture.

INTP:
And an alternator with a functioning voltage regulator isn't substantially 'noisy' to be worth worrying about.
Just think about all the delicate electronics that are in your car. Even 12V accessories like GPS or simple phone chargers that are on the 12V accessory line, which is sometimes only fused at 10A. These cheap devices somehow manage to survive this supposedly atrocious environment. Either these cheap devices are not as combustible as doomsayers like to preach, or the car's accessory line isn't as hellish that those doomsayers also like to preach. Can't be both.

Yeah, I'm probably thinking of them old cars with the mechanical voltage regulators -- old fart, here :stuck_out_tongue_winking_eye:

Yes, I would highly recommend NOT wiring your Arduino up to your ignition coil. Whew. Crisis averted.

INTP:
Yes, I would highly recommend NOT wiring your Arduino up to your ignition coil. Whew. Crisis averted.

No dufus [and I mean that with all due affection ;D] -- the ignition system can impart some nasty noise on the 12V system -- or is this no longer an issue? Again -- geezer, here. This used to be a problem -- is it no longer? It used to be that electronic gear, connected to a cars power supply, needed to be well protected with spike suppressors and passive EMI filters and such.

Update:

I did some Googling and found the following:

ETC!

I must agree with ReverseEMF, a constant topic with automotive electronics is that the power source (the 12V line) has the potential to be VERY dirty and under certain conditions may experience some rather nasty high voltage levels, spikes both high and low..

INTP:
And an alternator with a functioning voltage regulator isn't substantially 'noisy' to be worth worrying about.
Just think about all the delicate electronics that are in your car. Even 12V accessories like GPS or simple phone chargers that are on the 12V accessory line, which is sometimes only fused at 10A. These cheap devices somehow manage to survive this supposedly atrocious environment. Either these cheap devices are not as combustible as doomsayers like to preach, or the car's accessory line isn't as hellish that those doomsayers also like to preach. Can't be both.

The cheap devices have protection. Sometimes not entirely perfect - I have a phone charger that lets some noise through but it's only detectable if I'm using the headphone jack on the phone.

Jump starts are rare. Rarer still to hook the jumper leads up backwards or hook a 24V vehicle to a 12V vehicle. IF you did that, then I'm sure the cheaper devices would be melted.

Often "protected" means "will not set fire to other things." There's nothing that will survive a direct hit by lightning but the well-engineered devices will not pass on that destruction to a wide area.

Just for completeness sake, you should be aware of this snippet from the Atmel datasheet. (If you can call 400+ pages a "sheet").

Unless specifically provided otherwise, Atmel products are not suitable for, and shall not be used in, automotive applications.

INTP:
And an alternator with a functioning voltage regulator isn't substantially 'noisy' to be worth worrying about.
Just think about all the delicate electronics that are in your car. Even 12V accessories like GPS or simple phone chargers that are on the 12V accessory line, which is sometimes only fused at 10A. These cheap devices somehow manage to survive this supposedly atrocious environment. Either these cheap devices are not as combustible as doomsayers like to preach, or the car's accessory line isn't as hellish that those doomsayers also like to preach. Can't be both.

The sorts of accessories you describe would all have voltage regulator circuitry that mitigate voltage spikes on the automotive power bus. Such circuitry should be designed to tolerate voltage spike considerably higher than the nominal system voltage. For example even though the nominal voltage is about 14.5 V, the power interface might be specified to tolerate double that as it will almost certainly see very short voltage spikes on the order of 20 V. Doing so is not inherently expensive but not doing so may lead to failed electronics.

If one has an oscilloscope, it is instructive to hook it up to automotive power and take a look at the crap coming in.

Alternators are clamped to 40v iirc. Battery soaks up spikes. It's my opinion that the doomsaying about the hellish car environment is just 'better safe than sorry' run amok, just a safe wager where being wrong (and things go smoothly forever) has no consequence and being right ennobles an 'I told you so' cavalier.

Fuses deal with current, voltage spikes are not in the hundreds, I think a sensible caution be practiced, sure, but know what it is you're being cautious against. Keep them fears in check, is all I'm saying.

INTP:
Alternators are clamped to 40v iirc. Battery soaks up spikes. It's my opinion that the doomsaying about the hellish car environment is just 'better safe than sorry' run amok, just a safe wager where being wrong (and things go smoothly forever) has no consequence and being right ennobles an 'I told you so' cavalier.

Fuses deal with current, voltage spikes are not in the hundreds, I think a sensible caution be practiced, sure, but know what it is you're being cautious against. Keep them fears in check, is all I'm saying.

Then why do reputable "experts" like ST, TI, LittleFuse, ElectronicDesign, even Mouser, stress the need for unusual measures for automobile electronics? Are they all overreacting? Have a look at the elaborate TI "AUTOMOTIVE LINE TRANSIENT PROTECTION CIRCUIT SCHEMATIC" on page 5. Sure, it may be overkill, and a lot of it may deal with infrequent events, but heck, should such an event occur, do you really want your device to fail? And not all of the mitigations are for infrequent events. Have a look at Table 1 in the Electronic Design article.

Here's an EevBlog disscussion that kind of goes middle-of-the-road [ha! that's a pun!!]

Because fear sells products.

INTP:
Because fear sells products.

Ahhh...I see -- conspiracy theory :grinning:

I suppose you could be right. Perhaps a scope on a car would be revealing in these matters...'cept mine is a Hybrid, so possibly not representative. Or, one could try placing raw electronics across the 12V line on a typical car to see how long it survives. Perhaps a control would be, the same electronics on just a 12V battery, and/or a 12V battery with a trickle charger attached...? Or, the same electronics on numerous cars -- for a more statistical result.

BTW:

Battery soaks up spikes.

Not if the source of the "spike" is far from the battery, and other gadgets are connected, far from the battery and on the same line(s) as the spike generator. There is the possibility of a voltage drop on the line(s) to the battery, when a current surge occurs, and things like Reverse EMF from coils [like those in a Motor or Relay, if not sufficiently bypassed] might not be absorbed by the battery, due to propagation delay and wire inductance. Also, wire inductance can turn a current surge, into a voltage spike, which also wouldn't, immediately, be absorbed by the battery, because of the wire inductance between the source of the surge and the battery. These would be brief events, but electronics can be susceptible to even spikes that are this brief.