5V regulator for automotive source

I was answering about the diode to protect against reverse polarity. For the pther things of course you need something else.

Bye Jack

When I built my own audio equipment for my car a long time ago, I found that the horrors mentioned in this thread are more common than you'd expect. The brake light switch caused voltage to spike down to about 4V when pressed and spike up to over 40V when released. Nothing smaller than soda can sized capacitor could handle that directly. Starting the car caused surges stronger than any capacitor I had.

That's why small electronics have an RC filter and larger electronics an LC filter.

You can tame things down quite a lot, by refitting the battery connector with one that takes multiple wires. Don't put anything on your branch but your electronics... none of the car functions. ...Jim Thompson

I have measured the "bright" filament of an 1157 tail lamp at 0.5 ohms cold. Most cars that use that part number have either two or four of them in parallel. For a 14 V nominal voltage, that's 56 or 112 A for a few milliseconds, until the filaments start to warm up.

I don't know where the spike on release would come from, because that's "only" unloading by about 4.4 or 8.8 A. Maybe the alternator regulation isn't really that good.

Matt Roberds

Energy stored in the wire inductance, perhaps?

Must be pretty fast brake switch which drops the current to zero in a microsecond :-)

Assuming 1 uH/m wire inductance and 3 m wiring to the brake lights that would be 3 uH. With 8.8 A/us that would give about 30 V peak riding on top of the 12 V rail.

Which is close to his 40V statement. So, it seems to be a good estimate to gain understanding.

My point was to show that it would require a switch acting in 1 us after the foot was removed from the brake pedal and the spring starts to move the pedal to the normal position.

This is of course quite unrealistic.

Due to arcing, the disconnect time is several milliseconds instead of single microsecond.

Maybe. Contact arcing introduces another unknown rather than a known. If it is arcing, you can't say that the current diminishes in several milliseconds because it is at the mercy of many variables. I don't think you can anticipate the results of the unknown characteristics of arcing in the unknown properties of the wiring such as capacitance, transmission line effects, etc. I think you did a good job in your assumptions and resulting estimate.

Do you have data or a link to the waveform generated by an opening brake light switch? Or, any other information pertaining to this. I would be most interested in seeing the interruption of a few amps by a switch powering a resistive load.

Thanks.

A short search turned up this paper...

What are your thoughts on this?

I wonder what voltage is required to start the arc?

Excellent question! 12V is a bit low.

hen

ohms

of

for

hat's

lation

te

not really when it is at disconnect, it's like scratch start welding

-Lasse

On Sun, 10 May 2015 08:36:41 -0500, John S Gave us:

On Sun, 10 May 2015 06:48:45 -0700 (PDT), Lasse Langwadt Christensen Gave us:

Even old points contact ignition circuits only degraded because of the inductive back EMF from the coil, NOT the switching current or LV arcing. And they fire several hundred times a second and are quenched just fine as there is nothing (much) to quench.

My point was if it takes 40V then you will get 40V.

eroUno:

hopefully, or they wouldn't last very long

there was a capacitor across the points to limit arcing

-Lasse

If the brake lights draw that much current, the vehicle probably has a brake light relay. If the relay doesn't have a back EMF protection diode across the coil, opening the coil could produce a substantial voltage spike: The brake switch might also have an arc suppression capacitor across it, which is helpful for protecting the switch contacts, but also helps deliver the voltage spike to the rest of the 12V electrical system.

Brake light relay retrofit: inspired by a junk brake switch.

The 40V spike is quite real. I've seen similar spikes in my various vehicles although I don't recall the peak voltages and durations. ISO 7637-2:2011 covers electrical transient conduction along the 12v wiring:

My prize for the best automotive "big bang" goes to a friend when he accidentally removed a battery lug connection after a jump start while the car was running. The alternator regulator thought that the battery sense voltage was low, and preceded to supply maximum power and charge current until a blown fuse convinced it that this was a bad idea. I don't recall how much was damaged, but I do recall that all the 2-way radios were smoked because I was bribed into fixing them. Pretty much all the fuses, voltage regulators, RF output xsitors, audio power amps, and reverse protection diodes, were smoked.

A similar story:

Oops. Wrong link: