next topic »

[Reply #15 on:]

I was informed long ago that a car battery is anything but 12v.  You say "it has 13.8v peaks" but it can be A LOT higher than that.  For a beginner, there isn't really an easy answer here.  Using LM7809 or equivalent regulators is wasteful of battery power (excess voltage converts to heat) and still needs to be regulated by the Arduino to 5v, wasting more power still.  Also, the electrical environment in a car is very unreliable and dirty, so unless you're using a stout regulator, it may not survive very long.

The correct thing to do here is have a robust filtering stage to try and stabilize the 9 to 18v input, then use an efficient switching regulator to output a clean, regulated 5v directly to the Arduino's power bus rather than using the onboard DC input.  This isn't quite trivial, and I've been working on a design off and on over the last year to do just that.  I have it completed, I just need to have the PCBs made and run it through a series of tests.

Anyway, good luck with your project and be careful.  As with many things in electronics, the problem is often much more complex than you thought.  Makes for a steep and frustrating learning curve, but it gets fun when you start to understand it all.

[Reply #16 on:]

I was informed long ago that a car battery is anything but 12v.  You say "it has 13.8v peaks" but it can be A LOT higher than that.

Well to be clear it's not the battery causing those over 12.6vdc evnets but rather the alternator/regulator charging system of the car. When running the battery is not supplying any current but rather is just charging and the alternator/regulator is what is powering the auto.

Lefty

[Reply #17 on:]

To sum up

Put a regulator there and call it a day.

[Reply #18 on:]

Yeah, I bought 500 x zero-ohm resistors for $4.

I purchased some of these http://uk.farnell.com/panasonic/erj6gey0r00v/resistor-0805-0r-5-0-125w/dp/2057661RL recently. They are described as 0 ohm +/-5%. The ones I received had a resistance of at least 0.001 ohms, so I was thinking of sending them back for being out of tolerance. 

[Reply #19 on:]

Arduino board is a active element, ohm's law apply but little bit tricky in calculating power consumption. For Arduino board case, using linear regulator, it only draws current it needed (around 40mA without external shield connected) for a range of input voltage.

For converting 12V input to 5V VDD, power distribution is:
A) Total input power 12V x 40mA = 480mW
B) Dissipation on voltage regulator (12V-5V) x 40mA = 280mW
C) Dissipation on ATmega MCU and peripheral components 5V x 40mA = 200mW

If you add a led shield, current consumption goes high (>40mA) and power dissipation in both B and C part will increase proportionally. But, if input voltage increase (>12), only power dissipation on voltage regulator B increase.

Let's go back to your questions:
- An electronic device, when it's connected to a battery, only gets the amp required or the battery gives more and the device breaks?
Ans: It only gets the amp it required, for linearly regulated circuitry.

- A voltage stabilizer is good or i need one with current limit?
Ans: For the case of car battery, voltage pulse can up to 14V with running engine, stabilizer is recommended. From my experience, some voltage regulators cannot sustain such a high voltage drop, let's say from 13V to 5V. And high power dissipation across it (as in B above) will create heat and burn the regulator ultimately.

[Reply #20 on:]

Keep in mind that a simple linear regulator may lack sufficient protection to be directly connected to an automotive power source.  What you "should" see and what really goes on are not the same.  Alternators are not ideal sources, and the spark plugs and lots of inductive stuff generally do not help matters.  You will have people tell you "my LM7805 has been running for months without a problem", and it may continue to do so.  But counting on that is foolhardy, IMO.  At least chuck a diode, inductor, and cap on the input.  :-)