I've often seen devices with power requirements specified only in Volts (e.g. 7-12V) but never the amperage. I've wanted to run various embedded devices of wall warts and batteries (the devices have regulators don't worry) however I've been hesitant because I'm not aware of the amperage requirements for the devices.

My question is: Is there a standard amperage that is "understood" for microcontrollers and the like?

I have been told that amperage doesn't matter however I beg to differ, since I am quite sure that if I supplied a 7-12 volt device with 9 volts at 1 billion, amps that it would explode.

EDIT: To put it simply. A power supply is rated at the amps that it will tolerate before overheating and taking damage?

  • \$\begingroup\$ An example would be interesting \$\endgroup\$ – Brian Carlton Feb 25 '11 at 0:11
  • \$\begingroup\$ I think I finally understand this. As for a real world example: If I have a stepper motor rated at 1.2 amps per phase and I try to run it off of a powersupply rated at 650mili amps... The powersupply will fry. \$\endgroup\$ – user3045 Feb 25 '11 at 4:28
  • \$\begingroup\$ Not necessarily. It should be protected against an over-current situation, if it has been properly designed. \$\endgroup\$ – Leon Heller Feb 25 '11 at 15:52

Voltage (which is kinda like the strength of the supply), and Current (measured in Amps, which is the quantity of electricity), are two very different things.

Voltage: When trying to match a supply to a device, you need to get the voltage right... if the supply voltage is too high, then it will damage your device. If the supply voltage is too low, then your device just won't work.

Current: When looking at current, you need to ensure that the Amps rating is higher than the device needs as it will only use as much electricity as it needs. If the rating is too low for the device, it will be trying to get more electricity from the supply than the supply can provide, and so it will get hot and possibly explode. If you had a supply that was rated at 1 billion amps, then it would happily power a tiny bulb... it just means it could also power 1 billion bulbs or more a the same time!

So, the possible dangerous situations are:

  1. If the voltage is too high for the device.
  2. If the amps are too low for the device.

As a general rule, devices that produce a lot of heat or light or movement usually need a high current supply. Devices that control things, like a TV remote or some small gadget with maybe a few LEDs on it, won't need a lot of current.

To answer your question, the microcontroller itself probably only needs between 0.02 and 0.1 amps. If the microcontroller is controlling something else, and sharing the supply, then the current rating of the supply really depends on the device.

  • \$\begingroup\$ How exactly is it dangerous if the current is too low. Say a device needs 350mA and I have an adapter at 300mA what would happen? \$\endgroup\$ – Dean Feb 25 '11 at 10:02
  • 1
    \$\begingroup\$ A device doesn't 'need' 350 mA. It will let through 350 mA at a specified supply voltage. Ohm's law: I = U / R. With a constant resistance device, it will let a current directly proportional to the voltage applied on its ends. So, if a device takes 350 mA at 10 V, it will let through 700 mA if 20 V supplied through it. \$\endgroup\$ – Andrei Sosnin Feb 25 '11 at 10:12
  • 3
    \$\begingroup\$ Of course, the above is only true for purely resistive circuits (incadescent lamps, LEDs etc.). If you only get 300 mA of current through the same circuit, chances are that you're supplying only a fraction of the voltage needed. It's not dangerous in terms of safety, but it's dangerous in terms of device's functionality, especially if the device is not just a light bulb, but a microcontroller, for example. \$\endgroup\$ – Andrei Sosnin Feb 25 '11 at 10:21
  • 2
    \$\begingroup\$ @Dean: Maybe I should rephrase that to "possible dangerous situations". I was more thinking of something like powering a 2.5A kettle with a piece of cable rated at 0.5A... it would get hot, possibly melt and catch fire... hence dangerous. \$\endgroup\$ – BG100 Feb 25 '11 at 10:35
  • \$\begingroup\$ @BG100 ok I now understand. \$\endgroup\$ – Dean Feb 25 '11 at 11:12

If you connect a 5 V 100 mA device to a 5 V 1 billion amp power supply, the device will draw 100 mA.

  • \$\begingroup\$ i'd like to try that. where can i get a supply like that? \$\endgroup\$ – JustJeff Feb 25 '11 at 0:42
  • 5
    \$\begingroup\$ Just finish your dyson sphere, and line it with solar panels. \$\endgroup\$ – Connor Wolf Feb 25 '11 at 2:06
  • 4
    \$\begingroup\$ "commercial solar cell have short-circuit currents between about 28 mA/cm2 and 35 mA/cm2." If we assume this is at typical solar irradiation at Earth distance from the sun, the surface area of the Dyson sphere would be wolframalpha.com/input/?i=area+of+sphere%2C+radius+1+AU and the short circuit current would be wolframalpha.com/input/… 10^16 billion amps \$\endgroup\$ – endolith Feb 25 '11 at 3:23
  • 1
    \$\begingroup\$ Let's just call them giga-amps... :-) \$\endgroup\$ – JYelton Sep 7 '13 at 5:29

Of course it wouldn't, a device will only take as much current as it requires (Ohm's Law). The maximum current capability of the supply is irrelevant, as long as it is greater than the peak current rating of the device.


I agree with Leon. Just because a power supply can supply some maximum current doesn't mean that the device being powered will draw that much.

As for your question of an "understood" power rating for microcontrollers and the like, you can find the answer for microcontrollers by looking in the datasheet. This will, of course, vary greatly with the microcontrollers. Those typically discussed on this site (PICs, ARM Cortex-Mx, AVR, etc.) are relatively low power consumers (usually a few milliamps or tens of milliamps) compared to what a typical wall wart will supply. I'd wager you would be hard pressed to find a generic wall wart that supplies less than 100 mA at some retail outlet so, generally speaking, it won't be an issue. That being said, I can totally understand your frustration with the lack of good documentation.

  • \$\begingroup\$ It varies widely depending on how you use the micro, too. I remember they have actual charts of idle current consumption for different clock speeds, etc. \$\endgroup\$ – endolith Feb 25 '11 at 0:02
  • \$\begingroup\$ So then a regulated wall wart producing 7V at 1 amp should cover most situations? \$\endgroup\$ – user3045 Feb 25 '11 at 4:38

The "exploding" misconception, as far as I can see, basically comes from not understanding to what kind of ideal generator commonly used power sources can be approximated to.

Basically, we have two types of ideal generators. Ideal voltage generator and ideal current generator.

The ideal voltage generator has two contacts and provides constant voltage across them, no matter what kind of load we use. The output current comes from Ohm's law and that's why they mustn't be short-circuited at outputs. It basically makes current available to load connected at its output.

The ideal current generator provides constant current through its contacts, no matter what kind of load we use. The output voltage comes form Ohm's law and that's why they must always have a load or be short-circuited. It basically pumps current through its outputs.

To make yet another overused water analogy, ideal voltage source is like a lake from which you can just pick up as much water as you need, while ideal current source is like a pressurized pipe which will provide steady stream of water until it is closed.

In real world we don't have such ideal generators and the real sources which are generally available to common people are much closer to ideal voltage generator than to ideal current generator. So if you have a common power supply which is rated at say 9V and 1 GA, that means that you can approximate it as ideal 9 V voltage generator up to output currents of 1 GA. When output current needs to be higher, it will stop acting as an ideal voltage source and will start showing imperfections such as voltage drop, overheating, current limit and so on.


Voltage and resistance is all that matters.

For simple (non-reactive) devices like steppers/speakers/etc, the current is determined by a very simple equation:

current (amps) = voltage (volts) / resistance (ohms)

So, given a fixed voltage and the fixed resistance you can calculate amperage. It's that simple.

A Power supply with a rating for a certain amperage tells you a few things.

First, the power supply is only built to handle that amperage. Wiring, resistors, and other devices heat up based on more or less on much how current you put through them. A thicker wire heats up less, and can therefore handle more current without danger of melting or fire. This is because there's a larger cross-sectional area for the power to be distributed over. (although it's not quite so simple if the voltage has a high frequency AC component to it) So, you don't want the supply to go significantly beyond it's rated range. It may be built with thinner wire and burn up.

Second, many power supplies are fairly dumb devices (unregulated). If they're rated at 12V @ 1A, they may give you 16V at 0.25A, or 10V at 2A (if they don't burn up). You only know that you'll get 12V at exactly the rated voltage. This can give you problems if you put a 12V 5A supply on a device that only draws 100mA (it may end up giving the device 16V+)

Third, supplies also have an internal resistance. So: CURRENT = VOLTAGE / ( RESISTANCE_OF_LOAD + INTERNAL_RESISTANCE_OF_POWER_SUPPLY). So, the current it is able to supply to the load is limited somewhat by this internal resistance. Your 1.2A rated stepper example on a 650mA supply might only be able to draw 900mA for this reason. (For a stepper, that'd usually just mean it operates more slowly and has less torque)

Forth, the supply may have active current limiting. If your mentioned 650mA supply had current limiting, it may limit the maximum current (for safety) to 700mA.

The best power supplies are actively regulated. This means that a microcontroller or some feedback circuit is watching it's output and adjusting to always give you the rated voltage. They usually have current limits, too... so these are the safest type of power supplies. However, many are switching-mode power supplies instead of linear and may add noise, so they may be undesirable for certain devices (high performance audio comes to mind).

So... there's lots of factors that basically mean use a power supply that's close to what your load needs unless you know for sure that it's regulated. Never use a supply rated below what your load needs, unless you have a very good understanding of both the load and the supply and how they'd react to that.

Reactive devices (like microcontrollers) can dynamically change their resistance to suit their needs. Running these devices with less power than they need would usually mean some sort of incorrect operation.

  • \$\begingroup\$ Is the Active PFC on computer supplies an example of current limiting? \$\endgroup\$ – user3045 Feb 25 '11 at 16:59
  • \$\begingroup\$ No, PFC is correcting for a reactive load (where the current draw changes over time). It's closer to current 'smoothing'. A current limit circuit will effectively dynamically alter it's resistance to keep the current from rising above a set point. (INTERNAL_RESISTANCE_OF_POWER_SUPPLY mentioned above goes up) \$\endgroup\$ – darron Feb 25 '11 at 18:56

Current is something that is taken in by a device, Voltage is something that is provided by a source. A motor in stall condition for example, would require more current/juice to run, so it will draw more current from the battery, which is providing voltage and current. There are supplies or sources which limit the maximum current that can be drawn from them.


protected by Dave Tweed Aug 7 '14 at 19:21

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?