20
\$\begingroup\$

I am more of a software programmer, but I have a basic question on electronics.

What happens when a device draws more current than the power supply can provide? And under what circumstance can it happen - bad board design or not determining the worst case scenario of the device operation?
How does one actually see this on a scope? What in turn happens to the device if this is not rectified?

\$\endgroup\$
18
\$\begingroup\$

There are many different things that can happen when you pull too much power and this will be dependent on the technology being used. The typical things you will see is the voltage dropping below the spec'd output or cutting out completely. Some systems might have a fuse that trips when you pull too much power. Or the worst case is you exceed the safe rating of some components that result in something over heating and potentially causes lots of problems (fire, etc).

The circumstances that this can happen pretty much all fall on to the board designer. Many times it is cheaper to build a power supply that can provide less current, so the designer will minimize the cost as much as possible while still being safe. If there is some case that the board designer hasn't considered, then they might pull too much power. These cases might be things like running an RF module while doing some DSP work, while the designer originally intended for them to only operate at separate times.

The other time that you might pull too much power is under an actual fault condition. This is when something truly goes wrong, like an IC fails into a short condition, or someone accidentally shorts two connections together, or something along these lines.

As to what happens to the device, this also all depends on the device itself and the PSU being used. If the PSU drops its voltage, potentially the device might fail if the system can't handle the voltage.

\$\endgroup\$
1
  • 6
    \$\begingroup\$ As a little "information for the layman": Theoretical power supplies can output infinite current. Practical power supplies have an internal resistor that is the sum of all the wiring and other components. In the model of such a power supply, the resistor is what causes voltage to drop as current increases. The power lost (converted to heat) through this internal resistance is why a power supply needs cooling. \$\endgroup\$ – Vincent Vancalbergh Jan 27 '14 at 10:57
9
\$\begingroup\$

Everything Kellenjb has told you is absolutely true. What I would like to add is some further information on how do deal with this issue when building electronic circuits.

You have two pieces of gear involved, a power supply and the circuit under test. Now clearly if the circuit under test tries to draw too much power, and the power supply is not able to safely deal with it, then bad things (like a fire) can happen.

Consider what happens when there is a bug in the circuit under test and it draws excessive power from a supply (like a car battery) that is capable of providing lots of power safely. In this case it is not the supply that is the problem but rather the circuit under test that might catch fire, or more likely release it's "magic smoke".

The way this is generally dealt with in a lab where prototypes are built is to use a bench power supply that can be current limited. It will generally have two displays (meters or LEDs etc) on the front, one for voltage and one for current. It will also have a way to set both the desired voltage and or current. It will operate in either a voltage limited or current limited mode.

Now suppose you put together a circuit on a breadboard and you compute that it will require 5 V and use a maximum or 100 mA. You would put those values into your bench power supply which would guarantee that the voltage would never exceed 5 V (it could be less if you short the terminals) and that no more than 100mA of current will be delivered. If your circuit attempts to draw too much power the power supply will give an error indication and prevent excessive current from flowing.

While such a supply can not guarantee that the components in your circuit will never be damaged (you could wire the leads up backwards for example), they can greatly reduce the chances of destroying a component and perhaps more importantly prevent a fire.

\$\endgroup\$
2
  • 4
    \$\begingroup\$ A load circuit might, very briefly, draw more than the power supply can output when the load is first turned on. For instance if the load has a lot of capacitors that need to charge up, the charging current might flow in a large spike, which then settles down to much less than the supply's capacity once the capacitors are charged. For this reason it's better to set your bench supply above where you think it needs to be, and watch the current flowing at startup with an oscilloscope current probe. Then design the PCB supply to handle it. \$\endgroup\$ – Matt B. Feb 5 '12 at 23:39
  • 2
    \$\begingroup\$ @Matt I have seen boards act like they have a dead short because it didn't have enough current to start up. So +1. \$\endgroup\$ – Kellenjb Feb 6 '12 at 4:12

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.