2
\$\begingroup\$

I am a software engineer and I came across this article which states

If a processor attempts to draw more power than a power supply system can provide (by drawing more current than the system can supply), the result is typically a volt- age drop, which can cause the device to malfunction. Modern processors can vary widely in power consumption with high peak currents; hence, they provide voltage indexing methods that allow the processor to slow down and regulate voltage within a wider margin.

I have read other articles where they explain that you can imagine voltage and current relationship as water hose, where Voltage is equivalent to pressure, water flow is equivalent to current and the diameter of the pipe is equivalent to the thickness of the wire - or resistance. So my question is two fold:

  1. How does drawing more current leads to voltage drop, because voltage and current are directly related, right?
  2. We need both high voltage and high current to cause damage, then why the sudden drop in voltage causes any damage to computers. Causing a drop voltage should only mean less force with which the electrons flow, so it should be opposite to causing damage, isn't it?

I am sure I am missing some basic physics logic here. Can you please clear this for me?

\$\endgroup\$

5 Answers 5

Reset to default
5
\$\begingroup\$

You said yourself voltage is like pressure and current like flow rate. If you measure the pressure in your water line just before the sink valve, it will be maximum when the valve is off, and go down as the valve is opened more (letting more water flow). The pipes back to wherever the water pressure is maintained have some resistance. This resistance times the flow causes a pressure drop.

In the case of the processor and its power supply, the power supply is more like a pump. This is like a water pump that has a maximum flow it can sustain. If you tried to draw 10 gal/min from a 5 gal/min water pump, you're not going to get a lot of pressure, and you're not going to get 10 gal/min.

It's the same thing with a processor and its power supply. Let's say the power supply is rated for 3.3 V and 200 mA. That means it only promises to maintain 3.3 V if you draw 200 mA or less. If you try to draw more, various things can happen, but they all include dropping the voltage.

Note that the section you quote talks about the processor malfunctioning when it gets too low a voltage. This is not the same as getting damaged. Low voltage to a processor is not going to damage it, but you can't expect a processor designed to run on 3.0 to 3.6 V to work at 2.0 V. At too low a voltage, transistors can't be turned on fully anymore, which means they can't turn on other transistors fully, etc. The digital levels become undetermined and take longer to propagate thru the logic. The processor performs meaningless operations on meaningless values, if it can "operate" at all. At some points the clocks fail to clock.

\$\endgroup\$
2
  • \$\begingroup\$ I understood the part where processors fail to operate properly due to low voltage. Can you explain the voltage drop bit more? I still dont understand why would the voltage drop to lower value with more current? I am sorry for the noob question. \$\endgroup\$ Commented Feb 20, 2016 at 18:28
  • 1
    \$\begingroup\$ @Unde: Look up "Ohm's Law". Current thru a resistor causes a voltage drop across it. Also, again, a 200 mA power supply isn't going to work right when you try to draw 300 mA. If it did, it would be sold for more money as a 300 mA supply. \$\endgroup\$ Commented Feb 20, 2016 at 18:33
3
\$\begingroup\$

When you attempt to draw more power from a power supply than it is designed to deliver, something's gotta give. Usually the supply voltage will drop, as the supply attempts to keep the current within its capabilities.

The reduced voltage will rarely damage anything, but may cause the attached circuits to malfunction - logic circuits will see invalid levels, so counters may mis-count, processors will fetch invalid instructions, etc. When the voltage is restored to the correct level, the circuit will normally resume proper operation.

\$\endgroup\$
1
\$\begingroup\$

Voltage drop is voltage that is lost in the resistance inside the power supply and in the wires between the power supply and the load. The power supply can compensate for its own internal voltage drop, but there is a limit to its ability to compensate. That limit is considered when defining the rated output current.

Causing a device to malfunction and causing damage are not the same thing. Low voltage can cause a logic chip to give a false output or mis-interpret an input. That is a serious problem, but that is not damage. Logic errors that lead to data errors in a file may lead to a corrupted file.

Some loads, particularly motors, attempt to keep delivering the proper output when the input voltage is reduced. In doing that, they can draw more current and overheat. Only resistors obey Ohm's law. Active devices can behave as if the resistance is negative.

\$\endgroup\$
1
\$\begingroup\$

My answer to your question: "Can (I) clear this (confusion) up for (you) is: "Yes, I think I can.

The expression about the results of drawing excess power is in itself misleading. It suggests that by consuming more power you get to be acquiring less power. Power is a result rather than a cause. Maximum power expended in an electrical load when the load impedance is equal to the source impedance. Dealing more within determining factors, drawing excessive electrical current increases the voltage drop across the source impedance resulting in a reduced share of voltage across the load. The prime factor involved in having increased load current is to have reduced the impedance of the load. Continued increase of current beyond the matched impedance condition ultimately approaches the horrors of a short circuit.

Among the many extended complexities of circuitry is the fall-off of counter EMF for a motor whereby it functions as a generator resisting excess current. (An undercurrent can bring harm.)

\$\endgroup\$
0
\$\begingroup\$

Voltage drop will surely happen when the current draw exceeds the capability of the source. Almost all computer systems are powered by switching power supplies frequently at DC 5volts and 3.3volts.

This voltages are further bucked down by on board regulator to the regions of 1 volt to be used by processors, memory chips and chipsets. The exact voltages depends of the specifics.

Intel i7 processor requires 130w at 1 volt needs 130 Amps of current. The amazing thing about processor power requirement is, during idle it may require only 5Amps but at the moment a program is executed the current jumps to 100 Amps in microsecond. All available low ESR capacitors at the vicinity of the regulator will help to maintain stability during this sudden changes.

This demand will increase the main current at 5 volt lines proportionally. The current transients normally are well damped by capacitors.

Problem occurs mostly when the capacitors degrades. Voltage drop below minimum stable operation threshold during current steps from low to high.

Such circumstances cause indeterminate logic states in computer operation causing data corruption. There may not be hardware damage but data loss, program halted with blue screens are real cause of such instability.

\$\endgroup\$

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.