2
\$\begingroup\$

My apologies in advance if I make any wrong assumptions. Say you have an outlet that is rated at 120V and 10A. My power adapter for my MacBook says it is rated at 16.5V and 3.65A max. How was it determined that the power adapter needs to provide at most 16.5V and 3.65A? What is the process of actually figuring this out when designing electronics/circuits?

The only thing I have been able to think of so far is that based on the physical properties of a circuit, you can only handle a certain amount of current before burning out which can be figured out empirically. In addition to that maybe the amount of voltage required has to do with power efficiency and driving the max amount of current you can handle.

I understand Ohm's Law and I am not asking about how to calculate anything using it. This is more of a design question when you have a circuit in mind but need to figure out what resistors to apply and what voltage/current you will need to operate correctly.

\$\endgroup\$
3
  • \$\begingroup\$ That's what the adapter provides, not what it needs. \$\endgroup\$
    – Ignacio Vazquez-Abrams
    Commented Sep 24, 2014 at 23:38
  • 1
    \$\begingroup\$ Circuits are typically designed to work at an optimal voltage, or small range of voltages. It may also be designed to consume no more than a maximum current. If the current is not known, then once the circuit is completed, the current is measured. That is it. \$\endgroup\$
    – gbulmer
    Commented Sep 24, 2014 at 23:42
  • \$\begingroup\$ The MacBook designers determined, in the design of the MacBook, that the power adaptor should provide 16.5 volts. By measurement or calculation they determined that the adaptor should provide not less than 3.65 Amp. The power adaptor designers designed the adaptor to meet those specs. Hopefully the design considerations would prevent anything getting near to "burning out"! \$\endgroup\$
    – Peter Bennett
    Commented Sep 25, 2014 at 0:08

3 Answers 3

Reset to default
1
\$\begingroup\$

As you may have gathered from the other responses, this is generally not a trivial exercise.

At the individual component level, each manufacturer publishes a datasheet that lists, among many other things, the max and min supply voltages and the expected current for at least one recommended supply voltage. Add up all the components on the same internal power bus (don't forget passives) and back-calculate the current going into each regulator at the voltage that feeds the regulator. (more datasheets) Add up all of those, back-calculating again through cascaded regulators, until you end up at the power input.

As for the input voltage selection, they may have chosen something just higher than the highest regulated voltage + dropout of that regulator. I suspect though, that a laptop adapter would be just enough to charge the battery and the rest of the computer then runs off the battery terminals, even if there's not actually a battery plugged in.

Most of the time, this process is much more iterative in design than in testing. There may be a total power budget to start with (probably specified in watts), then parts are chosen somewhat by experience to try and add up to less than the budget. If it's over budget, then some of the parts are substituted for more efficient or less capable ones and the total power is calculated again. Once the numbers work out, in many more ways than just power consumption, then it goes to the first prototype.

In some designs, the power budget is a major factor in the design, like an all-day netbook. In others, it's more of an afterthought so long as it can still be cooled adequately, like a high-quality architectural drafting engine that is used both in the office and on a job site.

Of course, this is grossly simplified, but you get the idea.

\$\endgroup\$
10
  • \$\begingroup\$ Hey AaronD, best answer I've seen so far. Going a little bit further, for the individual components, how does the manufacturer originally figure out the min and max supply voltages and expected current at one of those voltages? Is it the same as how you described in the experience and iterative process? \$\endgroup\$
    – edaniels
    Commented Sep 25, 2014 at 3:16
  • \$\begingroup\$ P.S. I'm not really designing anything but through my learning I noticed that many of these circuits described give you all the parameters you need but this isn't the case when you're designing something from scratch. I like to know how they got to that point even with a simple circuit with a single battery and resistor where the voltage of the battery hasn't been determined. \$\endgroup\$
    – edaniels
    Commented Sep 25, 2014 at 3:18
  • \$\begingroup\$ Yep, it's mostly the same process for the silicon manufacturer as well. As for the battery and resistor (electric heater for example, just to get some kind of practical application), there might be a target power instead of max or min, so then the voltage and resistance are related through a simple equation: P = V^2 / R Pick one, and you know the other. \$\endgroup\$
    – AaronD
    Commented Sep 25, 2014 at 4:06
  • \$\begingroup\$ By pick one you mean choose a value for either V or R and then you can solve for R or V respectively? After deciding on target power, how would you choose V or R? Same as before? \$\endgroup\$
    – edaniels
    Commented Sep 25, 2014 at 4:12
  • \$\begingroup\$ There are several considerations that factor into the choice of V and/or R. V too high, and it becomes unsafe. V too low, and another two equations, P = I x V and P = I^2 x R (I is current), combined with the resistance of the wires makes the wires heat up (the second P). If this leaves an acceptable range, it can be narrowed down further by what you or your customer has available already. If not, then you'll have to accept that one of them can't be met and modify it carefully. (some customers have difficulty understanding this, especially if a naive salesperson sells them the moon) \$\endgroup\$
    – AaronD
    Commented Sep 25, 2014 at 4:35
0
\$\begingroup\$

The computer was probably designed with a power budget of 60 watts. Most of this is consumed by the CPU, GPU, screen, and battery charger. They also decided that the input voltage from the adapter would be 16.5 volts, requiring 60W/16.5V ~= 3.65 amps. The power draw of the components can be measured under full load to ensure the power budget is met.

\$\endgroup\$
2
  • \$\begingroup\$ How did they decide the input voltage would be 16.5V? How do you test at full load if you don't know what voltage to use as input? \$\endgroup\$
    – edaniels
    Commented Sep 25, 2014 at 1:49
  • \$\begingroup\$ 16.5 volts is a tradeoff between voltage and current and is likely related to the battery voltage so the battery charger will be efficient. Lots of current means resistive losses in the power cable and it requires beefier connectors. The battery is probably a 4 cell lithium polymer, with 4.2 volts per cell resulting in 16.8 volts at full voltage. 5 cell batteries at 4.2 volts per cell are 21 volts. If the AC adapter voltage is just about the same as the battery voltage, it is quite simple to charge the battery though a simple series transistor and current sense resistor. \$\endgroup\$
    – alex.forencich
    Commented Sep 25, 2014 at 5:23
0
\$\begingroup\$

A computer is a very complex beast, with power management circuits that control the amount of power being fed to various subsystems.

A laptop computer is designed to limit the amount of power it consumes. That is both to reduce the drain on it's battery and to not exceed the heat dissipation limits of the device.

A laptop has one or more voltage regulators that convert battery voltage (or external power supply voltage) to the different voltages needed inside the computer. It will also have a charging circuit to charge the internal battery.

You would need to figure out the maximum current needed by the whole system when it was running at it's max (Battery charging circuit at maximum demand, screen at full brightness, disk drive spinning and seeking, DVD drive spinning and seeking, screen at full brightness, all cores on the CPU powered up and at maximum draw, graphics subsystem maxed out, fans running at full speed, etc. Plus power loss for the internal voltage regulators...) Then you'd want to add a small amount as a margin of error. Note that the power management subsystem might prevent all those things from happening at once. For example, it might slow down or stop charging the battery if all those other things were going on at once. It might throttle performance on the GPUs if they start to get too hot or exceed the total current limits on the power supply. Etc.

\$\endgroup\$
2
  • \$\begingroup\$ So in order to run the system at its max, would you just hook it up to an arbitrary power source at some low voltage and increase it from there to figure out what you need? What I'm saying is what do you need to actually run the system at its max without knowing the voltage? Iterative testing? \$\endgroup\$
    – edaniels
    Commented Sep 25, 2014 at 1:47
  • \$\begingroup\$ No! Computer equipment needs very specific voltages. As I said, it almost certainly has one or more internal voltage regulators. Those have specific ratings for their input voltage. You give the computer the input voltage required by it's power supply. \$\endgroup\$
    – Duncan C
    Commented Sep 25, 2014 at 11:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.