1
\$\begingroup\$

I'm a designing a circuit that I'll prototype soon.

I'm curious about how should I determine my power budget for the circuit? Could you guys please suggest me how to do it? Should I calculate each power rail Watt and sum it? Do you have any suggestions for 'excel' table for this?

I would like to hear suggestions please or to know how the pros do it.

I assume that power budget estimation is important because according to this estimation I choose, for example, a voltage regulator. according to the power estimation (W) of rail - the power shouldn't deviate the power dissipation of the voltage regulator - Am I right?

Thank you very much, also forgive me about the newbish question :)

\$\endgroup\$
  • \$\begingroup\$ You have to do that multiple times, for each part of the circuit. Because heat dissipation and current distribution over the board is always important. One you have those numbers, adding them up is only little work. \$\endgroup\$ – Janka Jun 15 '18 at 20:52
  • \$\begingroup\$ Consider charge surge currents, power dissipation, heat rise, thermal resistance from inefficient linear regulator, consider buying SMPS (cheap) or use PC ATX PSU. Why make it hard? \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Jun 15 '18 at 21:34
0
\$\begingroup\$

There are a few ways to do this:
1) Predict and model the design. This is done a few ways. If it's a circuit, simulating each of the pieces gives a nice idea of how much it will use. A spice package such as LT Spice can be great for this (if you hover over a part in LT spice an hold down alt it can graph out a power for that component). Estimation and bounding goes a long ways, usually pouring through datasheets and finding quiescent currents (min current) and max currents. Then estimate something in between. With microprocessors it is much more difficult to determine how much power they will draw, so a max case is usually best.

Estimate power supply efficiencies and DC to DC converters, get a current draw for the total design. It's better to give yourself some margin (like add 30%) to make sure you size a supply right and don't overdraw current.

2) The next way is to actually get some parts with prototyping and measure the currents to give you an idea of how much current the total design will draw.

\$\endgroup\$
  • \$\begingroup\$ I believe using a simulator is not going to give any usable result. Most models use controlled generators which "do things" without loading power rails though. Educated guess of worst cases is often straight forward instead. \$\endgroup\$ – carloc Jun 16 '18 at 6:39
  • \$\begingroup\$ Funny, I used lt spice to estimate the power for the analog instruments of a satellite, and it did wonderfully, close to the measures after we built the design. It's a tool and if you know it's weaknesses and strengths you can get good results \$\endgroup\$ – Voltage Spike Jun 17 '18 at 4:11

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.