I am designing a 4 layer board (2 signal, ground, and power), and is there any reason not to have top layer power plane, middle layers signals, and bottom layer a ground plane? I am concerned about heat dissipation as about 1/2 of my power plane is broken up into about 10 medium squares with current in the neighborhood of 5 to 10 A @ 15 V for 100 - 500 msec pulses with a few minutes in between pulses flowing across the planes. I know the common stack up is signal, ground, power, signal. Also, I have all THT components and no SMD.

The highest speed signal is a 2.4 MHz SPI bus with data every few minutes - program a DAC from a Raspberry Pi to produce the current pulse of a certain level and duration. Everything else is slow (LED on or off) and mostly just on or off 3.3V logic signals. EMI is not a factor - the device sits by itself in a large field outside and far away from buildings or any other electronics. Power is from a battery pack. The purpose of the device is to ignite rocket engines from a distance over a local WiFi connection to the Raspberry Pi.

In the attached image, the high current power planes are at the top and the logic/raspberry Pi circuitry is at the bottom. The large plane at the bottom is for 3.3V.

The question is whether the high current planes should be on the surface of the board or can they be buried on an inner plane. I am a proponent of signal traces on the top and bottom to facilitate post pcb production changes to the circuitry (the oops factor in design). But, I am also worried about heat dissipation, hence my question. enter image description here

  • \$\begingroup\$ Do you have high speed signals requiring impedance control? \$\endgroup\$
    – Justme
    Oct 31, 2023 at 5:35
  • \$\begingroup\$ What general role is your PCB? Is EMI a relevant topic? Do you have RF-components? \$\endgroup\$
    – S_G
    Oct 31, 2023 at 6:49
  • \$\begingroup\$ I updated the original question to answer your questions and give a more complete description of the problem. My apologies for not being thorough in my original question. \$\endgroup\$ Oct 31, 2023 at 14:12

1 Answer 1


Power dissipation depends on the size of the squares, but more importantly, on the size of the trace connecting these squares. If you are worried, I don't see anything wrong by using the external bottom layer for power. Another consideration is the possibility, even remote, to accidentally have 10A permanently across the board instead of pulses. Design your board accordingly.


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.