I'm working on a project that is battery powered where run-time is an important consideration. The project has two distinct required voltage levels, both served from a single battery:

  1. 6.4V which powers the main load - draw is anywhere from 500mA to 5A (very rare, short-term load for anything over 2A)
  2. 3.3V which powers all the control IC's and the screen, nominal draw is about 1200mA, not much variation

Current design uses a 6.4V (nominal) 6.6Ah (LiFePO4) battery with a switching regulator to get to logic level and gives okay life. Space is a constraint to add more battery capacity.

It occurred to me that something like a 20V, 5Ah tool battery which are all over might be interesting to consider. I would have to add another switching regulator to get from battery voltage to load voltage, then keep the same existing switching regulator to get from load voltage to logic level. I'm not too concerned about being able to dissipate the heat generated from losses.

The crux of my question is this:

From an energy-density perspective, the 20V, 5Ah battery can generate (nominally) 100 Wh of energy where our current battery can only generate about 42 Wh. So, even with a 15 or 20% loss from the switching regulator, it seems like I could still get almost double the energy from the higher voltage battery - and as far as I can tell - double the life for the same power draw.

This makes sense to me, but it seems if it was true then other things (like phones) where battery longevity is important would utilize this concept. Maybe they just can't afford to dissipate the heat, or maybe there are products using it and I'm just unaware of them.

EDIT1: Corrected "power" to "energy" when discussing Wh

  • 1
    \$\begingroup\$ LiPo cells used in smart phones can have significantly greater energy density than your typical LiFePO4. Wikipedia claims around 220Wh/l for LiDePO4 vs up to 730Wh/l for LiPo. It's not the higher voltage battery giving you a better energy density - it's the battery technology. \$\endgroup\$
    – brhans
    Commented Jul 17, 2019 at 14:36
  • 1
    \$\begingroup\$ 1.2A seems rather a lot for the IC/screen component, presumably most of that is backlight? \$\endgroup\$
    – pjc50
    Commented Jul 17, 2019 at 14:45
  • \$\begingroup\$ Just to be pedantic, a Wh is a measure of energy, not power. \$\endgroup\$ Commented Jul 17, 2019 at 15:14
  • \$\begingroup\$ @pjc50 Yes, most of that is backlight on a 7" TFT. \$\endgroup\$
    – meshtron
    Commented Jul 17, 2019 at 15:17
  • 1
    \$\begingroup\$ @Elliot Anderson I suspected I was using the wrong term, I am quite satisfied if that's the worst mistake in my question. \$\endgroup\$
    – meshtron
    Commented Jul 17, 2019 at 15:17

1 Answer 1


Answering to let this get archived. As noted by @brhans, the power density of the battery is dependent on battery chemistry/construction, not other factors.

Therefore, for a given battery chemistry, the most efficient battery pack would be one which provides voltage closest to the required input voltage.


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.