Linear regulators in series

[okw]

I'm helping a friend review a schematic. It's a battery powered project with three rails: battery (AAA 3x1.5V) and two linear regulated rails (LP5907MFX-1.8 and LP5907MFX-3.3). The 3.3V reg is powered directly from the battery, the 1.8V reg is powered from the 3.3V output. The design is done by a serious player, but I don't understand why they don't parallell them (both powered from battery)? Then the current capabilities of 3.3V isn't limited by the 1.8V draw. Could it be that 1.8V draws much more power, so in this series setup, some heat is dissipated in the 3.3V regulator before the rest is dissipated in the 1.8V regulator?
Any other pros/cons with series regulators?

[Phoenix]

I can think of a few.

Share the power loss.
Input ripple rejection, cascading PSRR.
To meet one's input voltage rating, especially very small low output regulator.

[wraper]

Depending on current consumption on rails placing vregs in series can even out power dissipation on both (reduce power dissipation on lower voltage vreg). It can make things better or worse depending on circumstances.

[jwet]

Here are a couple of more points-

Overall higher efficiency better battery utilization- since Eff ~ Vout/Vin, taking two long drops is less efficient than the cascade.

Power sequencing is guaranteed.  Even if not absolutely required, having a defined sequence is better than having a race between two regulators.

One other comment- not so positive
Getting 3.3v plus some dropout is not a very good supply coming from 3 AA's.  You won't be getting very good utilization of total battery capacity.  I would estimate 30% or more less life than a 4 cell design or one where you can discharge to a volt per cell. Potential solutions, run at 2.7V vs. 3.3 or use a Cuk or other buck/boost type for 3.3v from 3 AA's.

[wraper]

Overall higher efficiency better battery utilization- since Eff ~ Vout/Vin, taking two long drops is less efficient than the cascade.

The law of conservation of energy disagrees with you.

[okw]

The regulator dropout increases with increasing current consumption. So the first 3.3V gets a unnecessary high dropout if the 1.8V consumes a lot. Which in turn reduces the battery life?
Maybe not a problem as the specific 3.3V regulator has a max dropout of 200mV@250mA and AA cells are completely drained at 1.3V (3.9V and dropout is at about 3.5V).

[tunk]

... and AA cells are completely drained at 1.3V ...

Depends on what you use them for - I had an MP3-player which shutdown at 0.9V.
Some discharge curves: https://lygte-info.dk/review/batteries2012/Ikea%20AA%20UK.html
From the link above, at 250mA you will only use maybe 20% of the capacity.

[Kim Christensen]

Another pro-series point: If the 3.3V goes into shutdown unexpectedly then the 1.8V one would be powered down too. Depending on the circuitry, having the 1.8V powered while the 3.3V part of the circuit was dead might be a problem.

[David Hess]

They are doing that either for power sequencing or to better share power dissipation between the two regulators, or both.  There is nothing wrong with it.

[ConKbot]

One other comment- not so positive
Getting 3.3v plus some dropout is not a very good supply coming from 3 AA's.  You won't be getting very good utilization of total battery capacity.

That depends on the specs of the chips on the 3.3v rail. If you don't need a perfect 3.3v, just to cap it from receiving 4.5v from 3 fresh out of the pack alkalines, use a 3.3v reg and let it sit in dropout on the tail end of discharge. Assuming the ground current of the 3.3v regs stays reasonable in dropout, and doesn't increase a ton and run the batteries down fast.