Use this forum to chat about hardware specific topics for the ESP8266 (peripherals, memory, clocks, JTAG, programming)

User avatar
By btidey
#83644 The quiescent current of that device can be confusing in some specs as it can refer to the current consumed when the device is disabled whereas some other regulators have the quiescent current as the current drawn when there is no load.

The ground current refers to the current coming out of the ground pin, i.e. current not going to the load. So the ground current under no load is equivalent to quiescent current on other devices.

The MaxLinear version of the spec for example has quiescent current as 90uA in the headline but 3uA in the table. The latter is for a disabled device. The ground current is 90uA for low loads and is therefore the equivalent of no load quiescent current.

For ultra low consumption with this device one would really need to disable the device under sleep conditions which adds extra complexity to the circuitry. One can achieve low quiescent current without this need in other devices.
User avatar
By davydnorris
#84082 I've been doing a lot of analysis on my battery powered devices and have come to a few realisations:

- using LDOs for low power solutions is not ideal if your supply voltage is much higher than your output.
The LDO is essentially pulling the current of the output side of the circuit (plus its own current, which will be small for a good LDO), so on the supply side that current is being pulled at the supply voltage. That means you're burning more power when your batteries are fresh or at full charge, and this is wasted as heat from the LDO. It's a weird thing when you power your circuit with 3 x 1.5V dry cells in series and it lasts less time than when you use 2 of them! Watching the power supply drain from 4.5V to 3.3V in days, and then sit at 3V-ish for the next several weeks is so frustrating!

- most of the ESP8266 modules have a minimum voltage of 2.6V, so if the rest of your circuit can work at lower voltage you can drop the supply to something like 2.8V and they will still work fine. This means a 15% power saving

- a switching regulator may be noisier than an LDO, and requires more external parts, but it makes the best use of your power. A good buck converter with proper filtering will be the most efficient use of any battery based power supply

- buck/boost converters look like a good plan, but fall down when you actually put them in place, and especially so when you consider dropping your voltage to 2.8V. Using a 3.3V buck/boost gives you virtually no extra time on battery because once you get below 3.3V the current draw starts to increase, draining your battery even faster. When most rechargeable batteries have to be cut off at about 3.0V or even higher, and most non-rechargeable cells will be sucked dry not long after, there is very little point in the boost side of the circuit.

As a result of these tests, I'm now dropping my circuit to 2.8V and replacing my regulators with 2.8V buck converters. If the price is comparable, I may just swap out my 3.3V buck/boost for a 2.8V buck/boost in that circuit, but will be running buck only for my main designs.

Would love to hear everyone's thoughts on this!