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

User avatar
By Caio.Viana
#80140 Hello, in my first post, I have been consulting the opinions of the most experienced users of esp8266 regarding the development of projects.
I have a project where I use an esp 8266 to send certain data over the wifi every time a button is pressed. the design will be powered by a lithium battery so energy savings become essential. so the sp8266 is only fed when the button is pressed and does not consume any current when the button is not pressed. the lithium battery has a voltage of 3.7v and reaches 4.2v when fully charged so it becomes necessary to use a means to reduce the voltage of the battery to 3.3v using one of two methods:

linear regulator: where the input current and the same of the output and the voltage and dissipated in the same, has low efficiency but low cost, also has a low dropout of voltage and low quiescent of current

buck converter: high efficiency where for higher input voltages the input current becomes lower, very high cost, low dropout of tesnao and high quiescent current

That said, I would like to ask for your opinion as to which method to use for the longer battery life and low cost
User avatar
By btidey
#80148 There is no one answer to this as a lot depends on how the app is being used.

There are basically two contributions to the average current drawn.

The quiescent current when nothing is happening. Typically this is the sleep current of the electronics plus the quiescent current of the regulator.

The current once an event occurs and is being processed. Typically this will be the average operating current of the esp8266, any peripheral current, plus the current of the regulator.

How these contribute to the overall average current is determined by the typical interval between events and how long an event takes to process.

Now a decent buck converter has an efficiency advantage over a linear regulator. In the linear case the battery current is the same as the 3.3V current whereas in the buck case it will be reduced by the conversion factor between the battery and the efficiency of the converter.

For some applications this conversion factor gives the buck a considerable advantage and when the battery voltage is 5V or 12V this makes a big difference. For a single LIPO battery the advantage is much reduced as the average input voltage over the battery life cycle is around 3.75V. So with an efficiency of say 0.95 the conversion factor is about 1.08

When I was looking int this a while ago I put together a simple spreadsheet to estimate average current.

BatteryESP8266.jpg


This shows a good LDO linear regulator with a low quiescent and two buck example scenarios. It is actually quite difficult to find buck regulators which have both low quiescent current, high efficiency and the ability to handle >300mA output current spikes which can occur with the esp8266.

The factors are the quiescent currents of the regulator and the 'sleeping' electronics, the operating current of the electronics, the time spent processing an event, and the average interval between events.

The buck examples enjoy a small advantage when the events are more frequent, but the lower quiescent current of the linear starts to win as the interval increases.

Now if we can find a really low quiescent current buck with the right other characteristics then that would win under most circumstances. Similarly if one had external low power electronics that detected the events and only switched the battery source through when required then again the buck would win.

In most of my battery powered apps the interval are quite long and I want to keep external electronics to the minimum so I normally choose the linear regulator methodology.

It is also worth pointing out that if the application allows for a rapid check for action before initiating wifi then this can reduce the event operation time in a lot of cases and reduce the average current further.
You do not have the required permissions to view the files attached to this post.
User avatar
By eriksl
#80962
Caio.Viana wrote:thanks for the tips, I will opt for the linear regulator because the activations are not frequent and the low current consumption without load of the regulator


I think if you'd use a combined converter (buck/boost, they're actually two circuits combined into one), you'd be able to use the ESP8266 until the battery is really empty. The buck/boost converter would step down the voltage if it's too high and step it up if it's too low. You could configure it to always convert to 5 V and then add a linear converter to 3.3 V, which would make your power supply rock solid (and the ESP8266 needs that always run without any problems).