Buck vs Buck-Boost Switching Voltage Regulators in low power desings

[Nikos A.]

Hi guys,

I design an ultra-low-power battery-operated hardware and I have concerns about whether I have to use a buck or buck-boost switching regulator to supply the system. The power source will be probably a LiOn or LiPo 3.7V battery and the system will operate at 3.3V. I was considering using a buck converter but I've been told that a buck-boost converter will increase the battery lifecycle. Our goal is for the battery to last as long as possible.

Do you have any experience with that?

Thanks

[Nikos A.]

Thanks evb149 for your reply!!

Another question would be if you really NEED 3.3V operation or if you can allow circuit operation down to 3.0V or 2.8V in which case you don't likely need a boosting function at all and your system will use even less energy.

This is a LoRa based project. My design integrates the LoRa E5 module that is based on the STM32WLE5J8 SOC.
https://wiki.seeedstudio.com/LoRa-E5_STM32WLE5JC_Module/
chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/viewer.html?pdfurl=https%3A%2F%2Ffiles.seeedstudio.com%2Fproducts%2F317990687%2Fres%2FSTM32WLE5JC%2520Datasheet.pdf&clen=2749497&chunk=true

The STM32WLE5J8 has an input power supply between 1.8V - 3.6V. Also, I will not integrate a charging system since my intention is for the battery to last for a couple of years (2-3). Then the user will replace the battery. Operating the system at 1.8V requires just a buck converter.

In general, if you had the option to operate the system either at 3.3V or 1.8V with a non-rechargeable battery, what you would choose in order to increase the battery lifetime?

[Ian.M]

You'll loose up to 5dB RF power operating at 1.8V (vs 3.3V) which will approximately halve the max. range.

[Nikos A.]

You'll loose up to 5dB RF power operating at 1.8V (vs 3.3V) which will approximately halve the max. range.

Thanks, I didn't spot that. Does the datasheet mention that kind of information?

[Ian.M]

Yep.  I looked at the transmit power graphs in it.

[Siwastaja]

Even 3.3V nominal is likely "in spec" at 3.2V or even 3.1V, look at the datasheet.

Empty li-ion open-circuit voltage is around 3.4V. If load current is low (look at maximum peak, though; capacitors can only supply tens of microseconds, not milliseconds), like C/20 or lower, actual under-load voltage is close to that open circuit voltage.

This means, if you pick the downconverting regulator of lowest possible dropout (linear LDO with, say, 200mV of dropout, they do exist; or buck with highest possible duty cycle you can find, PFET-based can even go to 100%), you won't be losing any battery capacity.

If it is an absolute requirement that 3.3V rail must not drop even to 3.2V, then even with very low dropout, you might be losing some 5-10% of battery capacity. This isn't a bad deal to make, usually.

Buck-boost or SEPIC can fully utilize the charge, but remember, in low power sensor designs, it's always the quiescent current which dominates, so pick the regulator accordingly.

[Psi]

Another option is a buck regulator with bypass feature.

It steps down, but once the input voltage is the same as the configure output it just connects them together.
So you get 3.3V output from lion cell 4.2V down to 3.3V then the output follows the battery voltage.
Most 3.3V devices run fine on 3.0V and below that your lipo is dead anyway.

It just saves needing to have two inductors for buck+boost.

[Siwastaja]

Another option is a buck regulator with bypass feature.

Yes, this is basically synonymous to 100% duty cycle.

Regulators that use high-side N-type MOSFETs (and as a result, an external bootstrap capacitor), won't support that unless they come with an additional charge pump. PFET-based buck regulators types support 100% duty if this isn't specifically prevented by the control logic for some reason.

[Psi]

Yeah, ya just need to read the datasheet and make sure it can do bypass and that it switches it on/off automatically. (The feature might be called something else depending on manufacturer)

I have seen some regs that don't do it automatically, they have an input pin to activate bypass. Like when EN pin is high switchmode is on, and when low bypass is on.
So with those ones you need to have something external to control it, those ones might not be suitable for your application.
Those ones are for when you have a MCU that can operate from any battery voltage that will be present but sometimes the MCU turns on an external peripheral with more fixed voltage requirements.
Usually they are boost controllers though, not buck.

[Nikos A.]

Thanks for your inputs guys!! I've used buck regulators with bypass feature in the past. The problem I see now is that if the system's voltage follows the battery voltage then I lose transmitting power while the battery's voltage reduces. So probably I will stick with a buck-boost converter with low quiescent current.

[Siwastaja]

Datasheet seems to state that high power (+22dBm) mode is not available at 1.8V, but that doesn't mean radio is forced into lower power modes or high power mode becomes nonfunctional if you deviate a bit from 3.3V.

Radio current consumption given as high as 120mA, the size of the li-ion cell we are talking about is critical. For example, if it is a 120mAh cell, we are talking about 1C discharge causing significant voltage drop in the cell near empty, especially in cold temperatures; so even if the cell is still at 50% charge, it could drop way below 3.3V during TX peaks. OTOH, if it is a 2400mAh cell, then this is no big deal, and at 3.3V under load, the cell is already empty.

[Picuino]

And what about a SEPIC converter?
https://en.wikipedia.org/wiki/Single-ended_primary-inductor_converter

"The single-ended primary-inductor converter (SEPIC) is a type of DC/DC converter that allows the electrical potential (voltage) at its output to be greater than, less than, or equal to that at its input. The output of the SEPIC is controlled by the duty cycle of the control switch (S1).

A SEPIC is essentially a boost converter followed by an inverted buck-boost converter, therefore it is similar to a traditional buck-boost converter, but has advantages of having non-inverted output (the output has the same voltage polarity as the input), using a series capacitor to couple energy from the input to the output (and thus can respond more gracefully to a short-circuit output), and being capable of true shutdown: when the switch S1 is turned off enough, the output (V0) drops to 0 V, following a fairly hefty transient dump of charge.[1]

SEPICs are useful in applications in which a battery voltage can be above and below that of the regulator's intended output. For example, a single lithium ion battery typically discharges from 4.2 volts to 3 volts; if other components require 3.3 volts, then the SEPIC would be effective. "

[thm_w]

Thanks for your inputs guys!! I've used buck regulators with bypass feature in the past. The problem I see now is that if the system's voltage follows the battery voltage then I lose transmitting power while the battery's voltage reduces. So probably I will stick with a buck-boost converter with low quiescent current.

As mentioned by Siwastaja/Psi, battery is likely dead at that point, so why would it matter?
Have the MCU monitor battery voltage and stop transmitting.

https://www.richtek.com/Design%20Support/Technical%20Document/AN024

[Psi]

The range difference between 3.3V and 3.0V will be pretty small.  I doubt you would ever notice it.