Author Topic: Regulating LiPo battery to 3.3V  (Read 6636 times)

0 Members and 1 Guest are viewing this topic.

Offline goreTopic starter

  • Regular Contributor
  • *
  • Posts: 95
  • Country: gb
Regulating LiPo battery to 3.3V
« on: December 28, 2022, 06:08:10 pm »
Hello,

I'm building a portable device which runs on a 3.7V 1500mAh LiPo battery. I'm trying to make it as efficient as possible to prolong battery life, of course. Previously I used a LT1308 to boost it to 5V and then a MCP1603 to take it down to 3.3V. Thing is I don't even need 5V. Only do it to keep it stable as battery charge drops.

Is there a better solution in a single package, perhaps? Or a chip which can step it down from 4.2V directly to 3.3V, but I assume I'd run into stability issues even with a low dropout regulator (200mV, for example). The device draws about 100mA @ 3.7V constant. Any device supporting about 500mA would be nice to handle short spikes when transmitting.

I can look for parts on TME, Farnell, but I'm trying to figure out the correct way to do it first. Any advice I'd appreciate.
 

Online Peabody

  • Super Contributor
  • ***
  • Posts: 2161
  • Country: us
Re: Regulating LiPo battery to 3.3V
« Reply #1 on: December 28, 2022, 06:40:45 pm »
I think a very low dropout linear regulator would work fine.  By the time your battery gets to the 3.5V dropout level, it will be almost fully discharged anyway.  But even then, most of these regulators will just follow the voltage on down.  You lose regulation, but that shouldn't matter much with a battery supply.

One option might be to go with a 3.0V regulator instead of 3.3V if your devices can handle that.  They will use less current at 3V than they do at 3.3V even at the same processor speed.

I really don't know what's available in the way of buck converters that would convert directly from 3.7V to 3.3V.  I think they need more headroom than that.  But others may know of options along those lines.

Are you able to put your processor to sleep?   That's often the biggest contributor to extending battery life, and that's where linear regulators really shine.
 

Offline goreTopic starter

  • Regular Contributor
  • *
  • Posts: 95
  • Country: gb
Re: Regulating LiPo battery to 3.3V
« Reply #2 on: December 28, 2022, 06:54:57 pm »
Are you able to put your processor to sleep?   That's often the biggest contributor to extending battery life, and that's where linear regulators really shine.

Unfortunately It needs to run all the time. Processing data, driving LCD user interface etc. I've already disabled most the peripherals I can afford to lose. Radio circuit gets shut off as well when not in use, which reduces power consumption considerably.
 

Offline Miti

  • Super Contributor
  • ***
  • Posts: 1357
  • Country: ca
Re: Regulating LiPo battery to 3.3V
« Reply #3 on: December 28, 2022, 08:53:23 pm »
I use MCP1703 in one of my projects, but it only draws few uA in sleep and few mA active. Take a look.
Fear does not stop death, it stops life.
 

Online thm_w

  • Super Contributor
  • ***
  • Posts: 7211
  • Country: ca
  • Non-expert
Profile -> Modify profile -> Look and Layout ->  Don't show users' signatures
 

Online Peabody

  • Super Contributor
  • ***
  • Posts: 2161
  • Country: us
Re: Regulating LiPo battery to 3.3V
« Reply #5 on: December 28, 2022, 10:59:04 pm »
I think pinning down your maximum current requirement may be important in selecting an LDO.  It appears that the dropout voltage typically rises as the maximum current goes up.  But it might be possible to use two or three ganged MCP1700s to provide enough current for everything while maintaining the 178mV dropout voltage.  Or maybe use a second regulator just for the radio.
 

Offline Karel

  • Super Contributor
  • ***
  • Posts: 2267
  • Country: 00
Re: Regulating LiPo battery to 3.3V
« Reply #6 on: December 29, 2022, 12:13:05 am »
Here you go: https://www.ti.com/product/TPS63051

Real buck or boost with seamless transition between buck and boost mode
2.5 V to 5.5 V Input voltage range
0.5-A Continuous output current: V IN ≥ 2.5 V, VOUT = 3.3 V
Adjustable and fixed output voltage version
Efficiency > 90% in boost mode and > 95% in buck mode


This way you can discharge the battery completely (which means down to 3V).
« Last Edit: December 29, 2022, 12:15:18 am by Karel »
 

Offline AndyBeez

  • Frequent Contributor
  • **
  • Posts: 858
  • Country: nu
Re: Regulating LiPo battery to 3.3V
« Reply #7 on: December 29, 2022, 12:20:47 am »
Don't forget LiFePO4 batteries which run at a nominal 3.2 volts in a working range of 3.0V to 3.3V

You could be efficient by removing the buck/boost circuit altogether - unless your system is 3.300 volts dependent.
 

Online SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15419
  • Country: fr
Re: Regulating LiPo battery to 3.3V
« Reply #8 on: December 29, 2022, 12:43:06 am »
Check your design, and chances are it can run at 3V or even slightly lower. Regulate to that. Done.
3.3V is not written in stone. Many if not most digital ICs these days have a largish operating range.
Think outside the box. So to speak.
 

Offline goreTopic starter

  • Regular Contributor
  • *
  • Posts: 95
  • Country: gb
Re: Regulating LiPo battery to 3.3V
« Reply #9 on: December 29, 2022, 10:37:51 am »
Great suggestions so far. I'll see how stable the system is on 3.0V and take it from there. If it works, a single 3.0V LDO it is then.
 

Online Peabody

  • Super Contributor
  • ***
  • Posts: 2161
  • Country: us
Re: Regulating LiPo battery to 3.3V
« Reply #10 on: December 29, 2022, 03:52:41 pm »
Here you go: https://www.ti.com/product/TPS63051

Is there anything like this in a package that would be useable with a breadboard?  Is there a VQFN-to-SIP adapter board?
 

Offline Karel

  • Super Contributor
  • ***
  • Posts: 2267
  • Country: 00
Re: Regulating LiPo battery to 3.3V
« Reply #11 on: December 29, 2022, 04:05:52 pm »
Here you go: https://www.ti.com/product/TPS63051

Is there anything like this in a package that would be useable with a breadboard?  Is there a VQFN-to-SIP adapter board?

I don't believe that's a practical option because the layout for that chip is critical.
For a breadboard it's better to use an LDO.
 

Offline bidrohini

  • Regular Contributor
  • *
  • Posts: 201
  • Country: bd
Re: Regulating LiPo battery to 3.3V
« Reply #12 on: December 31, 2022, 03:56:02 pm »
 

Online iMo

  • Super Contributor
  • ***
  • Posts: 5244
  • Country: bj
Re: Regulating LiPo battery to 3.3V
« Reply #13 on: December 31, 2022, 04:26:45 pm »
I used to use the MAX1724 (5pin sot23, 2caps, 1 inductor), it works from 0.6-0.7V input up, with fixed output for 3.0/3.3/5.0V ..
Readers discretion is advised..
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 8833
  • Country: fi
Re: Regulating LiPo battery to 3.3V
« Reply #14 on: December 31, 2022, 04:27:31 pm »
Do you really need voltage regulation?

If possible, use devices which can be supplied by up to 4.2V.

If not possible, e.g. most 3V3 microcontrollers with 3.6V recommended maximum, choose components which can run at tad below 3.3V (like 2.8V; most 3V3 parts can), then pick a very low dropout linear regulator (preferably: one where saturated/bypass operation is well defined, but most are fine anyway), and just let the output voltage drop, losing voltage regulation during the last 20-30% of charge or so.
 

Offline goreTopic starter

  • Regular Contributor
  • *
  • Posts: 95
  • Country: gb
Re: Regulating LiPo battery to 3.3V
« Reply #15 on: January 01, 2023, 12:18:14 am »
I can't take the direct battery supply route. Most devices on board are somewhere between 2.7 - 3.6V. So far I've tested the entire system on 3.0V and it's been running without issues for a couple of days. Going from 3.3 to 3.0V saves me a decent amount of energy as well (reading about 16% less power draw). Some of that has to do with a slight drop in LCD backlight. Screen is plenty bright though, so not a problem.

So far I've picked out a few very low drop LDO's and some devices suggested in the thread. They are good for 200mV or even less, which allows me to use up most of the charge with a stable output. Battery's getting cut @ 3.3V.

Example candidate:

https://www.tme.eu/Document/1985cfc4e26f4f29df9db32d8f1b1818/MCP1825-3002EAT.pdf
« Last Edit: January 01, 2023, 12:22:55 am by gore »
 

Online Peabody

  • Super Contributor
  • ***
  • Posts: 2161
  • Country: us
Re: Regulating LiPo battery to 3.3V
« Reply #16 on: January 01, 2023, 04:48:37 am »
It looks like that should work very well.  So is there no buck regulator solution in a usable package?
 

Online mariush

  • Super Contributor
  • ***
  • Posts: 5141
  • Country: ro
  • .
Re: Regulating LiPo battery to 3.3V
« Reply #17 on: January 01, 2023, 06:56:34 am »
If you want long battery life, a step-down/buck regulator can be more efficient than a linear regulator.
But few breadboard friendly regulators will actually be better than ldos, because regulators that run at high frequencies don't like long pins and the breadboard contacts. But you could make a tiny pcb at jlcpcb or somewhere and put the regulator chip and the components it needs on that pcb.
For example a chip that looks like it would do this 3.7..4.2v to 3..3.3v is this one, AP61100 : https://www.digikey.com/en/products/detail/diodes-incorporated/AP61100Z6-7/11696569
It's easy to use, it only requires a small inductor, a couple ceramic capacitors and a couple resistors to set the output voltage, but the components have to be quite close to the chip.
The chip can reach 90% efficiency which is better than a ldo (ex 3.7v in, 3v out = 81% efficiency) and the quiescent current (idle power consumption) is very low at 15uA

Here's a second example ... a slightly less efficient chip but easier package to solder by hand: https://www.digikey.com/en/products/detail/diodes-incorporated/AP3418KTR-G1/4470871

« Last Edit: January 01, 2023, 07:01:55 am by mariush »
 

Online Peabody

  • Super Contributor
  • ***
  • Posts: 2161
  • Country: us
Re: Regulating LiPo battery to 3.3V
« Reply #18 on: January 01, 2023, 07:23:52 pm »
I was hoping a module maker would deal with the layout issues.  But I don't see any buck converter modules that convert LIPO voltage to 3.3V.  I would have thought that if the switching oscillator could go to a 100% duty cycle, headroom wouldn't be an issue.  But apparently that's not the case.

Incidentally, what specs in the datasheets of those two Diodes Inc. buck converters you linked to would tell me what their dropout voltage or minimum headroom is?
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 8833
  • Country: fi
Re: Regulating LiPo battery to 3.3V
« Reply #19 on: January 01, 2023, 07:33:23 pm »
I would have thought that if the switching oscillator could go to a 100% duty cycle, headroom wouldn't be an issue.  But apparently that's not the case.

There are switcher ICs that can go to 100% duty operation, but it rules out any that uses N-channel switch with bootstrap gate drive supply. P-MOSFET based switchers usually can go to 100%. Maximum duty is mentioned in the datasheet.

If you pick simple, highly integrated (low pin count) parts, layout is not that difficult if you just follow the example strictly.
« Last Edit: January 01, 2023, 07:41:24 pm by Siwastaja »
 

Online Peabody

  • Super Contributor
  • ***
  • Posts: 2161
  • Country: us
Re: Regulating LiPo battery to 3.3V
« Reply #20 on: January 01, 2023, 08:04:38 pm »
I think this is what I was looking for:

https://www.adafruit.com/product/2745

It says it will output 3.3V with input voltage as low as 3.5V.  But I see nothing like this on Aliexpress or Ebay.  Perhaps I'm not searching for the right thing.  Or perhaps it's just not enough better than an LDO to be worthwhile, or not better at all for projects that sleep most of the time.


 

Online mariush

  • Super Contributor
  • ***
  • Posts: 5141
  • Country: ro
  • .
Re: Regulating LiPo battery to 3.3V1
« Reply #21 on: January 01, 2023, 08:20:08 pm »
For the AP61100, you can look in the datasheet: https://www.diodes.com/assets/Datasheets/AP61100-AP61102.pdf

It should work fine at close to 100% duty cycle though.

For example, see page 14 where it explains how you can determine the Ton period which like Toff must be minimum 70ns :

The AP61100/AP61102 device is a 2.3V-to-5.5V input, 1A output, fully integrated synchronous buck converter. Refer to the block diagram in
Figure 5. The device employs constant on-time control to provide fast transient response and easy loop stabilization. At the beginning of each
cycle, the one-shot pulse turns on the high-side power MOSFET, Q1, for a fixed on-time, tON. This one-shot on-pulse timing is calculated by the
converter’s input voltage and output voltage to maintain a pseudo-fixed frequency over the input voltage range.

Ton  = Vout / ( Vin x Fsw) 

where

 VIN is the input voltage
 VOUT is the output voltage
 fSW is the switching frequency
The off-time duration is tOFF and starts after the on-time expires. The off-time expires when the feedback voltage decreases below the reference voltage, which then triggers the on-time duration to start again. The minimum off-time is 70ns typical.

See Figure 15 on page 8 for  switching frequency vs load  - in your case at around 0.1A you're looking at around 500 kHz switching frequency.

So for example let's say 3.1v in, 3.0v out, 500 khz switching frequency ...
Ton = 3 / 3.1 x 500,000 = 1.935e-6 or 1935 nanoseconds

Page 19 has example of pcb layout  - it's only missing the connection to enable pin, which could be permanently connected to Vin

As for the second regulator I suggested, the datasheet says it can do 100% duty cycle, so no worries.
 

Online SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15419
  • Country: fr
Re: Regulating LiPo battery to 3.3V
« Reply #22 on: January 01, 2023, 08:26:40 pm »
In many cases, lowering the operating voltage and using a LDO is the simplest, cheapest and even sometimes more efficient approach (due to the fact that buck or buck/boost converters often have a quiescent current much higher than the low Iq LDOs that are ubiquitous these days.) Just do the maths and decide. No way around it.
 

Online Doctorandus_P

  • Super Contributor
  • ***
  • Posts: 3891
  • Country: nl
Re: Regulating LiPo battery to 3.3V
« Reply #23 on: January 02, 2023, 01:22:54 am »
I'm wondering why the device needs 100mA continuously to run. You give very little information about your device, but the way to get power consumption down is to analyze each part carefully.

For real low power devices the quiescent current of a voltage regulator can be a significant factor, but these days a few handfuls of uA is enough to run an SMPS. The TPS6305x mentioned earlier has apparently a 65uA quiescent current. Microcontrollers can also run on less than 1mA/MHz, so where is that other 99mA going? Radio's and LCD's can be very power hungry. Radio's typically run at a very low duty cycle (less than 0.1% of the time on) If the receiver has to be on, then you could revise your protocol. for example, a mains powered transmitter is always on, and your battery powered device only checks in when needed.

With LCD's, the backlight is often a power hungry thing, but there are a bunch of technologies that can suck less power out of your battery. A monochrome reflective LCD is an obvious choice, maybe E-paper is fit for your application. The Sharp Memory LCD's are also crisp and need very little power. If you need your display to emit light, maybe an O-led display is suitable, combined with turning on as few pixels as possible.

Software design is also an important factor. Writing "arduino" code and wasting cpu cycles in delay loops is just plain horrible.

But to me, trying to extends the battery life of a device that sucks 100mA and then only asking a question about some voltage regulator is very odd at least.
 

Online Doctorandus_P

  • Super Contributor
  • ***
  • Posts: 3891
  • Country: nl
Re: Regulating LiPo battery to 3.3V
« Reply #24 on: January 02, 2023, 01:30:12 am »
Screen is plenty bright though, so not a problem.

What sort of shortsighted attitude is that?
Of course it is a problem.
If you want to extend battery life, turn down the backlight as far as possible. Every emitted lumen sucks the battery dry.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf