3.3 volt supply from single L-Ion cell

[kenarton]

Hello,

I'm an electronics hobbyist, and I want to use a single 18650 Li-Ion cell to power a fixed 3.3 volt supply.
The cell voltage varies from about 4.3v down to 2.7v, so I need to use a buck-boost DC-DC converter which operates smoothly through buck mode to boost mode.  The TI TPS63000 series is ideal as it is specifically designed for this application.  Unfortunately it seems to be available only in a 3mm x 3mm leadless QFN package. 
I am quite used to SM soldering, but I reckon this package would be beyond my capabilities.
There are lots of other buck-boost and step-up step-down converters on the market (eg TI MC33063), but all the data sheets seem to indicate that they can be configured to operate in one mode or the other, but not both.

Can anyone suggest an alternative equivalent to the TPS63000 in a manageable package, eg DIL or SOIC ?

I could use two separate converters with circuitry to switch over when the battery voltage reaches 3.3.  But I really want the simplicity of a single IC.

Thanks.

[Siwastaja]

Are you pulling significant currents out of the cell? I.e., what is the highest peak (> 1ms) current in your application?

If you stay below say C/20 as is often the case with small gadgets, the end-of-discharge voltage is way higher than what is rated for the typical C/2 or C/5 discharge cutoff in the datasheet.

0% SoC Open-circuit voltage being some 3.4V for most typical li-ion cells, an LDO with dropout around 100mV-200mV would not necessary waste more than some 5-10% of cell capacity.

Switch mode converter likely has higher quiescent current loss. Switch mode converter also requires proper layout, in which case soldering the QFN isn't a big extra deal IMHO. I didn't find it hassle, there was a time I was in the same situation as you are now and just bought the QFN packaged components I needed and a $15 Chinese hot air Hakko clone (fix the broken protective earth first!) and had no issues.

[fcb]

Use a boost regulator set to boost to say 3.6V and then a linear regulator (LDO, low-drop out) to drop down to 3.3V

Not as efficient as buck-bost but can work really well.  The boost-converter will not operate until the battery gets to say 3.7-3.8V (depending on Vfwd of the boost diode).  Lots of choices for user friendly pagakged boost regulators.

 

[Siwastaja]

Also what is the actual voltage range the "3.3V" devices are specified to work with?

Most 3V3 parts are designed to work with some range like 3.0V to 3.6V, many even wider. In such case, having a very low-drop out linear regulator go out of regulation and drop the line to maybe 3.1V with 3.2V cell voltage could be fully acceptable.

[NiHaoMike]

You could also consider LiFePO4, which its 3-3.6V range is an exact match for 3.3V logic.

[georges80]

I use a Richtek RT6150A to buck/boost a solar charged lifepo4 cell to 3.3V to power an STM uC and a Lora radio. The RT6150A is quite efficient, high frequency -> small inductor/caps etc and also very low quiescent current.

You will find it difficult to find new components in dip/dil etc. Time to learn how to use solderpaste and a hot plate, it's actually quite easy and opens up the world of component choices.

cheers,
george.

[rfclown]

I suggest that you measure one of your Li-ion cells with the current required for your application and calculate how much more life you will get if you use the battery below 3.3v. What I think you will find is that the amount of extended life you might get from doing a buck-boost just isn't worth it. Below 3.3v there just isn't much energy left in a Li-ion battery. Figure out your problem first. Don't assume that since the battery goes down to 2.7v that you will get much more life out of the battery by trying to tap into that 3.3 to 2.7v range.

[kenarton]

Thanks to you all for these very useful replies; has given me some food for thought.
I was not aware that there is little capacity left below even 3.3v, which I had
thought was about middle of usage range [Edit: Brain failure, I do know that the nominal voltage is 3.7]
 So a low drop-out linear reg could do  the job, or a buck DC-DC reg.
I shall also look into LiFePO4, which I have not come across before.
The RT6150A seems to very similar to TPS63000, with similar footprint, and I have also
been thinking that I might try soldering it; could be a useful exercise for other
leadless QFN parts.
Thanks again for the advice.   Ken.

[David Hess]

Any switching regulator which supports the boost configuration can be configured as a SEPIC converter allowing the input voltage to be higher or lower than the output voltage.

Also, if the battery side can be floating with respect to the output, then an inverting switching regulator can be used to turn a floating negative input to a positive output.

[Peabody]

One other thing to think about is using a 3V regulator, or even 2.8V, instead of 3.3V.  Of course it will depend on what parts are involved, and what you're doing with them, but many processors can run at a range of voltages.

[Siwastaja]

Thanks to you all for these very useful replies; has given me some food for thought.
I was not aware that there is little capacity left below even 3.3v, which I had
thought was about middle of usage range [Edit: Brain failure, I do know that the nominal voltage is 3.7]

Open-circuit voltage, aka no-load voltage ranges from about 4.18V (100%) to about 3.4V (0%).

It's just that cells have internal series resistance, so voltage drops under load by U = R*I. Also, this series resistance isn't a fixed value, it grows higher near 0%, so you see more voltage drop. The resistance also increases at cold temperatures, and as the cell ages.

A typical 18650 has internal resistance of approx 50 mOhms, which starts growing to about 100-200 mOhms below about 10-20%. At cold temperatures, this can easily go up to say 500 mOhms, which could be a usable "not the absolute worst but quite bad case" number.

If you load the cell at 1A, 500mOhms is 0.5V voltage drop so 3.4V no-load voltage becomes 2.9V under-load voltage. But if you only load at 100mA, you are still at 3.35V and not losing almost any capacity with a very low-drop LDO!