AA LiIon battery - how does it work

[zelea2]

I've recently purchased some AA rechargeable batteries from Soshine.
They are lithium batteries, charge from a USB-C connector embedded in them and they even flash a LED while they do so.
I am intrigued how these work.
When you measure the voltage on the terminals is it exactly 1.5V. They do not self discharge (or if they do I haven't noticed it after only 2 weeks).
How do they get from 3.7V to 1.5V? With an LDO it would be a terrible energy loss. On the other hand a step-down switching converter will drain the battery
in a matter of hours (maybe a day?). The capacity of this tiny battery is also impressive at 2000mAh.
A voltmeter is probably drawing less than 1uA so how come you measure 1.5V with no load?

My only guess would be that they use and very low quintessence current LDO in parallel with a step-down which is only activated when the current increases over a few mA.

Does anyone know how there work because I don't want to take it apart.

[NiHaoMike]

[Bud]

The capacity of this tiny battery is also impressive at 2000mAh.

You are forgetting to apply the "Thou shalt downscale all Chinese specs by the factor of 2" rule.

The video NiHaoMike posted just confirms it.

[zelea2]

Soshine is not you regular chinese junk;
I've bought from them in the past other type of batteries (like the LiFePO4  3.2V AA) and they were all spot on.
I haven't measured the capacity of these yet but I will and report back.

[zelea2]

The battery in the video is not the same.

The one from Soshine has the charger inside the battery, you plug a 5V USB-C cable in it to charge it.
I'm not interested in the charging circuit, it's nothing interesting there.

The video does not explain how you get 1.5V on the terminals when you have only a 10Mohm load (the voltmeter).

[Bud]

The DC-DC converter works all the time, the video said it consumes 18uA at no load, so there is always voltage presents on the terminals regardless of a load.

[edpalmer42]

The DC-DC converter must operate at all times so that even a low current device can see a proper voltage and draw current from the battery.

Kerry missed a couple of points in his video.

1.  He showed the discharge voltage curve but didn't mention that his battery has a feature where the voltage drops from 1V5 to 1V1 when the battery is nearly discharged.  This is a good feature because the flashlight will dim or the smart device will report a low battery but still operate.  Without this feature, everything will just die with no warning.

2.  His capacity measurement is flawed.  Every battery, regardless of size or chemistry will show lower capacity at higher currents.  Capacity measurements are often made at the 20 hour rate.  For his batteries (2500 mah), that would be approx. 125 ma.  His measured capacity of 1850 mah @ 500 ma. doesn't seem unreasonable.  The battery may or may not get to 2500 mah @ 125 ma., but it'll be close enough that it wouldn't bother me.

Ed

[zelea2]

The DC-DC converter must operate at all times so that even a low current device can see a proper voltage and draw current from the battery.

Thanks for the explanation. My next question would be how do you make a step-down convertor that only draws such a low current?
It must have a transistor that switches current through a coil at least some tens of times per second.
I took some random datasheets and the lowest supply supply current I could find was 80uA (I'm excluding shutdown currents which are bellow 1uA)
Can you give me an example of a datasheet with 18uA no load current?

For my batteries shipment from China took 1.5 months because they came by boat and were at 1.5V when they arrived.
That's an impressively low discharge.

[edpalmer42]

I have a reasonable understanding of batteries, but my understanding of the details of DC-DC converters is small at best.

I was looking at some Youtube videos about these batteries and in one, the host was complaining about poor shelf life of the charged batteries.  He also complained about batteries that died and couldn't be recharged.  Part of that could be poor quality parts and/or assembly, but part is due to the design constraints.  Low idle current (as you stated), operating current range of microamps to amps, very small size, and of course, very low price all add up to a challenging design.  There might not be a part that is suitable.  It might exist, but doesn't have a published datasheet because it's only available in quantities of 10,000 or more.  It might even be a custom or semi-custom chip.

These things are relatively new on the market.  It'll take some time before everyone decides if the idea is practical and cost-effective.  Welcome to the bleeding edge!

Ed

[amyk]

The DC-DC converter must operate at all times so that even a low current device can see a proper voltage and draw current from the battery.

Thanks for the explanation. My next question would be how do you make a step-down convertor that only draws such a low current?
It must have a transistor that switches current through a coil at least some tens of times per second.
I took some random datasheets and the lowest supply supply current I could find was 80uA (I'm excluding shutdown currents which are bellow 1uA)
Can you give me an example of a datasheet with 18uA no load current?

For my batteries shipment from China took 1.5 months because they came by boat and were at 1.5V when they arrived.
That's an impressively low discharge.

https://fscdn.rohm.com/en/products/databook/datasheet/ic/power/switching_regulator/bd70522gul-e.pdf

Typical 180nA quiescent current, max 1uA.

[zelea2]

Typical 180nA quiescent current, max 1uA.

Wow! Thanks for that link, I didn't know such buck converters existed. You learn something every day.
With no load it switches to a low power mode where it generates just one pulse through the coil and then
monitors the output voltage. If it doesn't decrease fast enough in 8us then it remains in this low power state.
The switching frequency on that chip is probably measured in single digit Hz.

[tom66]

The DC-DC converter must operate at all times so that even a low current device can see a proper voltage and draw current from the battery.

Thanks for the explanation. My next question would be how do you make a step-down convertor that only draws such a low current?
It must have a transistor that switches current through a coil at least some tens of times per second.
I took some random datasheets and the lowest supply supply current I could find was 80uA (I'm excluding shutdown currents which are bellow 1uA)
Can you give me an example of a datasheet with 18uA no load current?

For my batteries shipment from China took 1.5 months because they came by boat and were at 1.5V when they arrived.
That's an impressively low discharge.

Pulse-skipping sometimes known as PFM mode.  The converter runs one or two cycles of inductor current then sleeps for some time, often 100ms or more.  The output voltage stays high because there's no load.  As the output self-discharges through leakage current, the converter will make the decision to switch to keep the output high.

As load increases the spacing between the pulses gets smaller until the converter switches continuously, sometimes the switch into continuous PWM operation is a 'hard' limit other times the control loop just decides to switch on every cycle, it depends on the architecture.  Often times the PFM mode is a separate control loop as it results in higher output voltage ripple than desired for loaded operation.

The operating current under no load will become some fraction of gate driver losses, capacitive losses in the switch node and whatever analog logic is required to make the controller function.  A good example of the opposite case (boost converter) is wireless mice.  Most of these use a single 1.5V AA, but that isn't even enough to light the red LED.  They will boost the battery up to around 3V for the radio, image sensor, etc.  And then sleep when the mouse isn't moving.  If you take the 2000mAh capacity and 12 months lifespan then the average current has to be 250uA so the standby current is probably in the range of 10-25uA and operating current around 1mA (will depend on the mouse). With the right converter they can achieve 12+ months standby from a single AA battery and work all the way down to about 0.7V on the cell, when it is effectively 100% exhausted.