Buck-boost and half-empty batteries?

Hello,

I still got piles of half-empty zinc-carbon batteries.
My idea was to connect a few of them together and then use a buck-boost converter to power clocks, thermometers and other "small stuff". I've been reading up on various information regarding this. The essence: There can be enormous loss differences between the models. However, the examples always worked with higher currents (50 mA - 2 A). Clocks or thermometers should be far below that.

My questions to you guys:
Will this work reasonably efficient?
Do you know of any inexpensive models that I could use here?
Or is the expected loss so big that the old batteries will simply die in no time?

Thanks for your help!

For such as a clock, I'd expect a simple buck-boost regulator would use vastly more power than the device it targets.

I'd consider either a simple emitter follower voltage regulator, using extremely low bias current.

Or if you want to build for the fun of it, a more sophisticated system using eg. a a high value cap or low leakage supercap for the output plus an ultra low power comparator to switch a buck-boost on and off as the cap discharged and recharged.
That could charge the cap at relatively high current for a brief time, so fairly efficient, then disable the bulk of the "lossy" components.

Even a "joule thief" type boost circuit may work with some simple comparator control?

I would suggest it's a complete waste on such low power items, which already have incredible long battery life. Those batteries already leak at the slightest provocation, and this kind of abuse is going to make that far worse.

rjenkinsgb said:

For such as a clock, I'd expect a simple buck-boost regulator would use vastly more power than the device it targets.

I'd consider either a simple emitter follower voltage regulator, using extremely low bias current.

Or if you want to build for the fun of it, a more sophisticated system using eg. a a high value cap or low leakage supercap for the output plus an ultra low power comparator to switch a buck-boost on and off as the cap discharged and recharged.
That could charge the cap at relatively high current for a brief time, so fairly efficient, then disable the bulk of the "lossy" components.

Even a "joule thief" type boost circuit may work with some simple comparator control?

Click to expand...

Thank you. I hadn't heard the term “joule thief” before.

About the capacitors: keeping a certain voltage would probably be an issue. Stable voltage, that's what I wanted the boosters for. I got some devices where the display is readable with 2x1.4V and already dark at 2x1.35V.
I'd need capacitors with a low voltage.
I got cartons full of capacitors, but they tend to be old. Only a few months ago I built a “not so smart-up” for an embedded system (140mA at 7V tops). It serves to shut down the device in a controlled manner on power loss, storing the collected data. I ended up connecting 12(!) capacitors. They barely manage to keep the device alive for a second and those were my best caps. The bigger ones lost their charge so fast that I couldn't even keep up reading the voltage drops out aloud.
There are cheap supercaps with low voltage. YouTube is full with videos of “improvised spot welding devices”. They even go down below 3V. But those are Elkos and they are really bad as they don't hold a charge for long. The ones who do are still VERY expensive. :-(
I killed one clock by simply using a resistance on three batteries instead of two. It lasted about two years before displaying all sorts of “weird behavior” and eventually quitting for good.
If you can't tell by now, I like to reuse things.

There are some “joule thief” boosters that appear to be very cheap (from AliExpress):
10PCS Solar Driver IC YX8018 8018 Solar Light Joule Thief DC DC Converter Booster IC 1.25V
US $0.58 / lot (10 pieces), Shipping: US $1.11
Those are basically the boosters I was inquiring about. The voltage on this one might already be too low for some of my devices, though.
Even though it has only 4 pins, the documentation is abysmal (unless you happen to speak Chinese).

The "joule thief" circuit in a cheap solar garden light produces at least 3.5V high frequency pulses at 20mA to light a white LED when its battery is only 1.0V. The voltage is not regulated so the voltage is higher then the current is less. I use the circuit with a Schottky rectifier and a 0.1uF filter capacitor to produce smooth 3.5VDC at 10mA to power a colors-changing LED driver IC.

There are some very clear datasheets for the XY8018 IC and others in Google. In China they spell it backwards (YX8018)..

Ah, thanks. That is good to know. I wasn't aware of the backwards trick. *lol*