How much voltage can we squeeze out of a CR2025? Turns out… plenty! I built a little Joule-thief-style boost circuit with two 1 mH inductors, an NPN, a Schottky, a 27 V zener clamp, and a storage cap to zap my LCD pad clean at around 27 V. I profiled it with a Nordic power monitor: about 11 mA during the boost and ~9.28 mC per click—so tens of thousands of presses from one coin cell. Then I added a second transistor tied to the zener to auto-limit the drive, dropping hold current to about 1.5 mA—huge win. Scoped it (~38.7 kHz, base swinging to nearly -6 V), laid out a PCB, 1:1 print check for fit, and sent the gerbers off to PCBWay.
How much voltage can we get from one of
these CR 2025 batteries? They’re rated
for 3 volts. But what if we want much
more voltage?
I can’t change the laws of physics.
How about 27 volts or even more? Not
bad. In this video, I’ll show you how
this little circuit works and we’ll lay
it out on a PCB and get it made up. It’s
pretty cool.
You may remember one of these from a
previous video. It’s one of these uh
clever little um LCD things that you can
draw on and then you can clear. Now, in
the previous video, we took this apart
and unfortunately I zapped it and broke
it. Uh now, it does need around 27
volts. You can go lower, but um 27 volts
seems to clear it nicely. So, what I’ve
done is I’ve made a little boost circuit
here. It’s a dual fee based thing. So,
if I click this button, it clears it.
So, if we uh bring in our multimeter and
see what voltage we get. I’ll draw some
stuff just to make sure you can see it
happening. Push the button. We go up to
around 27 volts. So, not bad. Um it’s a
really simple circuit. Now, with the
dual thief, I’m quite late to the party
on these. They were very popular on
YouTube around eight years ago. You can
use a toidal um inductor core. So you
can take off this copper wire and rewind
it or you can just use a couple of
inductors. So these are 1 millhenry
inductors. So on this PCB we got the
coin cell battery which is our power.
I’ve added a little resistor to limit
the current and a little um decoupling
capacitor to give it a little bit of
reservoir. Go through a switch go
through the two inductors. We have our
base resistor here going into our NP uh
NPN transistor. the shock diode going
from the collector. And here we have a
another capacitor to store the energy.
There’s a um zenia diode, a 27 volt
zenia diode and a little bleed resistor.
So does work quite nicely. So what I was
wondering is just how much current are
we drawing from this poor little coin
cell. So I’m going to wire up my um
Nordic power monitor and do a quick
capture.
Okay, so I’ve wired up the Nordic power
meter and I’ve opened up the software.
So I’m in ampair mode and we’ll hit
start, which will start recording. And
now let’s come in and click the button
and see what happens. I got my
multimeter hooked up. Okay, so that
worked. Uh let’s do another one.
And another one.
And another one. Okay, so we got some
good captures here. Let’s uh let’s
highlight um let’s highlight one of the
longer ones. How about this one? So, we
just shift and select. There we go. So,
that’s a nice zoomed in version. I’ve
got the actual button push selected. So,
we can see on average we’re drawing 11
milliamps, which is actually that’s not
too bad for a coin cell. That’s quite a
reasonable current draw for a coin cell.
And in total, we’ve got 9.28 millum
charge. So, I’ve gone off to our
friendly AI. I’ve got a CR 2025 battery.
I click a button. It uses 9.28 millum.
Let’s see how often can I click the
button. So, we’ve got 150 milliamp at 3
volts and times that by um I think that
looks like 60 * 60 get 540,000 millum.
So, it’s saying it’s very slow. So, we
can click our button 58,200
times. Uh, now, of course, there’s some
real world things. So, the coin cell
will go down over time. So, it’s saying
maybe we can get 40 to 50,000 reliable
presses. So, that’s not bad, is it?
Pretty good. So, I think we can turn
this little circuit into a PCB and maybe
we can fix this board up.
So I’ve also captured a trace on my
oscilloscope. So the the blue trace is
on the collector of the transistor and
the yellow trace is on the base of the
transistor. So you can see our base of
the transistor is pulled very low. So
it’s on two volts per division. So
that’s one two three it’s almost minus 6
volts on the base of our transistor.
that hard turns off the transistor which
means the magnetic field on the inductor
on the collector collapses and we get
this big spike of voltage. So that’s 10
20 20 something volts. Obviously that’s
being capped by the zenia diode on our
output and that happens at around sort
of 38.7 kHz. So that charges up our
output capacitor pretty quickly. So
that’s pretty interesting to look at.
It’s a very interesting circuit.
Definitely worth playing with.
So it turns out with a small addition to
this circuit, we can actually make a big
improvement. So it just requires one
more transistor. And what we do is we
connect the zenia diode through to the
base of that transistor. So when you go
over the voltage that’s required, the
transistor gets turned off. So our
switching transistor here, the base gets
pulled down low by the other transistor
when you go over the voltage. So, let’s
try this out. So, I’ll start the
recording going and let’s push the
button. So, I’ll do it once
and again
and again.
So, let’s look at our traces now.
So, if we just go and zoom in on this
one. So, you can see, let’s zoom in a
little bit.
What we can see is initially if I can
get this to zoom nicely. This uh this
this UI is really difficult to use. Very
sensitive on the zoom.
Okay. Well, that’s that’s probably as
good as we’re going to get. I’ll keep
doing this. We’re going to be here
forever. Okay. So, if we select this
first half, then you can see it’s very
similar to the measurements we got on
the original circuit. But you can see
here as soon as it hits the 27 volts
that’s limited by our Zenia diode, it
drops straight down and we hold the
button for this period here. But
actually, we just average
1.14 milliamps. So that’s absolutely
amazing. So the total time when we got
the button held,
we actually average 1.53 milliamps,
which is 1.27 27 millons.
So, I think we could click the button a
lot of times and it would carry on
working. So, I’m going to knock both
these PCBs up and get them sent off to
PCB way and we’ll see how well they work
when they arrive.
Okay, so here is our PCB. So, we have
the the pads for mounting the two um LCD
plates, I guess we’ll call them. We have
the blister button and we have the
mounting holes. So, let’s have a quick
look at the 3D viewer. Make sure that
looks good. So, these are our two
inductors that will be um magnetically
coupled. Here’s our shock key diode.
Here’s the zener diode. This is our
reserve capacitor that gets charged up
and it’s bleed resistor. We have the two
connectors for the battery. So, they are
gosh here and here. Here’s our
decoupling capacitor. And here’s our
current limiting resistor. So, and then
on the other side, here’s the blister
button and the two pads. We’ll have to
cut these um fairly short so they don’t
poke through, but um it’s looking good.
So, the best thing to do to really check
this is to actually print it out at
one:one scale and compare it with the
actual PCB. So, let’s do that now.
So, this is a really good way of
checking your PCBs. You just print them
out at one:one scale. So, I’ve got the
original PCB here. So, if we just lay
that on top, you can see that my um my
mounting holes match up exactly to the
PCB. So, that’s pretty good. So, if we
get this PCB made, it should mount
exactly where this one mounts. And these
two pads here for the battery connector
are also pretty much in the right place.
So, I can desolder these battery
connectors and solder them onto my new
board. Now, if we flip the board over
and bring the print out uh the other
side,
then we can see that these tabs should
pretty much line up. So, we can solder
the two um tabs for the um two plates
straight onto these. And the blister
button is pretty much in the right place
as well. We can be fairly flexible with
this. So, these blister buttons, you can
just peel this off. It’s just on some
sticky plastic. So, we pull it off from
this board and stick it onto our board
and we’ll have the nice clicky switch.
So, let’s get back to the computer and
actually submit this to PCB way.
So, everything looks good. Let’s get
this submitted to PCB way. So, what
we’ll do is we’ll export the um gerbers.
So, go to fabrication
gerbers. Um we’ll create a folder to put
these in. So new folder
togethers
and yep we want to do that everything is
already selected properly.
So when you do the plot it’s generated
all the gerbers and also generate the
drill files. So everything’s fine here.
It’s going to go into the same folder.
Just hit generate. So there we go. Okay.
So we’re in our folder. We’ll just
compress this folder. So turn it into a
zip file. Says our Gerbers file. And now
we go off to PCB way. Okay. So we’re on
the PCB way. We just need to upload our
Gerber file. So click on this.
And then we’ll just add Gerber files. So
I’ve got my Gerber zip file here. Double
click on that. It uploads. It should
pick up all the parameters
automatically. So I know it’s a two
layer board. It’s got the size. Then we
can choose what we actually want to do.
So, I think we’ll just get five pieces.
It’s two layer board. Um, thickness.
We’ll make it slightly thinner. Looking
at the existing PCBs, they are
incredibly thin. Let’s see how thin we
can do without actually going up in
price. So, at the moment, we’re going
for $5, which is pretty amazing. So, how
thin can we go before the price goes up?
Okay. So, looks like a 6 is the thinnest
we can go and still stick with the $5.
Um, and that’ll take three to four days
to build. So, the rest of it’s all fine.
What if we can get a different color?
No, that’s more expensive.
No, no, no. I think we’ll stick to green
then. Stick to the $5. Perfect. Matte
green. No. More expensive purple. No.
Okay. So, we’re going for green and
we’re stick with white silk screen.
We’re not doing any UV color printing.
If you want to see some of that, look at
my ZX Spectrum. Uh, and everything else
is fine. We’re not doing SMD assembly.
So calculate
and I’m shipping to the UK. So, United
Kingdom, but what I’m going to do, I’m
going to store up a whole bunch of PCBs
that I’m going to get um designed and
then I’ll get them all shipped in one
go. So that will be kind of done later.
So I’ll just add a whole bunch of orders
to this and then I’ll recalculate the
shipping so that we get quite good
shipping costs. So hopefully those PCBs
will arrive sometime in the future. I’ll
see you in the next video.