Imagine if chefs couldn't taste, artists
couldn't see, and musicians couldn't hear. Those jobs would be basically impossible,
but that's exactly how things work for anyone building electronics. Us humans just
can't perceive electric current, except y'know, when we're getting zapped by it. If I
could sense the current traveling through a wire, I could literally poke around in my projects
and feel the power flowing and the chips chattering. Today, I'm going
to give myself a super power. ...right after I deal with this hangover. [Jazzy funk tune goes "la la la la la la"] Huge announcement: I have just
launched my Patreon campaign! I'll tell you more at the end, but tl;dr, give
me all your money. Link is in the description. So check this out. Whenever electricity flows
through a circuit, it creates a magnetic field around the conductor. The more current flows
through the wire, the bigger and stronger the magnetic field surrounding it. This works the
other way too; when we cut the current through the wire, the magnetic field will collapse back
into electricity and send a bit of voltage in the opposite direction. All we have to do is
put the right sensor into the magnetic field and it'll tell us how powerful it is, which
in turn tells us how much current created it. So this is a digital compass, which is
a type of magnetometer you'll find in smartphones and watches. This thing
detects magnetic fields themselves, so we don't have to wait for them to collapse
back into electricity to measure them. So let's demonstrate. I'm going to hook
this thing up to a Teensy, of course, and I'll stream each axis of the compass to
this graph. You can see that as I turn it around, it picks up the earth's magnetic field
just fine. Watch as I wave a very, very tiny magnet around, and again, it
makes a huge impact on the readings. So I'll put this wire over it and run some
current through it, and it doesn't make a dent... ...actually it does. The resolution' s not great,
but this thing totally picks up the current traveling through the wire. I really should do
the experiments BEFORE I write the episodes. We're going to need a more sensitive sensor.
Another option would be Hall-effect sensors, like the ones I use to detect my fingers in this data
glove. Those won't work either, unfortunately; you can absolutely buy Hall sensors that can
detect super-small DC currents, but you have to put a ferrite ring around the wire to concentrate
the magnetic field on the sensor. I can't clamp a sensor over a wire and a circuit board,
because the wires are... in a circuit board! [Goofy-style G'HYUCK!] What we need is a fluxgate magnetometer. That
sounds like some star trek bull[REDACTED] but trust me, it's a very real and very precise
magnetic sensor. This is a Texas Instruments DRV425 fluxgate magnetic-field sensor.
We just power it here, ground it here, this is our reference voltage, and this voltage
is proportional to the strength of-vuvuv the of the magnetic field... the strength of the magnetic
field... all right, I'm done with this scene. You can clearly see that when I turn on the
power through this wire, the fluxgate picks up that very faint magnetic field. As I turn up
the juice, the reading increases in proportion. The fluxgate itself is just a tiny piece of
magnetizable material wrapped in two coils of wire. The drive coil magnetizes the core with
an electric current and then shuts off. That magnetic field collapses back into an electric
current, which is picked up by the sense coil, and the device measures it. Then the
drive coil magnetizes the core again, but in the OPPOSITE DIRECTION. It shuts
off, sense coil measures again. So, back and forth it goes, creating a magnetic
field, allowing it to collapse, measuring it, creating the opposite magnetic field, allowing
it to collapse, and measuring that too. So these readings should be symmetrical; in
fact, we should always get the same amount of current out of the sense coil as we put in
through the drive coil because, like, y'know, where the hell else is it gonna go? But if
we expose the sensor to a magnetic field, heh-hey hey! The field helps the drive
coil in one direction, so it charges the core faster and returns a higher reading. The
field FIGHTS the coil in the other direction, which prevents it from charging the core
all the way, and it reduces the reading. It's just... a current transformer... in disguise. [Palpable awkwardness] ...arright... This thing is sensitive enough to pick up the
faint magnetic field around a current-carrying wire with, like, a really high level of accuracy.
By measuring the difference between the readings in each direction, we can tell not only how
powerful the magnetic field, is but also which direction it's pointing. In this sensor the core
runs from the left side to the right side, so it detects current on wires going from the top side
to the bottom side. That's because current follows a right hand rule, so if the current is flowing
thiiis way, magnetic field forms thiiis way. By the way, I usually skip over the theory, but
in this episode I decided to mix it up and dive a little deeper. Did you enjoy learning a bit
about electromagnetism, or should I just shut up and jump into the project? Did I do a good job
of explaining this stuff? Drop me a comment and roast me up real good; I am still trying to find
the perfect blend of education and "education." So for this project we're NOT using a Teensy! [Gasps from live studio audience] We are using the Adafruit NRF52 Feather because
it integrates the battery circuitry that we need to turn this into a wearable. This thing also
acts as a Bluetooth Low Energy master device, which is important, because it lets us connect
to this Buzz wristband. I am using this vibrating wristband instead of, just, you know, adding
vibrating motors to the project, because drunk me drunkenly entered a developer contest on Hackster
and if I ghost them one more time, they're gonna stop sending me free hardware. I tried to convince
the company that makes this, Neosensory, to sponsor this episode, but the CTO called the
hell out of this bluff. Enjoy the free PR, Scott. All right, so what the hell is this
thing anyways? Well, out of the box, the Neosensory Buzz translates the volume
and pitch of incoming sounds into vibrating patterns on these four rumble motors. It lets
you touch sound. If you're hard of hearing, this sensory substitution can improve
your lip reading and alert you to cars skidding out of control behind you.
If your hearing is fine, it's... it's kind of useless. But I guess it makes
your speakers sound like a bass booming rave? [Tacky 90's trance] [Oontz oontz oontz] [Trance still playing in background] Give this thing to Elon Musk and throw
them in a closet with three tabs of acid, he'll get us to Mars in like 20 minutes. All you have to do is call some functions and
the Feather connects straight to the Buzz, overrides the sound-to-vibes functionality, and
directly tells it when it's ready to rumble. Somewhere along the way, fun fact, your sketch
has to agree to Neosensory's terms and conditions. I Am Not A Lawyer, but I really doubt that you can
sue an embedded device when it breaks an end-user license agreement. It does this by literally
saying the word 'agree' over Bluetooth serial. Somebody overthought the HELL
out of this and I love it. This thing's really simple to build. We
just wire the sensor up to power and ground, we run the output and reference pins to the
Feather's analog inputs, we tell the thing to vibrate when the thing senses the thing, and
there we go! I can now feel electric current. I mean, like, you have to trust me on that; uh,
you can't feel anything through YouTube. I wish I wish you could, though. I'll tell you
what... uh... put your phone on your wrist, and on the count of three, I
will turn on the current and you shake your phone around. Desktop and TV
people, like, I don't know. Figure it out. One... two... three... [Buzzy mouth noises]
[bvvvvvvvvvvvvvvvvv] Fine. We'll visualize it too. Here's
an LED shield - I mean an Adafruit DotStar Matrix FeatherWing™, all rights reserved. Ermegerd, merchernical erngernering! I fired up Fusion 360 and I modeled a nice
three-part enclosure that glues right onto this weightlifter's glove. This is actually
the leftover glove from the Somatic project, which is why you should never throw anything away, ever. This sandwiches the boards and braces the
battery, so we avoid some lithium-splosions. I'm going to glue this onto the glove
with some E6000 and screw it all shut. Now, we gotta make the wearable sensor
wearable. Magnetic fields weaken with distance, so I designed this little ring that exposes the
sensor so we can really mash it into the wire we're measuring. I printed this in flexible TPU,
which makes it a lot more comfortable to wear, but also lets me design a little pocket
to snug the sensor board in there. I add a little cable management on the fingernail,
so I can route the wires gracefully over my hand. It's important to leave some
slack so I can bend my finger, so I printed this little clippy dingus to
keep the wires from snagging on protuberances. I will just display a simple
linear plot of current over time and add some flamboyant color
coding to make it camera-friendly. [In the manner of an elevated individual]
Now you can, like, see what I feel, man. Electronic status: Done. The code is also
done, but my analytics show that about 30% of you people run away as soon as I put a
line of code on screen. You can download it yourself and see what it looks like. Links
to everything are down in the description. I will give you the spoiler
though... it looks like code. [Triumphant music]
All that's left is to give this thing
a name! Ladies, gentlemen, and cyborgs, say hello... to the THUNDER FINGER!
[Record scratch] I REALLY suck at naming things. So we're powered up, linked up, locked, loaded.
there's a chance that I'm currently the only homo sapiens capable of feeling direct
current. Let's finger some electrons! This thing actually works alright for a ten-hour
project. It isn't as sensitive as it should be, but that has more to do with the deadline than
the hardware. If I had some time to tweak, I could calibrate the DRV425's response
range to better pick up what I'm looking for you. I could also feed the output reference
voltages into a differential amplifier, which would increase our resolution. Also,
I'm polling the sensor super-slow because it makes the code easier to write, but we can
actually read the fluxgate up to 23,000 times a second. With the right code, we could decode
PWM pulses, we could pick up audio transmissions, we could even sniff 9600-baud serial
traffic. I could convert each one to a different tactile texture so I can
debug and hack electronics with my finger. This finger, in fact. This one. Of course, the Thunder Finger's vibrating
motor should really be thundering my finger but not my... wrist... but come on, the Buzz
was free. It's free hardware. Speaking of, I am putting a link to the Hackster Neosensory
contest in the description, because Hackster has promoted a bunch of my projects, so it's
about time I return the favor. There's still two weeks left in the contest. I'm looking forward
to getting stuffed by each and every one of you. If you want to build the
Thunder Finger for yourself, I'm putting the links in the description. I'm pretty sure you can actually buy all the
parts for this project, so knock yourself out! Big announcement, my Patreon campaign is
live! The super-duper-rad perks include a monthly Q&A stream, sneak
peeks at upcoming projects, Brooke's trolltastic blooper reel. I need your
support to buy parts and make awesome projects, so if you want to see more videos, it would
totally rule if you can kick in a few bucks. My glorious Lab Patrons get access to the
top-secret supporter-only channels in the Discord, which has 2,600 friendly members and
five jerks. It doesn't make any money, but it does make friends, and isn't... isn't the
real money the friends you make along the way? ...No, it's money. You
can't pay rent with friends. Thanks for watching, and
I'll feel you in the future!
