Debra Ansell

[MUSIC PLAYING]

Welcome to the Circuit Python Show.

I'm your host, Paul Cutler.

This episode, I welcome Debra Ansell.

Debra studied physics and applied math

before becoming a software engineer in the mid '90s.

She quit to stay home with her three boys

after the internet bust, then rediscovered

her love of technology as a first LEGO League robotics

coach.

She has been making open source projects ever since.

Debra, welcome to the show.

Thank you, Paul.

It's really nice to be here.

Thanks for having me on.

How did you first get started with computers and electronics?

Well, I got started with computers.

Those are actually two slightly different topics.

I got started with computers in graduate school.

I had taken a year of coding in college

and felt a little bit intimidated, frankly,

because I think most of the people who took the class--

I was a novice, and a lot of people

had been hacking either TS-80s or whatever in high school

before I got there.

But I knew enough to code.

And then in graduate school, I did a fair amount

of coding with Fortran and numerical recipes

to analyze my data.

So I was in graduate school for physics

and decided against an academic career

and came back to Los Angeles and ended up

working for an internet startup as a software engineer.

And I've been fairly comfortable with coding for a while,

though I did retire from that job about 24 years ago

and have only coded as a hobby since.

I had technically learned electronics in graduate school.

You know, I'd learned a little bit about circuits and things

like that, but only in the lab and only in very small amounts.

So I really didn't have any practical experience with that.

That part all started when my middle son turned eight

and was an enormous LEGO fan and decided

he wanted to join a LEGO robotics team.

But we couldn't find a local team.

And given that I was a stay-at-home mom with coding

experience, I figured I could coach it.

And I have told this story before, but it doesn't get old.

It was a blast to coach, but I was very jealous of the kids.

Because if you're a good coach, your job

is to stand back and let them do the work while you just stop

them from doing anything dangerous.

But I really, really wanted to get in there and build robots

and make them do the things the kids were doing.

And somebody said to me--

another coach from another team said,

well, if you think the robots are cool,

you should look into Arduino.

And I said, what's that?

And he went on to explain.

And I did a little research online

and found a great deal of information on people's blogs

on the internet and also managed to attend

a workshop nearby that had an Arduino lily pad.

And so my first exposure to Arduino in person

was blinky lights.

And it seems it never really left.

I never really kind of graduated past that.

Just the complexity has grown, but the focus has stayed.

But that just started everything.

I started building projects on my own

and then writing my own blog once I started doing projects

that were original so I could kind of give back

because I'd gained so much from what I'd

write on other people's blogs.

You're well known for your LED projects.

And earlier this year, you shared the orb-cessed LED

sphere.

How did you make an LED cube into an orb?

Thank you for asking.

I'm very proud of that project.

That was one of those aha moments that just worked.

In general, I have a lot of ideas.

And about 80% of them just never go anywhere.

But they're fun to think about.

I think most people, when you're talking about electronics,

it's just cooler to see them in shapes that aren't square,

that don't have sharp corners.

So a sphere is obviously the ultimate interesting shape.

But it's very hard to get circuit boards

into a spherical shape or get lights

on the exterior of a sphere unless you're pushing a string

through holes or something.

And so it always been in the back of my mind

to do an LED sphere.

And I'd never really come across a good idea until--

unrelated, I can't remember quite what I was looking for.

I was looking for a way to map, just map computer graphics.

And I found this projection mapping

that people often use to go from the surface of a cube

to a sphere called a quad sphere.

And they use that rather than the traditional latitude,

longitude mapping because you don't have the singularity

at the poles that you normally do.

And I saw that and I said, well, wait a second.

I know how to pipe light.

I spent a lot of my projects on how do I get light

from one place to another.

And you get very interesting diffusion effects.

So this is a mapping that if I could make a cube,

would really make it look nice in the sphere.

And I just began to obsessively work on it.

And there's a reason--

it's funny.

It's a joke, but it's also true.

The title of the article is Or obsession.

And it really did become an obsession for a very long time.

It's just seeing how far I could take that project.

So yeah, it was kind of that aha moment that said,

wait, I got to try this.

And everything really-- it was the click

that everything fell into place from.

It was so satisfying.

It only took a couple of iterations

to really get a working shell that I

could place over on the outside of a pre-built LED cube.

People have done LED cubes for a long time.

And I've seen LED spheres, by the way, too, in clever ways,

but nothing quite like the way I did it.

So that was very, very satisfying

when it all came together.

How did CircuitPython help with the project?

Well, I used CircuitPython for the version

I wrote up for Make Magazine.

And I always prototype, with few exceptions.

Most of my LED projects are prototyped in CircuitPython

because I like the dynamic interactive nature of being

able to change the code and see the changes instantly

in the LED patterns.

I was using a small microcontroller, one

of the Jao's-- I think the one with Bluetooth,

NRF 52840 that runs CircuitPython.

And what's nice, of course, is there's a wonderful LED

animations library that Adafruit has written.

So it's got a bunch of pre-built animations.

And I often will use that and extend onto that library.

So it was made mapping the sphere much easier

because the order in which the-- I

wanted to know where the pixels were in 3D space

because it's much cooler if the patterns respond,

have a 3D spatial orientation.

I was able to use the CircuitPython libraries.

First of all, the controller I used-- ah, that's right.

I used the NRF cents.

I'm going to get the numbers wrong.

So it had a built-in accelerometer and microphone.

And there's an Adafruit library already ready

to how to get that data out of the accelerometer.

And that was great.

So I'm very lazy.

I'm not going to do any more work than I have to.

So I could use the library for the accelerometer, the LED

animations library, and then combine the two very simply

to create a new animation that responded

to orientation of the sphere.

And then when I'm mapping the sphere,

if I were very organized, I would

have planned ahead for where the orientation of the pixels

are going to be given the mapping.

But of course, I didn't.

But the nice thing about mapping the sphere

is to figure out the orientation of the matrices that

made up the side, I could just light up a few pixels

at a time with just a few CircuitPython commands

and figure out what the orientation was as I went.

So it just makes my life a lot easier.

I'm a big fan of the interactive nature

of the language for prototyping and the vast array of libraries

that let me take advantage of a lot of hardware that's

out there.

Your projects often combine a variety of skills,

including CAD, 3D printing, coding, sewing, and more.

Is there one skill you enjoy more than the others or one

that you find more challenging?

The answer to that is I like all the skills.

I love gaining new skills.

And probably my favorite skill is whatever

I'm working with at the time.

And I don't like anyone more than any of the others,

but I love to combine them in unusual ways.

Like sewing and electronics is one of my favorites.

I love wearables where you're creating an accessory

or a clothing item that's designed specifically

to hold the LED strip or string that you're

wearing with a pocket for the controller,

or 3D printing and electronics, which a lot of people do.

That's a great way to protect your electronics.

But if I can come up with a new way

to combine skills that I've learned,

it makes me really happy.

And frankly, the more, the better.

I think the one I published recently in Make Magazine

was LED tote bag we called Back to the Future, which

combined 3D printing and sewing and electronics

and a Pixel Blaze controller to make the patterns work.

So that was really fun to see what you can do.

Anytime I learn a new skill, it kind of

guarantees that I'm going to come up

with a new idea, which I really enjoy.

You mentioned sewing and Pixel Blaze.

And I wanted to ask you about the Pixel Blaze pillow project

that you did earlier this year, which uses the Pixel Blaze

controller.

What is Pixel Blaze, and how did it help with the project?

Pixel Blaze is a wonderful, wonderful controller

that makes-- like Circuit Python, actually,

just makes my life easier when doing LED projects.

It was designed by a very smart programmer named Ben Henke, who

does wonderful LED installations.

And it is its own controller.

It works as its own access point to launch a web-based coding

interface, which has a JavaScript-based editor window.

And you can code your LED patterns

and see the changes on the fly, which is wonderful.

Again, makes seeing how the code you're writing

affects the patterns very easy, because it's sometimes

hard to predict how things will look versus the code you're

writing on the page.

And it has a large number of preconfigured patterns

that people contribute to the library and a very active forum

that you can ask and answer questions in.

So it's a very active kind of user base that's very helpful.

My favorite thing about the Pixel Blaze

is that it has a pattern of libraries

separate from the pixel map.

So you can specify for any project you do.

You can go ahead and use whatever patterns you've

written.

And then in a separate area of the controller,

you can specify, well, here's where my pixels are in space.

And it will automatically map whatever pattern

you've created, as long as it has the correct number

of dimensions, to the location of your pixels.

So that's very nice, too.

I can take patterns that I've written, and I frequently do.

There's some patterns that run on my sphere on my orb

that the large version of my orb uses the Pixel Blaze.

The smaller version uses Circuit Python.

But I can take patterns for the large version of the orb that

will also run on my tote bag, that will also run on a pendant.

The versatility and the ease of use

and the built-ins on that controller

are very, very nice for anyone who works with LEDs.

One of your first Circuit Python projects

combined a Trinket M0 and a Raspberry Pi 0W

to control LEDs sewn into a jacket.

How did they work together?

That was a great project.

I really enjoyed that.

There were a couple of steps in that process.

I'd been playing with jackets and LEDs,

and they'd all had controllers based

on Circuit Python and the LED anima-- using the LED

animations library.

And I was going to wear the jacket to Maker Faire,

the last one right before everything shut down.

And last minute, I decided, well, it'd

be really cool if people could program this jacket on the fly.

So I wanted to use an access point

and have people via a web browser

be able to log in to the jacket and change

the animations remotely.

So I set up a Raspberry Pi 0 and didn't then and still

know very, very little about Apache,

but set up an Apache server.

And I could figure out just enough about the Apache server.

I could get it to serve a web page that

allowed you to use a drag and drop block code to create

Circuit Python patterns.

The block code is based on the Blockly coding language.

That I could get set up.

But I couldn't, for the life of me,

figure out how to get those new patterns onto the jacket

until I realized that the--

I believe it was a Trinket M0 I was running the Circuit Python

code on initially-- worked as a separate drive

for the Raspberry Pi 0.

So all I had to do was then, once the code was generated

in the web interface, save the Python file directly

to the drive.

And that I could do relatively simply

without having to go into any kind of operating system

or anything like that.

Because you just save the file and Circuit Python

automatically picks up the code and runs it on the jacket.

So there's probably a better way to do it.

But the way that works is always the best way at the time.

And it worked.

And I got it.

And I was thrilled.

And it was just a nice little bonus

of just the convenience of Circuit Python showing up

as--

the controller showing up as a drive on whatever computer

you attach it to.

So that was a lot of fun.

And I was very proud of myself for getting it working.

And I still have no idea how Apache actually works.

I don't think anyone knows how Apache really works.

Probably true.

If people want to learn more about your projects,

where should they go?

So my older projects are available on my blog,

geekmomprojects.com.

Those are mostly the kid-friendly ones.

And the really well-documented ones, you can find--

I've written seven or eight articles from Make Magazine.

You can look at-- that's a good place

to go to look up archives of some

of my more interesting projects.

Those are the two best places to go.

If you'd like a full instruction set,

I do tend to post finished builds to social media,

Instagram and Mastodon.

But for instructions, my blog and Make Magazine

are best places to go.

I'll make sure I link to those in the show notes as well.

Thank you.

Last question I ask each guest.

You're starting a new project or a prototype.

Which board do you reach for?

It's almost always the smallest Circuit Python controller

I have on hand.

So these days, it's a QDPI or a Zhao.

Because there's such a range of both of those.

I think they just came out with a Zhao with a camera.

And I've been dying to work that into a project.

And the QDPI is great.

I'm a big fan of that too.

I like the little connector.

What's that?

They call it the Stemic-QT, I think, connector.

Which makes it very easy to, when you're prototyping,

attach a variety of peripherals very easily and access them.

So both of those are kind of my go-to boards.

And they're cheap, which is great.

Debra, thanks so much for being on the show.

Thank you for having me, Paul.

I really enjoyed it.

Thank you for listening.

Transcripts are available in most podcast players.

And show notes are available at www.circuitpythonshow.com.

Until next time, stay positive.

( calling for help )