Kevin Matocha joins the show and shares some of his projects, including the Hackpad and TinyLogicFriend.

Show Notes:

00:20 Welcome

00:27 How Kevin started with computers and electronics

3:50 HackPad

8:20 Wiring the HackPad

9:55 TinyLogicFriend - a logic analyzer using Sigrok

15:28 Which microcontroller?

Welcome to the circuit Python Show. I'm your

host Paul Cutler. This episode I'm joined by Kevin Matocha.

Kevin is a longtime CircuitPython community member,

and currently works as a vintage stereo repair technician in

Austin, Texas.

Kevin, welcome to the show.

Hey, thanks for having me. Thanks for

invitation, I hope I've got something useful I can share

how did you first get into computers and

electronics?

probably like me and 12 million other families, I

had a Commodore 64 as a kid. And my, my brother was actually the

better programmer and, you know, studying how to do the

programming at but a little bit rubbed off on me as a kid

playing games and typing in code from magazines, with the family

has we traded off typing, and eventually did some courses in

high school learning Pascal, which was kind of the

educational language of choice then, and thought I wanted to go

into computers, I started in a double E program, electrical

engineering program with computer emphasis thinking I

want to do computer stuff, but then got distracted with more of

the semiconductor design processing aspects of things. So

kind of, you know, took some more programming classes in

college but geared towards more of the, you know, how the chips

are made, and designed, and fabricated in commercial

environments. So so I kind of strayed from computing, you

know, pure computing, but always retain that sort of software

coding capability. In particular, how do you analyze

big groups of data? How do you visualize it? And that got me

into MATLAB and eventually into Python after that? So so that's

how I sort of was had an always an interest in computing, maybe

straight from that, but I always use sort of coding or to sort of

help me in my, in my day to day work. How did you discover

circuit Python, I think I got here through Arduino, my

daughter had a Arduino Uno, I think one of the basic boards

stuck down to her bed for a long time that she got as a gift, I

pulled it out and and realized, wow, the tools that are

available now, which was Arduino at that time, are so much more

than what I had when I was in, you know, first looking at

microcontrollers. And in undergrad, I remember starting

with Motorola, HC 11 programming and assembly, using buffalo to

talk to some arcane in a way of analyzing what the chip was

doing. And that was actually on a project sort of assistance

technology project with a professor who was trying to make

it where you could push a button, and it you know, say

something or highlight a picture or something like that. But it

was near impossible to get anywhere with with so low level

coding. So, so fast forward 20 years later, and see Arduino and

all these libraries. And you know, there's support for things

that visit any chip you wanted to talk to or use, you could

find somebody who had done that to help you along. So that was

eye opening. And after that, doing some basic projects, you

know, even some graphics projects with Arduino, then

heard about circuit Python, I thought, wow, this is even, you

know, the next level if I thought Arduino was, you know, a

huge leap forward in this circuit Python thing is, is even

further. So just got involved with that first a user and then

didn't help them contribute on on especially some graphics

things. And just been enjoying it with circuit Python, I guess

is awareness that basically the chips are getting so powerful,

that there's been an oversupply of capability, right? Like how

fast the process did you do? Do you need to flash LEDs, right?

It's probably not not that much. But you can use that processing

power to make it so you can flash LEDs, you know, in two

minutes, instead of taking 20 minutes to figure out how to do

that. So I think that's that's what I like about it that you

get something working in a few minutes rather than maybe an

hour or something like that with other other things. So so that's

why I was just kind of stuck with it and trying to trying to

move things along little ways that I can. Well, that's a great

segue to one of your next projects that it does involve

graphics, that probably takes a little more than a couple of

minutes, but you've been working on something you call the hack

pad. Tell me about that. I've always been interested in

displays. And sort of a big jump for me was when the Adafruit PI

portal came out, it was a chip and display all connected. So

basically, it's easy to prototype. You don't have to

have a bunch of wires hanging out there to connect the

display, which always seems to be as the wires are always a

stumbling block or a place to break. So with that it made it

clear, hey, you have this ship and circuit Python and display

capability. But it's pretty small. So having any kind of

user interface is pretty limited, just based on how big a

thing fingers are. So I've been looking for how to get a big

display, working in circuit Python. And in particular, the

problem with big displays is they're more expensive

processors a small cost, but now the display is sort of over

overwhelming costs. So from just a prototyping standpoint, I was

able to find some displays that were basically these conference

room scheduling units that I guess they're supposed to be

mounted on the wall in a conference room. And so you can

see oh, is this room available or how much time

I'm I've actually never used them. But I think that's the

purpose. So they're fairly big, and they have a capacitive

touchscreen on them. So that's really what I was looking for.

So, and even better, I could find them on eBay for 20 bucks

apiece, so, so probably even cheaper than I could buy a raw

display from, you know, overseas to get it. So you know, sort of

hit all the right targets or criteria for what I was looking

for cheap and big display and capacitive touch. And as I said,

one of the challenges with making a, I guess a product with

a display is the display cost. Part of that display cost is

usually the processor that goes between your microcontroller and

the display itself. And actually, this display that I

found is actually one of the simplest displays, so called

like an RGB display, or sometimes called dot clock

display, where you've got to give it a lot of signals

basically, for you got to send it signals to read draw every

time you want to draw frame, as opposed to other displays have

some controller in between has memory, and you just send it

like maybe what parts you want to update. So it has some some

displays have some intelligence built in. But these so called

RGB displays are cheaper because they don't have that, that

controller, okay, but it takes more work to be able to run

them. And at the same time that was fortuitous, a new chip came

out from espresso called the ESP 32 s three, which actually

inside the main processor has the capability to drive this

kind of simple RGB display. So So in essence, for you know, one

price and the microprocessor, you don't need another chip in

between to actually run the display. So in essence, it's

kind of a combination of realizing this s3 Chip now has

the capability of driving these simple displays, and finding a

suitable display to hack around on so that that's how the heck

tablet came about. So most of the work of getting it to work

in circuit Python has been how to, to understand this new LCD

driver built into this s3 chip. And it's still new things coming

every day from this press of team to try and how to better

use that. So that's, that's my final goal is to make something

that you can, you know, kind of make iPad like demos on, it's

probably never gonna be as fast as you know, what Apple can do

with its, you know, amazing chips and stuff like that. But

at least you could toy around with some touch interaction and

a reasonable size of a display that you can touch and, and

modify and react to. Right, I don't think I've seen anything

bigger than a PI portal title that has touch on it. So that

would be great to see are these about a seven inch tablet like

the old Google unexercised? Yeah, seven inches, so there's

one I've got, of course, if you, if you have that capability for

the running an RGB display, then you can pick whatever size you

want. So another benefit of this, this new chip is that it

got, or at least the demo boards that that I've been able to get

as they got quite a bit of RAM and also storage space on them.

So you'd have plenty of space for you know, large memory space

for frame buffers for these displays, too. So it's a sort of

combination of all the features of these new chips, or are

making it easier to drive these big displays. How hard was it to

connect that SP 32? s three chip to it? Was it just a matter of

simple soldering? Or was it a lot of reverse engineering

involved? Well, as you could get here, from our previous comment

about wiring the hands, sometimes the difficult, most

difficult problem, my first attempt, used a bunch of

adapters, you know, you've got first you got the chip or the

demo board with headers, and you got to convert that some maybe

40 pin cable connector, so the you know, connecting 40 Or maybe

it's not quite 40 but you know, 30 pins between the demo board

and cable adapter that's that's usually where the problem is.

Plus you got to drive the backlight. And so there's just a

bunch of wires so it's a real pain eventually at this my first

demo circuit board so I made a circuit board adapter which made

my life a heck of a lot easier where I could solder it and just

wires it's nothing more complicated than that. So if you

think of it that it just saves you from having to pull wires

out of your breadboard on a daily basis so so the circuit

board even though it's a you know, kind of takes a while to

get it made and delivered is made my life so much easier to

make some progress on the project.

Did you submit the hack pad to the latest Hackaday

contest that they're running?

Yep, sure did on the figure this the the one that

just finished up basically, how do you reuse user recycle? I

can't remember the name of it. But yeah, submitted to the heads

up and I noticed that a similar project got selected for the

next one, disabuse pupi, which is a kind of a gamepad, but also

reusing displays. So I've been watching his project for a while

and actually learned a lot from what he's done. So it's good to

see the see what he's going to come up with next.

I've been watching his projects as well and he'll

be a guest on an upcoming episode. I'm looking forward to

having him on.

Oh, cool.

One of your other projects is tiny logic friend.

What is that project?

This was a project that was started by

Scott shockcraft. Attend Nuit that works on the circuit Python

project, he had this concept of how to use low cost,

microcontroller boards like Adafruit, and many others sell,

but use the capability to do logic analysis are a formal

logic analyzer. And what that is, is basically, when you have

two chips talking to each other, they have a certain language of

ones and zeroes and timing, so that they can communicate, but

if something goes wrong, you can't see what's going wrong

with the communication, right? You may just get nothing back or

garbage or, or what when you're just you know, talking with from

your processor to house a a sensor is talking back, often

you'll need some way to debug, you know, what is it actually

talking, you know, what is the microprocessor sending to the

sensor, and what is the sensor sending back? First, you might

think you need an oscilloscope, which they say is at a time

measurement of signals, which that might work. But

oscilloscope, it basically measures like I said, a voltage

level over time, but it's usually used for repeating

signals. So that echoes the same waveform used to analog signals,

not necessarily, but usually some kind of repeating

measurements, that's the same thing over and over. But in

contrast, the chips talking back and forth, you know, it's a one

in zero of, you know, different timing between each one and, and

so usually you'll need to string like a long amount of ones and

zeros to see, you know, one transaction or you know, what

the chips sending one request for data, and then a long thing

have all the data back from the sensor say, So in essence, what

you really just need is, okay, how long was the zero for how

long was the next one for you know, so in essence, you just

want to know the ones and zeros and the duration of them. Okay,

so that's what a logic analyzer does, it basically tells you,

okay, you know, how long that those ones and zeros are for, so

you can then decode that into something that makes sense to

you. So this project was actually conceived as an add on

to an existing piece of software called Sigrok. So SIG signal

meaning Sigrok is, is a combination of signal analysis,

which the logic analyzer is, and grok, which is a word, I think,

developed in the 50s, or 60s that was used to say, like, hey,

understanding something, and I think I think coders use that a

lot. So So anyway, combination of signal understanding, is a

software. And in essence, what that does is takes, you know,

even just a raw file basically has to take that data of ones

and zeros over time, but then convert that into something you

can understand like, whether it be visually of seeing the

timing, which there's a tool inside of that tool suite to do

that called Pulse view, or build on top of that, like, knowing

that when that ship sends a hex code 32, that means I'm asking

for data or you know, are converting even a bit to numbers

or, you know, or letters, you know, Hex hex codes or something

else. So it has capability to build on top of that, not just

what ones and zeros and how long, but what does that mean,

on in multiple different levels. So so basically helps you even

further analyze how these two chips are talking back and

forth. So long story, but in essence, it's how to how to,

instead of buying a, you know, several $100, you know, logic

analyzer that's, you know, custom made for that, can we

take these $15 boards, and also use them because they can

measure fast, you know, they're spitting out signals at the same

rates? Why can't they read them and report them back. So So in

essence, it's, it's how to connect, you know, simple

microcontrollers, to Sigrok. So you can use the existing tool

set, but these fairly low cost boards, and this tiny logic

friend is one of the several others that are trying to do the

same same kind of thing. So was there a way to contribute that

back into upstream into the Sigrok project? Yeah, there,

there is a way but since we're talking to circuitpython, you

know, one thing that's, you know, promoted about

circuitpython is that there's a big community wrapped around it,

right? There's a forums where there's people crawling around

there that are willing to, you know, contribute their time to

do that, when I worked on this tiny logic friend, it felt a

little more like little more lonely of a project, I would say

that basically, I'm contributing, it's not a huge

community. And the best way to communicate is through an email

server to get some help. But it's kind of hard difficult to

elicit feedback from that, I guess, because people are doing

their own things. And, and most of that the project, actually, a

cig rock is focused on commercial or, you know,

analyzers, that people are using that so so but I have seen as

sort of a group of different projects that are kind of

heading in the same direction. So I'm hopeful that maybe those

can converge around maybe make an easier that, that the concept

is the Sigrok software can have a generic capability to be able

to accept data from a lot of different microcontroller

boards, maybe a standard way of talking to it. So you don't

really need to be changing cigarettes a lot. You just need

to you can take any kind of board as long as you fit or

behave a certain way. Can you get the cigarette to understand

that so I see a few projects You know, kind of heading in that

direction? So question is, Can Can we all sort of come together

and figure out how to make that work? So basically, can anybody

take any kind of small board that they want and still have

hooked that into the Sigrok project, which has a great

capability to be able to analyze and, and understand these these

signals? So. So that's, that's the ultimate goal. I can't say

achieve that yet. But I think that's what the concept of this,

this is

a last question for you. You're about to start a new

project or prototype. Which microcontroller Are you reaching

for?

Yeah, for me, it's no question. I mean, that

PI portal, it's got a special place on my desk, I haven't made

a bracket. So it sits right on top of my monitor. And it just

stays there looking at staring at me with a with a blinker

right there and repple. So no question. That's, that's my

first one to go to. You know, it's good to have a display

connected so you can get different outputs if you if you

want to debug what's going on. So yeah, that that cortex in for

CHIP, and it's got a fair amount of RAM. It's got touch response

on the display. So that's definitely got to be my go to.

I'm right there with you. I've talked about on the show before

how I have a pipe horrible title on my desk, and it's one of my

favorite circuit Python devices. Yeah, exactly. Yeah. Kevin,

thanks so much for being on the show. Yeah, thanks a lot. Thanks

for having me.

Thank you for listening to the circuit Python

Show. For show notes, transcripts and to support the

show. Visit circuit Python show.com. Until next episode,