532: It's Still Real Fire
Transcript from 532: It's Still Real Fire with Kevin Santo Cappuccio, Christopher White, and Elecia White.
EW (00:00:07):
Welcome to Embedded. I am Elecia White, alongside Christopher White. Our guest this week is Kevin Santo Cappuccio, founder of Architeuthis Flux. Which, as you know, must relate to the giant squid, because the giant squid's name is Architeuthis dux. I am so excited.
CW (00:00:25):
Hi Kevin. I do not think we are going to talk about squid.
KSC (00:00:29):
Oh, it is going to end up there.
CW (00:00:30):
But Elecia thinks we are going to talk about squid.
KSC (00:00:32):
We will get there. We will find a way.
EW (00:00:35):
Could you tell us about yourself, as if you were not a squid?
KSC (00:00:39):
Okay. Yeah. So I am Kevin. I obviously make the Jumperless and made a whole life around it. Yeah. I do not know. Art school dropout. Used to work on squid, doing science and commercial fishing. So the squid thing is there. And yeah, it is what I do. Just sit around and be a weirdo.
EW (00:01:08):
What is Jumperless?
KSC (00:01:09):
Oh. Okay. So Jumperless is a breadboard that is jumperless. It does not need jumper wires. It is weird because it is not a unique idea. I feel like anyone in college is thinking about this like, "Why am I dealing with these stupid <beep> jumpers?" Yeah. It uses analog crossbar switches to basically just make the wires a software thing.
EW (00:01:39):
Make the wires a software thing.
KSC (00:01:43):
Okay. Yeah. There is always the, "We have to start at breadboards." I guess not for this audience, but it basically allows you to add and remove wires very quickly and really easily. So, instead of actually using a physical wire, you are using a very complicated set of electronics on the back to make a virtual wire.
EW (00:02:08):
But it looks like a breadboard with a lot of extra LEDs. And you plug things in, capacitors, resistors, EEPROMs, whatever, and you can plug them in somewhat haphazardly. And then in the software, instead of having all of these little jumper wires that go everywhere and make it look like a troll head doll, you do that all in software. And then you show it in LED.
KSC (00:02:32):
Yeah.
EW (00:02:33):
Okay.
KSC (00:02:34):
Yeah. Got a lot of LEDs.
EW (00:02:35):
Yes.
KSC (00:02:35):
I think it is 445 of them. So every hole has an LED.
EW (00:02:44):
And to show the connections, you need to light them all. So yeah.
KSC (00:02:49):
Yeah.
EW (00:02:50):
All right. So we are going to talk more about Jumperless, but first we want to do lightning round. Are you ready?
KSC (00:02:56):
Yeah, we are starting with that. All right. Cool. Yeah.
CW (00:02:59):
Favorite LED color?
KSC (00:03:01):
Pink.
EW (00:03:03):
Squid, anchovies or sardines?
KSC (00:03:05):
Squid.
CW (00:03:06):
Wire wrap, solder or breadboard.
KSC (00:03:09):
Oh God, I do all of them at the same time, like, in the same product.
CW (00:03:12):
<laugh>
KSC (00:03:13):
Yeah. I love wire wrap though.
EW (00:03:17):
Which is cooler? Transparency, iridescence, bioluminescence, or fluorescence?
KSC (00:03:24):
I really like the fluorescence that cuttlefish do. I do not even know if they call it fluorescence. But the bottoms of them actually take in or emit more light than the tops do, so they do not cast a shadow. It is a really cool thing on the bellies of cuttlefish. So I am going to say- Do they call it fluorescence? I think it is fluorescence. Yeah.
CW (00:03:46):
Complete one project or start a dozen?
KSC (00:03:48):
Both. I really like going in and spending years on the same project, that whole optimization thing. But I also start a dozen projects at the same time, and also complete them. I just have free time, so I could just do it.
EW (00:04:06):
If you could teach a college course, what would you want to teach?
KSC (00:04:08):
I am going to make up a class. It is going to be doing electronics like an art school dropout.
EW (00:04:15):
Hmm. I like that.
CW (00:04:17):
Favorite fictional robot?
KSC (00:04:20):
God, I did not think of this one. Okay. What was that "Flubber," with the Weebo little flying girl?
CW (00:04:31):
I do not know.
KSC (00:04:32):
It looked like a Walkman that was flying around, and it talked.
CW (00:04:36):
Oh. Okay. I do not know that one.
KSC (00:04:38):
Yeah.
CW (00:04:39):
It is an acceptable answer. All answers are acceptable.
KSC (00:04:40):
<laugh>
EW (00:04:44):
And do you have a tip everyone should know?
KSC (00:04:47):
I feel like I have a lot of them. I do not think they are really important. Oh yeah, just do it. You will figure it out. Just do stuff. It is cool. Especially with electronics. You do not have to learn. Just do it and you will learn.
EW (00:05:01):
I think that that is an interesting thing from somebody whose video for the Hackaday Challenge, or Hackaday Prize, involved snipping a wire and having things light on fire. Was that a real- Really or did you fake that?
KSC (00:05:19):
I did fake it. Yeah. I actually- I tried it a few times, getting- But no, that was not. So what was behind the breadboard was a little tray of vape juice, like propylene glycol with nichrome wrapped around like a spongy thing in there. And yeah, so I am just hitting my power supply right after I cut it. You do it hard enough and it actually just catches on fire.
EW (00:05:46):
I am so disappointed. This is a lie. I am done.
KSC (00:05:48):
<laugh>
CW (00:05:50):
It is still real fire.
KSC (00:05:51):
I did consider- There were a few ways where you are like, "How do you-" You have a huge MOSFET and that is holding back 10 amps of current through it.
EW (00:06:01):
No. No, no.
CW (00:06:01):
But that is dangerous.
EW (00:06:03):
No, but you put a couple capacitors there in backwards.
CW (00:06:05):
Also dangerous.
EW (00:06:06):
And that is really fun. That is not that dangerous. They pop and it is a little explosion.
KSC (00:06:11):
That still scares the crap out of me.
CW (00:06:13):
You get boiling electrolytic fluid all over you.
EW (00:06:16):
Well, if you put them in series, they go, "Pop, pop, pop, pop," which is even better.
KSC (00:06:20):
Wait, can you do that reliably? I feel like I have never been able to do that on purpose.
EW (00:06:25):
It depends on the capacitor. You have to get the right kind.
KSC (00:06:28):
Yeah.
CW (00:06:28):
You have not done this intentionally. This is all accidents you have had, right?
KSC (00:06:31):
Yeah. Yeah!
EW (00:06:32):
This is all- Well, no, we actually did it intentionally after the third accident.
KSC (00:06:39):
<laugh> It is one of those things where you will always be surprised at how much they can actually handle before that happens. And then sometimes they will just go off for no reason. Yeah.
CW (00:06:48):
<laugh>
EW (00:06:50):
Okay.
KSC (00:06:51):
That would have been a fun project. I would have liked to do that.
EW (00:06:53):
So we talked to Peter Griffin about Jumperless, where he made an Excel interface.
CW (00:06:58):
<laugh>
EW (00:07:01):
Peter, you are such a wild man. He made an Excel interface so that he could connect the breadboard up according to the spreadsheet. And then that goes down via serial interface to your Jumperless. And then-
KSC (00:07:22):
Via a plugin that only works if you have a Microsoft 365 subscription and are on Windows.
EW (00:07:29):
Yeah. It does not surprise me. And then it tells the Jumperless, "Connect this point to that point." And the Jumperless uses magical electronics I do not care about to do that.
CW (00:07:46):
You care. We are going to ask about the magical electronics.
EW (00:07:49):
Oh, I was not. It is just a bunch of the chips that act like SoC.
CW (00:07:53):
Yeah, but I have questions about that. Later on, we will ask.
KSC (00:07:56):
Okay.
EW (00:07:57):
Okay. It is like a field programmable array.
KSC (00:08:00):
Yeah.
CW (00:08:01):
Okay. Well-
KSC (00:08:03):
Okay. I have a bone to pick with the FPGA characterization.
EW (00:08:06):
Okay.
KSC (00:08:06):
It is okay. It is just the story that they tell you with FPGAs is different from what is actually going on. They are actually just a bunch of lookup tables. It is not wiring anything from one side to another. Yeah.
EW (00:08:21):
Does the Jumperless?
KSC (00:08:23):
The Jumperless is- It is a real- It is carrying current. It is an analog switch. Yeah.
EW (00:08:27):
Okay.
KSC (00:08:27):
No, I wish FPGAs worked like that. That is always my thing. Yeah.
EW (00:08:35):
I like it either way.
CW (00:08:36):
<laugh>
KSC (00:08:36):
<laugh>
EW (00:08:36):
And I am worried this question is going to come out wrong, but I am just going to go for it. Is this a tool or a toy?
KSC (00:08:46):
Absolutely a <beep> toy. That is what it is for. Yeah, no. You can find things to do with it that are fun. But no, do not put this in a mission critical.
CW (00:08:58):
Oh, sure.
KSC (00:08:59):
Yeah.
CW (00:08:59):
But there is a whole vast continuum of "toy" up to things, like I consider- This may be my last podcast. But I consider the Arduino a toy, in some sense. Sometimes.
KSC (00:09:12):
Yeah.
EW (00:09:14):
I am not going to disagree with that. I have one of the most seen articles that says, "Do not use Arduino." Unprofessional.
CW (00:09:21):
But that is not to say you cannot use for useful things.
EW (00:09:23):
Situations.
CW (00:09:24):
Or learn a lot by using it.
EW (00:09:27):
Oh. No. I have no problem with it. But it does make a difference. Am I buying this for myself? Or am I expecting my company to buy a half dozen and use them in different products?
CW (00:09:37):
Oh, sure. Well, if you expect them to do that, you should probably tell them that it is not a toy.
EW (00:09:41):
<laugh> Well, yes.
KSC (00:09:42):
<laugh> Yeah. Yeah. Do you want to cut a second version of this where I say, "It is an absolutely necessary tool for your company," if someone needs to justify it to their boss?
CW (00:09:52):
<laugh>
EW (00:09:54):
Okay. So we have established it is a breadboard that makes software connections, instead of hardware. And that it has lots of LEDs and it is very pretty. Of course, the blinking LEDs makes it hard for me to watch the videos. Some of the videos were pretty fun because I mentioned the fire.
(00:10:08):
But how do you explain this to people outside makers and hardware engineers?
KSC (00:10:15):
Oh, I totally just choke on it and barely answer the question.
EW (00:10:18):
Huh. Good, good.
KSC (00:10:18):
And then someone else who understands it will come in and save me, usually. I have normie friends that will be like, "Here is what it does." And I am like, "Oh, thank God. Thank you." Because when you have to start at what a breadboard is, it is hard.
CW (00:10:35):
Yeah.
EW (00:10:35):
Yeah.
CW (00:10:35):
Unlike a passive breadboard, however you want to call it, those are pre-wired in columns, whereas-
KSC (00:10:45):
These ones are too.
CW (00:10:45):
Oh, they are. Okay. So you are mapping column to column.
KSC (00:10:48):
Yeah.
CW (00:10:49):
Okay. That makes sense. It would be really, really a lot of chips to map any point to any other point.
KSC (00:10:57):
Yeah. There is a lot of it in this where you want to keep it as familiar as possible. So it is set up like a real breadboard, even keeping some of the mistakes of the past that are- You know like the power rails are offset on a regular breadboard?
EW (00:11:12):
Yes. So wrong. Yes.
KSC (00:11:12):
It is very wrong. Oh, the worst one? I am going to ruin everybody's life right now. Okay. You know how it is marked every five rows on a breadboard?
CW (00:11:22):
Yeah.
KSC (00:11:23):
There are three rows between one and five, and there are four between the rest of them. It is ridiculous. It is like one indexed kind of, but then every five. So between one and five is shorter than the other ones. It is so gnarly and I had to do it.
EW (00:11:46):
When I was watching some of the videos, I saw that it was using Wokwi, which is an online-
CW (00:11:53):
Electronic simulator.
EW (00:11:53):
Simulator.
CW (00:11:53):
Yeah.
KSC (00:11:53):
Yeah.
EW (00:11:57):
But why do you need physical hardware, if you are using a simulator?
KSC (00:12:02):
Yeah, because it is fun. No. So, that works fine if you are like- Actually it does not work fine.
EW (00:12:10):
<laugh>
KSC (00:12:10):
Even an OLED. Wokwi does not actually do electrical simulation under the hood. It is just faking it. You send signals to an OLED and it just interprets those like the OLED should. But there is no- Most of the things that do not work, are not going to be caught in a lot of simulators like that.
EW (00:12:34):
No, Wokwi is for writing software.
KSC (00:12:36):
Yeah. Yeah, it is great for that. It is awesome.
EW (00:12:38):
If your resistor network is not quite right, I am not sure it cares.
KSC (00:12:45):
Yeah, no, it does not.
EW (00:12:46):
And you definitely can power, "fake power," emulate power to a lot more motors and LEDs than you can if you are actually dealing with current.
KSC (00:12:57):
Yeah.
EW (00:12:58):
Wokwi is for the software. It is not for the hardware.
KSC (00:13:00):
Totally. And I love that. My favorite circuit simulator is Falstad.
CW (00:13:06):
Mm-hmm. Yeah, I use that a lot.
KSC (00:13:08):
Yeah, I do not care what op amp I am using. I just want an op amp. I am just checking my math.
(00:13:14):
And I also think, "Okay, so for Wokwi, why have it at all?" is most parts are not on there, and it is a lot of work to put in a fresh part. And it is just better. I do not know. Hardware is just nice. It is more tactile.
EW (00:13:35):
I agree sometimes. Sometimes it is irritating, but usually what is irritating is the wiring part of it. And it seems like that is the part you are fixing. So, cool.
KSC (00:13:44):
Yeah. Yeah. And I guess maybe it is more of just a me thing, but I always go and get all the most obscure parts I can find. Every LCSE order is like, my cart is filled with parts you had never heard of, just to have samples of everything.
CW (00:14:00):
<laugh>
KSC (00:14:04):
Yeah, I am almost always using something obscure and it would never be like- No one wrote the thing in the simulator for it.
CW (00:14:12):
Right.
KSC (00:14:14):
Yeah. It still works with Wokwi, but- Oh yeah, I think this is probably a distinction we should make in this early on, is that there is the OG Jumperless, the original one, which is the video with the cutting the wires. And then there is the Jumperless V5, which is the more like Apple ad kind of thing.
EW (00:14:34):
The more- I am sorry. "Apple ad"?
CW (00:14:37):
Shiny?
KSC (00:14:37):
The launch video for that does not actually say anything. It is just shows you what it can do, but it is not an explainer like the old video is.
CW (00:14:48):
Got it. Got it.
KSC (00:14:49):
Everything still applies. Crossbar matrix is basically the same. The big thing I did with V5 though is scraping Wokwi. It was very clunky. And with V5, I was like, it all comes down to actually being able to tell you what is going on. So now you can just use it on the breadboard. You do not have to use Wokwi.
EW (00:15:13):
Okay. But you also added the little probe?
KSC (00:15:18):
The probe. Yeah. Okay.
EW (00:15:21):
How do you explain the probe? I cannot explain that one. I am sorry. There is this little thing and it makes the other little thing work, thing, better, thing, something connect thing?
CW (00:15:31):
<laugh>
KSC (00:15:32):
Yeah. Exactly. Nailed it.
EW (00:15:34):
<laugh>
KSC (00:15:34):
Yeah. Okay. So on the original Jumperless, after I shipped- I had already made the hardware. Never thought of a probe. Then I came up with a way to do a probe. I had some extra GPIO pins broken out, and turns out with the Jumperless, you can scan every row a hundred times a second. So it is looking for-
(00:15:56):
So one of those GPIO pins just emits a PWM signal, and it just checks one at a time. Connects it to an ADC. "Hey, do I see a wiggly PWM there?" And then go to the next one. So the probe was a hacked in thing. Then turns out that is the best way to use it, because I do not like looking at a computer if I am breadboarding.
(00:16:20):
So then the V5 was all about that. It was like, "All right. We are going to make the probe part of it. You should be able to use it without touching a computer. Just plug it in for power."
EW (00:16:31):
So now the probe, instead of sending over a serial, "Connect A1 to B3 and C4 to D12," and playing Battleships, now you have this probe that you touch one point and then you touch it another. And then magically they create an invisible wire.
KSC (00:16:55):
Yeah.
CW (00:16:55):
Except it is visible, because it is LEDs.
EW (00:16:57):
It is LEDs.
KSC (00:16:58):
Yeah. <laugh> Yeah. It is invisible until I tell you about it. Yeah.
EW (00:17:05):
And so you do not need a computer anymore.
KSC (00:17:08):
Yeah.
EW (00:17:08):
But one of the things I really liked about the computer and the ideas I got while looking at it, was scriptable circuits.
KSC (00:17:19):
Oh, yes. Yeah, that too.
EW (00:17:22):
Have you done anything? It seems like an odd and magical idea. But I keep trying to fit it into what I know. I think what I need to do is treat it like a toy that helps with imaginary play, that I can think of more big things.
CW (00:17:38):
I still do not think it is a toy. Prototyping is a real thing. Anything that speeds up prototyping or is less error prone-
EW (00:17:45):
True.
CW (00:17:45):
When I was doing my synth stuff, after a while it gets really tiring to have all those wires. Or maybe you want to make a change and reroute something, like you are saying. Once something gets complicated enough, it is sometimes hard to do that, without getting lost.
KSC (00:18:01):
Yeah. So the scriptable thing though, this is actually- I have to thank Peter a lot for that, because he was doing the Excel GUI. We are coming up like, "How do you want to control it?" and stuff like that. I had these single character commands. And then eventually- Now it has got a MicroPython interpreter in it. I was like, "Oh, this is so much easier. I could just have it run Python."
EW (00:18:23):
Yeah, because you have a Raspberry Pi Pico in there, that is controlling all those crossbars, right?
KSC (00:18:29):
Yeah.
EW (00:18:30):
So MicroPython is a very compatible system to put on there.
KSC (00:18:37):
Yeah. And it has its own module for all the very specific Jumperlessy stuff. So, "Connect that. Measure that. Whatever."
EW (00:18:46):
So what have you seen done, or what have you heard done, that the scripting becomes cool?
KSC (00:18:53):
Oh. Yeah. Okay. So we have one guy, George Architech. He is doing high-end audio stuff, but he got really into the idea of, for audio- I had never heard of this. Switched capacitors and switched resistors.
CW (00:19:07):
Aah. Great idea. Yep.
KSC (00:19:09):
Yeah. Yeah. It only actually cares about- You could duty cycle a capacitor, and it would be half the capacitor, if you do it quick enough. So he is doing that, where he is simulating components by switching the crossbars on and off super crazy fast.
EW (00:19:29):
That sounds like a terrible idea. Explain it to me like I am a software engineer.
KSC (00:19:34):
I think what he is working on is that there is a thing where it is like- I have no idea. Some audio person is going to be listening to this and be like, "Dude, what are you talking?"
EW (00:19:42):
I think Christopher should have to explain it.
CW (00:19:44):
Well, I was just thinking of guitar pedals or something like that, where you have a filtering network and you want to remove things completely, or add new things without having necessarily a pot or a variable capacitor.
(00:19:57):
Or like you are saying- What I was not considering what you are saying, where you are rapidly switching something in and out. I am not sure what that would- That is terrifying, but probably sounds really cool. It probably can make some very strange effects by actually altering the circuit at audio rate, or close to it.
KSC (00:20:15):
Yeah, I can do 500 kilohertz of on and off.
CW (00:20:19):
So the switches are so fast.
KSC (00:20:22):
Yeah. Okay, so I think what he was looking at, and the reason this was relevant, was they do these things- There are these really expensive machines you would put on some particular pedal, some vintage pedal they do not make anymore. And it will twist all the knobs and characterize it, so you can simulate it into your audio software.
(00:20:42):
But it physically twiddles all the knobs and everything. So he wanted to do something that would do that more softwarey. I think the switch capacitor was so he can twiddle the knobs without a robotic arm.
EW (00:20:59):
Okay.
CW (00:20:59):
Probably building an impulse response for some sort of convolution thing.
KSC (00:21:03):
Yeah.
CW (00:21:04):
And then put that into software, and now you have got a virtual pedal that does the same thing. Yeah.
KSC (00:21:10):
Yeah.
EW (00:21:14):
I do not know if Kevin has noticed, but the next question on the outline was, "So I am a software engineer and I want to make a guitar pedal." And I want to point out-
CW (00:21:21):
Oh, I did not even notice that.
EW (00:21:22):
That Christopher does not read the outlines. <laugh>
CW (00:21:27):
<laugh> It is just a natural place for us to go.
EW (00:21:29):
It really is. It makes a lot of sense.
CW (00:21:31):
So my first question about making guitar pedals with this is, do the networks do anything bad for impedance or anything else?
KSC (00:21:41):
No.
CW (00:21:41):
Oh, good.
KSC (00:21:42):
Someone actually was nice enough to characterize these. Yeah, he had all the equipment at work. I would not even know where to start on how to- Those high-end scopes and stuff. So a breadboard just without any connections, like a 3 dB roll-off, you get half the signal out, is 13 megahertz.
CW (00:22:03):
Oh. Okay.
KSC (00:22:04):
And then doing it through a Jumperless is eight megahertz.
CW (00:22:08):
I do not-
KSC (00:22:09):
Yeah, it is so high that I am actually- It is surprising. Yeah.
CW (00:22:13):
Oh. That is great. Okay.
KSC (00:22:14):
It is more than anyone would expect you could actually push through a breadboard in the first place. Finding out that it is, oh my God, that is ten times higher than I expected.
CW (00:22:25):
Yeah, I would think it would-
KSC (00:22:26):
The Jumperless takes off half that almost. Yeah.
CW (00:22:29):
I would have thought breadboards crap out at, I do not know, a couple megahertz.
KSC (00:22:33):
It is actually, you can do it on the thing because it lets you do PWM. And I did not cap the PWM. So from the Pico you can do 62 megahertz coming out of it. And you could watch it on your scope, the signal getting smaller and smaller as it gets- Yeah.
CW (00:22:48):
Oh, okay. Okay. Huh. Well, I am thinking of a lot of things to do with this now. <laugh>
KSC (00:22:56):
That was the point, is to have a fun little- There is a lot of weird stuff you can do because you have not thought of all of them.
CW (00:23:03):
Yeah, because you could have a- I am totally spitballing, but I could think of an oscillator, right? Some little synthesizer kind of thing. You could probably fit a simple voltage control oscillator on this breadboard. But then you could mess with the components.
(00:23:19):
People do things called "circuit bending," where they will take an old kids' toy or something that made sounds and they will go in there and pop resistors off or caps or replace them. Go in and put switches in places and turn them into synthesizers and make weird sounds.
(00:23:32):
But if you can modulate the actual components themselves being rerouted and hit audio rate, you could do some really weird stuff.
KSC (00:23:44):
I know. I wish I did more like I was a synth guy. I have to just ask the bunch of synth people in my Discord. This is a thing you do? How does this? Yeah.
EW (00:23:57):
Are there other areas that have come up like that? Music and synth and guitar pedals all seem like they would be a natural fit. Are there other areas of natural fit you have found?
KSC (00:24:10):
Here is the most insane one. Dyslexic people. Half of the people that use this a lot are dyslexic.
EW (00:24:18):
Why?
KSC (00:24:18):
Something with dyslexia and wires, it does not work well.
CW (00:24:24):
Oh. Okay.
KSC (00:24:25):
In your head. That is what they say, it actually is like, "Oh, I can actually see a circuit now." I would have never expected that. Yeah.
EW (00:24:38):
On the other hand, color-blind folks probably hate you.
CW (00:24:42):
Well, you can use colors that are not conflicting, right?
EW (00:24:47):
Yeah.
CW (00:24:47):
Yeah.
KSC (00:24:48):
Yeah. If someone asked, I would do that. Which kind of colorblindness? It is always so funny with colorblindness though, because people are like, "Oh, that is not color-" There is a type of colorblind that will make any two colors not recognizable. Oh yeah, you just need three options.
EW (00:25:09):
Okay. So dyslexic folks find it easier to build hardware because you do not have to connect A4 to B12, because now you just connect this to that and there are no moving letters in between.
KSC (00:25:27):
Yeah. I think it comes down to the wires being flat on the breadboard, versus them being a- I guess this is true for all of us, but the little loopiness, I guess.
CW (00:25:38):
I confess, I do not know how dyslexia works, but it probably existed before writing. So it is probably some spatial thing that is confusing for people. It is a brain spatial thing. So I can understand being- I am overwhelmed by a breadboard of sufficient complexity and wires. Yeah, I can see stuff getting flipped in your mind impossibly if it is too complicated.
KSC (00:26:03):
I actually remember in school- I did not really go to school for engineering, but I did take a bunch of classes. But I just remember being so amazed that the teacher during a lab could walk up to someone's breadboard and be like, "Oh yeah, yeah. That is wrong." Just parse it just walking by, is so crazy. I am still amazed by that.
EW (00:26:29):
I suspect that is practice <laugh> more than anything else.
KSC (00:26:32):
Yeah. Absolutely.
CW (00:26:33):
"I have seen this before. I know what mistakes you people make."
EW (00:26:39):
Have you seen anything else with the scriptable circuits?
KSC (00:26:42):
Yes, but I am totally drawing a blank.
EW (00:26:44):
That is cool.
KSC (00:26:44):
Let me see what people are working on.
EW (00:26:46):
Well, you also mentioned AI circuits.
KSC (00:26:49):
Oh! Yeah! That is crazy now. What I mentioned there- This was pre-AI. This codebase was written before Claude. But it was evolvable hardware. I do not know if you have ever heard of that.
(00:27:04):
It is this thing where they are putting random gates into an FPGA and then they would make a tone discriminator or something. It would put a pin high if it was 400 hertz and low if it was 100 hertz and gave it nothing else. That was it. They would give it a tone and then they would see what comes out with just a completely random set of gates. And so then they would just run that as an evolutionary algorithm for long enough and it would totally do it.
CW (00:27:35):
It reminds me of genetic programming.
EW (00:27:36):
Genetic programming.
CW (00:27:37):
Yeah. Genetic programming.
KSC (00:27:37):
Yeah, exactly. But yeah, doing that in hardware is so cool. Yeah.
CW (00:27:44):
So you could totally do that. You could put a bunch of parts on and just, okay.
KSC (00:27:49):
Yeah. Nowadays, now with LLMs, they are amazing at this stuff. I have a whole page in Docs for people who want to use LLMs, because it is really good at that. If you are working something like, "Hey, something is wrong." It has a bunch of calls for it to check measurements on every place. It could check all of your assumptions just by querying the board, which has been amazing.
(00:28:13):
We just had someone in the Discord was like, they put down a transistor and it did the whole transistor tester thing, told them the gain and everything. LLMs are getting way less horrible at hardware, apparently.
EW (00:28:31):
Going back to the tool part, being able to measure resistors and transistors and other inputs to circuits that need to be precise, that could be a tool. I do not know that it is a tool that is better than what already exists, but it may be more general, which sometimes can be better for small shops.
KSC (00:28:55):
Yeah. Yeah. There is also the stuff like EEPROM dumpers. Those make Jumperlesses look cheap, actually.
EW (00:29:06):
Very much so.
KSC (00:29:08):
But yeah, it is like, "You just twiddle the address lines, and read the data lines."
EW (00:29:14):
There are times if you were reverse engineering something, that it would be easier to do this sort of connector.
KSC (00:29:22):
Oh yeah. I do that all the time, because I always have, as I said, the most random LEDs, and just getting the pinout of those. Sit there and just try it. You just keep connected till stuff lights up.
CW (00:29:37):
Those are really considered brute force electronics.
KSC (00:29:40):
Yeah.
CW (00:29:41):
"I do not know how this works. I will connect it in every possible permutation."
KSC (00:29:44):
Yeah. It works great.
EW (00:29:46):
But you also do have some current-
CW (00:29:50):
Limits?
EW (00:29:51):
Prevention of actual fires.
CW (00:29:53):
Oh.
KSC (00:29:54):
Oh, yeah. Yeah. So the crossbar resistance does take care of that, which is very nice. I wish there was zero resistance and that has been my whole thing. But yeah, so it is like one connection is about 80 ohms. How it gets that lower is it will just fill up all the available routing space with parallel paths.
(00:30:19):
But even with 20 ohms, you cannot really blow anything up. Even if you just hook up an LED straight to power, it is not going to blow up, which is nice.
EW (00:30:30):
You can source enough current to turn on LEDs.
KSC (00:30:35):
Yeah.
EW (00:30:36):
Or do I need to add a MOSFET for everything?
KSC (00:30:38):
Oh no, it can totally- I think it is, I do not know, it is something like, it is like 300 to 500, 400 milliamps before the power rails start to really sag. Yeah, I can push some current. I am always surprised that a charge pump can do that.
EW (00:30:57):
Do you need the charge pump for what the user is doing? Or do you need a charge pump for the, what was it, 150 LEDs you have on there?
CW (00:31:04):
450.
KSC (00:31:06):
Yeah. Oh no, those are powered just straight off five volts. But the charge pump is- So how crossbar switches are is they can only switch stuff within their power supply envelope. Their plus and minus that sets your window of what-
(00:31:22):
Anything above that, they start being very weird. They do not blow up anymore, but they will leak stuff to the rows next to it, just completely undefined behavior. So you need- I have a plus-minus 10 volt power supply on the board powering all the crossbar switches, which is also the voltage rails and everything. So that gives you plus-minus eight volts on the rails, stuff like that.
EW (00:31:52):
I have never used a crossbar switch. My mental model of them is actually really similar to my programmable logic array mental model. Which is you tell it to connect two things and it does and it tries to pretend it is a wire. And there is not a lot more to them, that they carry whatever current they are supposed to carry, and they truly try to act like wires.
(00:32:20):
Is that a good mental model? Or are there things I should know about?
KSC (00:32:25):
Yeah, no. Just an analog CMOS switch, like that. It is not sourcing it from the crossbars. They do not draw basically any current, like all the crossbars. It is actually just- It has to pass through some silicon, but no, it is just like you are turning a switch on in an ideal sense. Yeah.
(00:32:50):
If you want how it is hooked up, it really is named after it being a bunch of crossed bars.That is where they came from. Yeah.
CW (00:33:01):
What are they usually used for?
EW (00:33:02):
Power switching?
KSC (00:33:07):
So the original crossbar, you can find old pictures of them from Bell Labs. These were for telephone networks. This is how they would route calls from-
CW (00:33:15):
Right. Okay.
KSC (00:33:17):
So you would have those crossbars laid out and before they even had the electronic control, you would get this clip that would just grab both of them. Where they crossed, it was like, "Okay, that is coming from Louisiana, and that wants to go to this next place," and you just clip them together.
EW (00:33:33):
Like the old operators that actually physically moved the connectors.
CW (00:33:38):
I remember there were crossbars in some of the high-end networking products that I was adjacent to, but I do not remember what they- I think they were for switching one port to it literally, for establishing a permanent connection for one port to another port.
KSC (00:33:56):
Or it is like if you wanted to flip them, when the whole crossover cable thing happened, you can just do that.
CW (00:34:02):
Oh. Okay.
EW (00:34:03):
I still have an ethernet crossover thing.
CW (00:34:05):
You should throw it in the trash. It does not matter anymore.
EW (00:34:08):
I know, but.
KSC (00:34:08):
Thank crossbars for that.
EW (00:34:09):
But I actually got a pink one, so that people would stop stealing it. And I would always know which one was mine. So I cannot get rid of it.
KSC (00:34:16):
That is the correct move is make your stuff pink. No one will take it.
EW (00:34:20):
Often works for me. Okay. So I make a guitar pedal and I like what it is doing, but now I need to make a schematic. Does the Jumperless help me with that at all? Is it a question that you keep getting? Are people like, "Okay, I want to use my Jumperless now. I have to move it off of here. Now what do I do?"
KSC (00:34:48):
Yeah, that is an interesting one. Okay. So when you gave me this question, I started doing that. I made a schematic capture-
EW (00:34:56):
<laugh>
KSC (00:34:58):
Thing in Jumper IDE. Yeah. I was like, "Oh yeah. You could just do that now."
CW (00:35:02):
Is that podcast sniping? What is the term for this?
KSC (00:35:06):
Yeah. It is funny because, yeah, Jumperless, a lot of people use it as a teaching tool, but I am such a curmudgeon about this stuff where I am like, "Dammit, KiCad should be hard. You should take the time to do it. You will learn so much more."
CW (00:35:25):
"Builds character."
KSC (00:35:28):
Yeah. You will think about it. If you make it too easy, you will miss stuff, and then you will order boards and they are all <beep>, because you never were quietly routing it and thinking like, "Oh, should that be there?" All that. I do not know. I guess that is different people's style.
(00:35:44):
But there is also just not the information to make a schematic. So now how I did the schematic capture is it has to ask you where the chips are, because it only knows where the wires go.
CW (00:35:56):
Right. Of course.
KSC (00:35:58):
Yeah. Yeah. It is an interesting problem of what information is contained in a netlist and what is in a schematic.
CW (00:36:08):
But wait, you have got all this modulation stuff. Can you not modulate each pin?
EW (00:36:13):
No!
CW (00:36:14):
And then check the response and then have a database of parts? It is like, "Oh, that looks like how a potentiometer would respond to a PWM of five megahertz." No, I was kidding.
KSC (00:36:24):
I have started that project.
CW (00:36:26):
Oh my God!
KSC (00:36:27):
I should finish that. Because yeah, of course. You could. Yeah. It would be awesome.
EW (00:36:32):
You mentioned education, and I can see how this would be easier for instructors who cannot just walk up and see what is wrong with your circuit. Being able to look at the typed instructions would make a lot easier, or if it dumps the pinout.
KSC (00:36:48):
Yeah.
EW (00:36:50):
Is it mainly for educational use? Or? Who are you selling to? Educational institutions? Makers? Companies? Who is buying them?
KSC (00:37:04):
People who saw them and thought they were cool. Yeah, I have no idea actually. Honestly, I wish I collected that kind of data, but I do not. It is just engineers people. I do not know if people are using them for education. I think it is a lot more hobbyist stuff for that.
(00:37:23):
A lot of software people wanting to get into hardware are getting into these. A lot of it is just it being so extra as just how it is. I have heard that a lot, where it keeps people engaged to play with it. It will keep you going doing it.
EW (00:37:44):
There is a point in every breadboard project where I just want to toss it. Whether it is for work or for my own amusement, it is the one wire that breaks the camel's back. It is too many. This is it. So I get that.
(00:38:04):
So you are totally a one person operation.
KSC (00:38:13):
Yeah. Yeah, just me.
EW (00:38:16):
Do you spend a lot of time in marketing? Do you mostly do designing? Is it fulfillment? Is it supporting? You have really good communication for the designs on GitHub and the videos and all of that. So that part I know you spend a lot of time on. But where do you end up spending your time, and where do you wish you spent more of your time?
KSC (00:38:45):
I am always bouncing around all of those at once. I always have Discord open for anyone asking a question. I am like, I will be right there immediately. Doing support and then getting orders and shipping them, that is always fun.
(00:39:02):
Where I wish I spent more time? I do not know. Right now specifically, I guess this gets into the next thing coming, but I really need to finish designing my next one. I have all the parts, kind of. I just need to say it is good.
EW (00:39:24):
Is this the Barnacle Board?
KSC (00:39:26):
Yeah. Okay. Oh! All right. This kind of talks about what you are doing earlier. Okay, you have a breadboard and how do you make it a schematic? All right, I still cannot help you there.
(00:39:35):
But the newest one, I am making dumb breadboards too. This is the Jumperless optional part. It is just a normal breadboard with all of the things. Okay, the power rails are continuous and lined up, because I made six-hole spring clips that fit together. They are actually on their way right now, just so exciting. Yeah. Had spring clip stamping molds made.
CW (00:40:05):
Wow.
KSC (00:40:05):
But the top of the breadboard, instead of it being like the plastic part of the molded shell, is a protoboard. It is a PCB, so you-
EW (00:40:15):
Okay. These are the normal five holes are connected and there are a whole bunch of columns that are not connected. And I have seen these being handed out as business cards. We have a stack of them.
CW (00:40:31):
Oh, just the perfboard.
EW (00:40:32):
The perfboard and you solder it, but there are already connections on the perfboard.
CW (00:40:35):
Yeah. Okay. Yeah, yeah, yeah.
KSC (00:40:36):
That mimics a breadboard.
CW (00:40:37):
Yeah.
EW (00:40:38):
Yeah.
KSC (00:40:39):
Yeah. It is one of those. It has countersunk holes. Working on a breadboard with a flat surface without some countersink-
CW (00:40:47):
Oh, right.
KSC (00:40:48):
JLC will do countersunk plated holes for you. Amazing.
EW (00:40:52):
Huh. That would make life easier. Okay.
KSC (00:40:54):
Yeah. So yeah, you breadboard on this optionally dumb breadboard. And when you are ready to go, you are like, "All right, I like this. I made a pedal." You just solder it right there in place.
CW (00:41:04):
Oh, that is a good idea.
KSC (00:41:07):
Yeah. You remove the protoboard, and just swap it out. So you now have your breadboard is permanent, and you could throw that in a drawer instead of your whole breadboard.
EW (00:41:17):
Instead of your whole Jumperless.
KSC (00:41:19):
Yeah. Yeah, exactly. And then the next one, that will be- It will just be that. There is just a regular breadboard, but it will have a port. It has a port on the top with a card edge connector, and on the back.
(00:41:34):
So you could basically- You have that board and you can stick a Jumperless as a module- It is just like a PCIe card or something. And that is all of your crossbar switching and your LEDs and stuff are all controlled on that. And that is completely separate from your breadboard. So you can have a few going at the same time, different projects and whatnot.
EW (00:41:58):
I am sorry. I do not understand.
KSC (00:42:01):
Sorry. It is hard to- Yeah. Okay. So you saw the Barnacle Board. It is just a breadboard with a big card edge connector on the very top of it.
CW (00:42:09):
Okay.
EW (00:42:09):
Okay. So I am good with that. I understand. Given no Jumperless, I could wire a circuit together on this like I would a perfboard with the connectors already there. I have seen lots of those. Okay. But now it has got the side thing and it connects somehow to the Jumperless. It becomes a Jumperless?
KSC (00:42:31):
Okay. So that port on the top of it is mapped one to one with every row in the breadboard. So it is a 70 pin header and that is like, all right, pin one is the top row. So it is connected to all those rows. So anything you plug into the top can be electrically connected to every row on the breadboard below it.
CW (00:42:56):
So you can have a project and then the Jumperless attaches to it, and does more stuff or routes the rows that are on the other board.
KSC (00:43:07):
So it could show you that circuit. It could do all the LED stuff. You can wire it with regular wires, and then it could show you one at a time where to put physical wires if you want.
CW (00:43:17):
Okay.
KSC (00:43:17):
Then you can remove the Jumperless as a module. So now it is just a breadboard again. And then even further still, you could solder it all together and remove the soldered protoboard.
EW (00:43:29):
Could I then plug back in the soldered protoboard, and have it tell me if I did it right?
KSC (00:43:34):
Yes. Yes, you can.
EW (00:43:36):
You are just making that up right now, are not you?
CW (00:43:37):
<laugh>
KSC (00:43:38):
I am. I am. Yeah.
EW (00:43:39):
<laugh>
KSC (00:43:39):
<laugh> Good idea. Yeah. Oh my God. Yeah. That one comes down to style. It is really weird. The card edge connector part is ugly. It is really killing me trying to figure out how to do this where it does not look like crap, just visually.
EW (00:44:02):
Yeah. No, I can see that. The Jumperless, it has style. It has opinions about how it should look. For all that I am not in favor of blinking lights, I do appreciate that there is some coolness there. You are hitting the black and lights and synth wave, almost, aesthetic. But the Barnacle is aptly named.
KSC (00:44:34):
Yes. You call that the tops barnacles. Yeah? Yeah.
CW (00:44:39):
Some things just need to be functional.
KSC (00:44:42):
Yeah. No! That is not. No.
EW (00:44:46):
No?
KSC (00:44:47):
Strongly disagree. Not everything needs- No. Yeah. It has to be functional too. It just also has to look good.
CW (00:44:52):
I think it looks cool. It is still black. It is still has that sinister- Does not look like an '80s piece of electronics. It is not green.
KSC (00:45:01):
Yeah. Yeah. Right? Yeah, that is just that SLS just looks so good. The 3D printed nylon that they do. It is just so strong. It is just amazing as a 3D printing material.
CW (00:45:16):
I do think if somebody has that fear of holes thing...
KSC (00:45:19):
Oh my God. Yeah. I have some worse ones for that one too. Yeah.
EW (00:45:27):
Okay. So when does the Barnacle come out? Are you doing Crowd Supply again? Is this in conjunction with Jumperless V6?
KSC (00:45:36):
Yeah. I think that is all going to be mashed up into one. Yeah. I am going to say it here and I am like, I have no idea. When it is done, I will launch it. Yeah. Probably in a few months it will probably be ready, and there will be some way to explain what it does better than what I just did now. Yeah.
EW (00:45:58):
It is okay. This can be a practice way.
KSC (00:46:00):
I did. I made a bunch of prototypes and handed them out at DEF CON. That was fun. Just as regular breadboards is fun.
EW (00:46:09):
When you do a V6, are you going to try to reduce the cost? Or?
KSC (00:46:15):
No. No, not at all. No. Okay. That is what this thing is about. So the idea would be you can- The problem is these actually cost a ton to make.
EW (00:46:26):
Just in LEDs.
KSC (00:46:29):
Yeah. That is like 20 bucks a BOM cost, just LEDs. It is crazy. And then the spring clips, I have them custom-made and they are like, I do not know, 10 cents a piece. They are so good though. They are so buttery smooth. I will not compromise on that.
CW (00:46:51):
You can destroy the ones on cheap breadboards with certain parts, right?
EW (00:46:55):
Yes. I can say for sure you can.
CW (00:46:57):
Bigger chip pins and stuff, because they are expecting resistors.
EW (00:47:01):
Pushing them in and up. And if you get big little dev boards and then you try to jiggle them in and out, yeah, it goes bad.
KSC (00:47:09):
No, they are so good. It is phosphor bronze. It is what they should be made out of, but obviously there is a reason they do not because it is expensive.
EW (00:47:20):
Has anybody come to you and said, "We want to use your amazingly buttery smooth pins"?
KSC (00:47:26):
Yeah, that is why I started Barnacle Board was people just kept asking me like, "Could you just make a regular breadboard?" I use regular breadboards. Yeah. And I cannot go back. I have a bunch of prototypes that I use for my own stuff. If I am doing- Most of the stuff I do is tube circuits. That will blow up a Jumperless, so I cannot use those.
(00:47:47):
But I just use my own breadboards all the time. I cannot use a regular one. They are annoying now. So that is what people wanted. They wanted a luxury breadboard. Yeah, that is what I am doing.
(00:48:04):
That was the whole idea of having that PCB at the top removable. I cannot really get around it being expensive, but I can make it so you only need one. You do not toss the whole breadboard in the drawer.
CW (00:48:19):
Got it.
KSC (00:48:19):
You could have that one and then you can have different projects even going at the same time.
CW (00:48:24):
It is your little template thing. And then the other part is the consumable that is cheap. That is a great idea. I like that idea. Huh.
KSC (00:48:31):
Yeah.
EW (00:48:33):
Okay. Now let us talk about the thing that makes no sense at all.
KSC (00:48:37):
Okay.
CW (00:48:37):
Good.
EW (00:48:40):
This is totally open source.
KSC (00:48:44):
Yeah. Oh yeah. Oh yeah.
CW (00:48:47):
You always ask this question.
EW (00:48:49):
I still do not understand how you can make a living.
CW (00:48:53):
Because nobody is going to build this, even if it is open source.
KSC (00:48:56):
It is amazing. Yeah. No one has. I want them to so bad. Seriously, I am like-
CW (00:49:02):
It is like open sourcing a jet engine. Yeah. Okay.
KSC (00:49:05):
That is the problem. Yeah. There are just so many lower hanging fruits with better margin than a Jumperless. You have to want to make this. It has to be a crusade. It cannot be like a thing that is like, "Oh, we are just going to produce these and sell them on Amazon." But if anyone is listening to this and wants to do that, please do. I will help.
EW (00:49:36):
So basically this is something you want, more than you want to have be your career?
KSC (00:49:41):
No, it is just like- Yeah, I get enough from selling them. I just want there to be some adoption. There are a thousand of them out there, and I am assuming most of them are in someone's drawer who has never touched it.
(00:49:59):
But this thing could do so much stuff. If there are enough people actually using it, even if it was not something I sold, they could just take my firmware and put it on there. It is fine.
EW (00:50:14):
Okay.
KSC (00:50:14):
It would just be cooler.
EW (00:50:16):
So let us say we have some listeners and they have them in their drawer.
KSC (00:50:21):
Whip it out right now, guys.
EW (00:50:24):
What do you want them to do with it this weekend?
KSC (00:50:27):
Oh, that is a good one. Okay. Whatever you have wanted to do for a while. I think you should start with a 555.
CW (00:50:34):
Mmm.
KSC (00:50:34):
I think that is a good way to get comfortable. If you go on Jumper IDE, first update your firmware. I push firmware updates every day. Go on Jumper IDE, go to the JumperNet Registry and open 555_Guided.
(00:50:49):
It will guide you through putting a 555 circuit together, one at a time like, "Here, tap there. Tell me where it is," and it will do it for you. Once you get that, you will keep going.
EW (00:51:05):
How many components is that?
KSC (00:51:08):
555, like three.
EW (00:51:10):
Yeah. That is what I thought. Okay.
KSC (00:51:11):
Yeah. Yeah, it is very small. No, you should do anything you want to do. Just make a circuit. Connect an Arduino to the coolest little display module you have laying around and some buttons, and make a little UI in MicroPython and you are all good.
EW (00:51:34):
We mentioned MicroPython and the Raspberry Pi Pico that drives the crossbars. One can program for that. You can program the crossbars and then you can write something to do beyond that. Or scriptable crossbars so that you can change your components.
KSC (00:51:54):
Yeah.
EW (00:51:55):
But you also have a header that is off the main breadboard, a fixed header that is a Nano footprint.
KSC (00:52:04):
Yes.
EW (00:52:07):
If what you are trying to get to is a Nano controlled circuit. Or, the Nano footprint has a whole bunch of different boards. When you say, "Pops a MicroPython on there," you have two places to put it.
KSC (00:52:23):
Oh yeah. Yeah, you do. MicroPython, that is running on the Jumperless. You can- Oh my God, it goes very deep. You can actually tag things over UART from the- The Arduino can control the Jumperless too, if it sends something over UART.
(00:52:43):
But yeah, MicroPython is running on the RP2350. It has access to all the connections. It has ten GPIO that you can put anywhere you want. It has five ADCs, two DACs, current sensors. There is all MicroPython API for that.
(00:53:02):
So you do not need the Arduino, but if you wanted to put it on something else, yes, you would program on the Arduino. And that can be done through the same wire, because Arduinos will take code over UART. So you can just flash it.
CW (00:53:18):
Right.
KSC (00:53:19):
Yeah. Jumperless shows up as four ports. So one of the ports is like one of those serial adapter things. It just acts like that.
EW (00:53:30):
Oh, I was thinking the Nano would be for when I wanted to control the circuit, when I had taken the system off the Jumperless. Was that not the use mechanism?
KSC (00:53:45):
The Nano header just actually acts like more breadboard, for the most part.
EW (00:53:49):
Okay.
KSC (00:53:50):
It has got a couple of the hardwired pins for power. But yeah, that would be what you would do. So if you wanted your Arduino- Yeah, you are totally right. You would take it off of the Jumperless and then it would work that way. So yeah, if that is your intention, not to just make it on the Jumperless and then like, "Hey, it works," and then move along. Yeah, you would probably want to program on the Arduino.
(00:54:14):
You can actually do that from Wokwi, which is actually really fun. So the app scrapes Wokwi, but also scrapes the code and compiles it. So it will flash the Arduino as you type. As you enter code in Wokwi, you can just flash it from the web browser, which is very fun. Through the same wire. Yeah, you do not have to plug in the Arduino. Crazy that works.
CW (00:54:42):
Elecia looks perplexed.
EW (00:54:44):
I am not sure that is a good idea, but that is okay. I do tend to type faster than I can think.
KSC (00:54:50):
Oh. Well, yeah. You have to hit save. That would be crazy for C to just try to compile every keystroke. It is just like, "Error. Error." But yeah, that is a workflow that people like, I think.
(00:55:06):
Basically just making Wokwi real. Because you could have the diagram and your code right there, and you can- Doing all your Arduino stuff right there and the wiring, which will just be copied over to the Jumperless.
EW (00:55:22):
Then you get to make that emulation into a real hardware. And then with the Barnacle, you can lift up that hardware, solder it together and connect it back to an Arduino. And now you have a circuit. Now you have a guitar pedal!
KSC (00:55:40):
Yeah. Finally, we got there.
CW (00:55:41):
<laugh>
KSC (00:55:41):
That took us a while to make this guitar pedal. Yeah.
EW (00:55:50):
Okay. So people who have one in the drawer, you try the 555 timer. Then if you are interested in synths, you go look up a couple of circuits, because that is going to be easier to change there. You do not end up with the mass of wires.
(00:56:05):
And you can do the oscillator to oscillator thing, that usually ends up with fun sounds. What oscillator to oscillators are the best? It is not the triangle to triangle. It is like-
CW (00:56:16):
You can try all kinds.
EW (00:56:16):
Square to sine and-
CW (00:56:18):
Basically, you are doing FM modulation at that point.
EW (00:56:22):
Yeah. But it is audio range, so it is just weird.
CW (00:56:24):
Square wave. Start with a square wave and modulate it with- Yeah, you can do all sorts of stuff. What sounds good.
KSC (00:56:31):
Wait, so you are saying. Yeah, I do not know what that is actually. Is this like you just have two oscillators running on top of each other?
CW (00:56:38):
Basically, yeah, you take one as the input to the other.
EW (00:56:43):
If you think about a sine wave, not with another sine wave added to it, but with another sine wave multiplied. So then your sine wave has wiggles on it, and that is frequency modulation.
KSC (00:56:55):
Ahh.
CW (00:56:56):
It adds all sorts of weird harmonics and stuff. Then if the rates change, the waveform can evolve. Yeah, it is fun.
EW (00:57:04):
Dogbotic is not offering classes anymore, but they still have that book which was really good.
CW (00:57:07):
They still have the book, and they still have in-person workshops they are doing. I think the synth class may still exist. They were doing just the synth class still.
EW (00:57:20):
That is making your own synthesizer. That is the one you did, where -
CW (00:57:22):
No, I did the drum machine, which is very similar.
EW (00:57:25):
Okay.
CW (00:57:26):
But similar circuits. Yeah, similar circuits.
EW (00:57:28):
But that had a big breadboard, that would have been vastly simplified by this.
CW (00:57:34):
Well, the circuits, there were a lot of circuits, so it would cost me a fortune to do it all on Jumperless. I think I had five or six standard breadboards at the end of it, that were all wired together. But this would be like- I was just thinking about using Jumperless to
(00:57:52):
Because synthesizers tend to be modular, you have pieces you want to switch in and out, or you want to connect in different ways to each other. So you could have normal breadboard modules or the Barnacles, and then have the Jumperless be just a router between them.
(00:58:08):
Or have some filtering on it or something that fits, and then connect those in different ways and experiment that way. Because you are not just necessarily experimenting with the individual components. You can have circuits talk to each other.
KSC (00:58:26):
Yeah. And you could do- How I tend to use it is I am usually doing logic, 7400 logic circuits and stuff, but I do not need the clock circuit anymore. That is what you use MicroPython on the Jumperless to do.
(00:58:43):
So you can almost record what would be coming out of the module to the left, and capture what would be going out of it. And so yeah, you can do them one at a time. You can basically simulate the other modules if you need to.
CW (00:58:59):
Yeah. There is a ton of stuff you can do with synths, that a purist doing analog synths is like, "Well, I want to make a voltage controlled oscillator, but I have something over here that generates the voltage and analog."
(00:59:10):
But you can replace all that with the Pico and go out of DAC to do the oscillator control stuff. Similar things can be pushed into the microcontroller. Nobody is going to be able to tell a difference.
KSC (00:59:23):
Yeah.
CW (00:59:24):
The important part is the analog oscillator. Yeah.
KSC (00:59:27):
Yeah. It is tubes or bust. We are not doing this- We have to do tubes. That is my dream of the Jumperless, being able to handle high enough voltages to do tubes. Because everything I do that is not Jumperless, is tubes.
EW (00:59:45):
I could see this being useful. Like, you have an electronics kit that has a whole bunch of different parts and it comes with the book. Christopher is looking at me like I am crazy, but it is in the garage. Probably has been since he was 15.
CW (00:59:58):
Oh. That old thing with the components laid out, and you wire between the components?
EW (01:00:02):
Yes.
CW (01:00:03):
For kids. It was a kids' electronics kit.
KSC (01:00:04):
With the little spring things?
CW (01:00:05):
Yeah, the spring clips.
KSC (01:00:06):
Oh, those are awesome. Yeah, I had one of those.
EW (01:00:09):
But that also tends to get pretty messy pretty quickly.
CW (01:00:13):
Yeah.
KSC (01:00:13):
Oh my God. And taking stuff out of those little springy things.
CW (01:00:17):
It was a nightmare.
KSC (01:00:17):
Oh my God. Yeah.
CW (01:00:19):
Because it was always stranded wire, too.
KSC (01:00:23):
Yeah. I need to do that as a bit, of hook up a Jumperless just underneath one of those. On their back, just put whole bunch of wires into it. Yeah.
EW (01:00:33):
But I could see- I was headed towards it. You said hook up a 555, which is a pretty common introductory project. I could see going through the whole book, and not having the wires be the defeating component.
CW (01:00:51):
Holding you back. Yeah.
KSC (01:00:52):
Yeah.
EW (01:00:54):
It would be so much better. Anyway.
KSC (01:00:58):
I also have examples on the Jumperless. It has a file system. If you go into /python_scripts/examples/ it will just give you a bunch of little projects to do.
CW (01:01:11):
Oh, cool. Okay.
KSC (01:01:12):
Yeah. Yeah, I figured that would be helpful.
EW (01:01:16):
There is so much more that I did not find. I think-
KSC (01:01:20):
It goes too deep. I am sorry.
EW (01:01:21):
It goes too deep. It is very cool. Yet it is still a little hard to explain.
KSC (01:01:28):
Yeah, especially without visuals, because it is a very, very visual thing. I could be showing you.
EW (01:01:32):
There were a lot of blinking lights.
KSC (01:01:34):
Yeah. Nothing really does anything that makes noises.
EW (01:01:44):
You started this- Wait a minute. I guess the question is, did you start this for the Hackaday Prize in 2023? Or had you already developed it, and that was just one of the places you showed it off?
KSC (01:01:54):
That got me- It was good timing, because I was already working on it. I had I think two rounds of prototypes at that point. Then the Hackaday Prize came and I was like, "Okay, this is time to finish it. There is finally a deadline." Yeah, I was already doing it.
EW (01:02:15):
Did you win any prizes or any of the subcategories?
KSC (01:02:18):
Yeah. 2023, I got second place.
EW (01:02:22):
Oh! Wow!
KSC (01:02:23):
20 grand funding for more Jumperless stuff. Woo-hoo.
EW (01:02:28):
Sweet.
KSC (01:02:29):
Yeah, it was awesome. It was the last one they ever did. It was great.
EW (01:02:33):
The Supercon does have their call for proposals open until August 26th.
KSC (01:02:38):
Oh, that is good for other people to hear. Yes. I will never do a talk. It is so terrifying. Even the concept of getting on stage and pontificating like that, would drive me crazy.
EW (01:02:51):
Okay. Then I hate to tell you this, but the audience for this show is a lot larger than that.
CW (01:02:57):
It is too late. He has already done it.
EW (01:02:58):
He has already imagined them all naked. That is just weird.
KSC (01:03:00):
Yeah. Yeah. We will just do it over the phone. I will just do a talk, but not even be in the same room. I will just be talking into a microphone in front of my computer.
EW (01:03:12):
Oh, one of our listeners asked me to ask you about lenticular silk-screen shirts.
CW (01:03:21):
What?
KSC (01:03:21):
Oh yeah. I did those for last Supercon. It was a test and they barely worked, but I think we got it.
CW (01:03:30):
Wait, those are the things where it is like a little animation if you see it a different angle?
KSC (01:03:35):
Yeah.
CW (01:03:35):
Yeah. Okay.
EW (01:03:36):
Oh. Okay.
KSC (01:03:39):
We did a test run of these. I do so many Jumperless shirts. I just like designing shirts and my friend owns a print shop, so we are just always- Almost any event, they will have a bunch of Jumperless shirts to give out.
(01:03:50):
But at last Supercon, we did this test of- You can use puff ink. Use a layer of puff ink as lines, and then you print colors on top of it. Then when you puff it in the dryer, it will become like how a lenticular thing is.
(01:04:11):
Ideally, it is like a little vertical thing with printing on either side. So you look at it from one side and it is one image, and you look at it from the other and it is a different image. Yeah, that is what I made at Supercon. I did make lenticular shirts. They did not work that great, but the next one will.
CW (01:04:35):
Seems difficult, because shirts are not flat, unless you are a door wearing it.
KSC (01:04:44):
Yeah.
CW (01:04:44):
The curve seems like it would cause some problems with the effect.
KSC (01:04:49):
Right. Yeah. No one has ever done it that way. There are a few ways that people have done it, but they usually do it on simple logos. Because I am an idiot, I just went full CMYK image of front and back side of the breadboard. Yeah, crazy.
(01:05:03):
But they work. It was cool. They were lenticular shirts. The next ones are going to be better. We learned what we needed to learn for the next run.
EW (01:05:14):
Cool. All right. I am going to close up the show because we have lots more to do. We did not talk about squid at all, so we may have to have you back at some point just to talk about squid.
KSC (01:05:32):
All right. Yes. I will come in as a squid enthusiast.
EW (01:05:37):
And you offered a coupon for buying Jumperlesses.
KSC (01:05:43):
Yeah.
CW (01:05:43):
Jumperlesses?
EW (01:05:44):
Is that the plural? Or is it Jumperi? It is an octopus joke. Come on!
KSC (01:05:49):
Yeah, I know. I really want that to be a controversial thing, because they are all right.
CW (01:05:53):
Could be Hungarian, Jumperlesk.
KSC (01:05:55):
It is actually Jumpoderdies.
EW (01:05:55):
<laugh> Exactly. Okay. So embeddedfm is the coupon code. It will get you 15% off for at least the next month. After that, Kevin gets the opportunity to turn us off.
KSC (01:06:13):
I always forget to turn them off. It is cool. It will go on forever. Yeah, it is all good.
EW (01:06:21):
Kevin, do you have any thoughts you would like to leave us with?
KSC (01:06:23):
Okay. Did you ask my favorite acronym? I have a bunch of squid thoughts I want to talk about.
CW (01:06:28):
No. No, no. Go for it.
KSC (01:06:31):
If I already said it earlier, just cut out the first one. But the favorite acronym of SQUID, Superconducting Quantum Interference Devices.
CW (01:06:39):
Ahh.
KSC (01:06:40):
That totally bridges the world of what I do.
CW (01:06:44):
Huh. I have to look that up.
KSC (01:06:45):
So that is my final thought. It is a really sensitive magnetometer. You get little rings of superconductor and they measure a magnet driving by on a truck a mile away, kind of stuff.
CW (01:07:02):
All right.
KSC (01:07:03):
That is a weird closing thought. We are just going to go back to the lightning round.
EW (01:07:08):
Cool.
KSC (01:07:08):
At least I got squid in there.
EW (01:07:13):
Our guest has been Kevin Santo Cappuccio, founder of Architeuthis Flux, creator of Jumperless, and doer of silly things.
CW (01:07:24):
Thanks, Kevin. It was really fun to talk to you.
KSC (01:07:25):
Thanks.
EW (01:07:27):
Thank you to Christopher for producing and co-hosting. Thank you to Peter for the introduction, and to Seth for his questions. Thank you to our Patreon Listener Slack group for their support and their amusement. Also, thank you for listening. You can always contact us at show@embedded.fm or hit the contact link on embedded.fm.
(01:07:50):
And now I do not have a quote to leave you with. I have facts. The giant squid, Architeuthis dux, is a species of deep ocean-dwelling squid. It can grow to a tremendous size, offering an example of abyssal gigantism. Recent estimates put the body size at around five meters, 16 feet, for females. Males are slightly shorter.
KSC (01:08:20):
Makes it longer than the colossal squid, which is at an estimate of only 4.2 meters, 14 feet, but it is substantially lighter. Much less robust arms. And realistically, the arms are the whole thing. The mantle of the giant squid is quite a bit smaller than the mantle of the colossal squid.
(01:08:44):
I have strong opinions on- The giant squid is way cooler. The colossal squid, they are like stubby and round.
CW (01:08:51):
I think it is misnamed. I expected something- Every time I see colossal squid in the science news and stuff, I am like, "Is that it? It does not look that big compared to a giant squid. Colossal should be bigger, right? It should be like a kaiju."
EW (01:09:08):
They are both kaiju.
KSC (01:09:10):
Then there are the Humboldt squid down in Mexico. They are huge, but not even close, but almost people call those "giant squids" sometimes too.
EW (01:09:20):
Well, the Humboldt squid, they come up closer to the surface and we see them more often, so they are quite a bit more terrifying. A 14-foot squid in the deep sea that I am never ever going to see, is not as scary as a seven-foot squid that I might accidentally swim with.
KSC (01:09:37):
Yeah. Because you always see giant squid when they are dead, washing up. Like, "Ah, you are not so tough now. Come meet me on land."