Parker is working on the front panel for the SSPS, Rigol DP832 woes, and Stephen does a guest lecture for the Iron Yard.
Podcast Transcript
Unknown Speaker
Hello, and welcome to the macro fab engineering podcast. I'm your host, Parker Dohmen.
Host 3
Unknown Speaker
So this week, I've been working more on the super simple power supply. Yeah, mainly on the front panel design. So basically, as we said before, is we want to put tons of LEDs on this thing. Hmm, because why not? Yeah, lots of lots of LEDs, different colors. Been looking at RGB segmented displays. I kind of want to do 16 segments, and I haven't found RGBs segments displays that are 16 segment yet. Might have to make a design decision there. We can make our own. Make your own segment. Yeah, why not? Well, I've been trying to keep this thing. Ah, the parts wise is stuff that you can just buy off the shelf. And so other people can replicate the project.
Host 3
How about how about LED boards in 3d printed? trouts?
Unknown Speaker
Yeah, that could work. We can Oh, yeah. Like that. He we can talk about this? Yeah, um, I guess? Yeah, that'd be good idea. Yeah. But um, some LEDs I was looking at where the king king breaks, WP 1040 threes, which are like bar graph LEDs. 100 individual. So you can make your own I guess style bar graphs. Okay. And I was going to put them in sweeps. And so when you turn the encoder, so the LEDs around the encoder, and so it would follow your encoder around. Ah, cool. Little cool little effect there. And then, and I was thinking, like, how many decimal places do we want to go to? On for accuracy? Yeah, accuracy, because I've had that many displays.
Host 3
Well, okay, so the, on the power supply we have right now, which is the Rygel, we bought the option where it has an accuracy of 10 millivolt. Or, I'm sorry, a resolution of 10 millivolt. And then a current of believe it was one milliamp. I can't I can't remember the current, but 10 millivolt on the voltage and that's probably more than enough for what we do.
Unknown Speaker
Alright, so I was thinking, I was gonna do to two decimal places on on both amperage and voltage. Enough display. Oh, yeah, that'd
Host 3
be perfect. Yeah, yeah. Cool.
Unknown Speaker
And then, so yeah, an encoder so that you could select what voltage? Well, you want set the voltage and current limiting to, and then a keypad for? You can just punch stuff in? Yeah. So let's because sometimes that's faster if you have to, you know, instead of rolling the encoder from one volt to 40, punch in 40, enter and
Host 3
Oh, come on. You don't want to program velocity into the
Unknown Speaker
oil have that but sometimes it's you know, the like, the wriggle has both. Yeah, yeah. You can just punch it in except the rural has a weird, like, the numbers go in a circle around the knob. Yeah. Which is really funky.
Host 3
It's funky. It seems like marketing crap.
Unknown Speaker
Yeah. It seems like a an engineer didn't make that design decision. Oh, no, no, no, because it's really annoying. The key stuff in on it. Ours is going to have a standard phone style. Yeah, absolutely. The keypad on it. And then we were talking about last week about making an internet of thing, an IoT device. Yeah. And so I was I've been doing more research on the ESP 82. Six, which is a very popular low cost Wi Fi module. Hmm. I'm doing some low end research. It looks like you just talk serial to it. You know, TX RX. Pretty simple stuff.
Host 3
So I actually internet of things real quick deviation, I met a guy who created a barbecue that had a feedback loop. So it kept itself in a constant temperature. And it was connected to the internet. And it would, why was it connected in it? Because it would tweet its temperature and what what he was cooking. At that time, he could somehow enter it in and say, Hey, I'm cooking burgers, and they're there at this temperature.
Unknown Speaker
And so your friends would show up right when it's ready. Exactly.
Host 3
That's cool. The Internet of Things is is so weird.
Unknown Speaker
Yeah. It's the Internet of Things is one of those. It's like, about a decade ago, it was like, Oh, my God, everything will be connected to the internet be awesome. And nowadays, it's like, I don't want anything connected to the internet ever. Like why why would you want your frigerator have a screen on it? To show you the news?
Host 3
Yeah, well, it's that kind of thing. Here's the thing. Here's the only way that I think that barbecue could be more IoT is if it had a screen on it, and you could scan all the ingredients of whatever you're putting on it. And then it would and then it would tweet that out. That would be you know, maximum IoT want to
Unknown Speaker
have One surprised it hasn't happened yet. Is a barbecue with a little screen on it that's connected up to ESPN. So when you're like watching sports on the big screen, you can run out and as you're checking the you know, the meat on the grill. Yeah, it has the game already on it.
Host 3
Yeah, yeah, yeah. And then a second one underneath the lid. So if you have to pull the lid up. You got another screen high temp screen. A high temp screen. Yeah, exactly. Never missed the game.
Unknown Speaker
Have you ever seen I'm gonna keep going on. I guess. The gas stations that actually have a the the pumps have a high resolution display in them?
Host 3
Oh, yeah. And they play ads while you're pumping? Yeah.
Unknown Speaker
We need more ads. And oh, yeah. Sometimes they do news and stuff like that. I've seen elevators, where that kind of stuff in the to where they display ads. It's like, it's kind of weird. That kind of marketing is how, I guess there must have some kind of metric on whether or not that's effective or not.
Host 3
Hmm. Probably, if you can make money on it. Someone's gonna do it.
Unknown Speaker
Exactly. Because at the gas station, one did pop up with like, an energy drink. And I'm like, I'm thirsty now.
Host 3
No, no, here's where it gets crazy, is if you pull up to the gas station, and it starts like, like a Mouser commercial comes up. I'm mad. I'm gonna get scary.
Unknown Speaker
Oh, so it's like, no no's. Yeah.
Host 3
Unknown Speaker
I guess I could be true if your car was also connected to a new your there. Like a RFID tag their car? Yeah. Good thing I drive an old car that doesn't have any computers in it.
Host 3
Okay, so that data harvesting we won't stay on this long. But, but I got a funny story. I before I got married on Facebook, it'd be like you want to meet local singles or think something like that. Now that I've got married, and I changed my status on there. It's like, Would you like to buy a Silla scopes?
Unknown Speaker
I'm like, do you want to reverse mortgage?
Host 3
Yeah, great. Thanks, Facebook. That's actually
Unknown Speaker
really funny. Anyway, so going back is been doing research on how to make IoT work. So I've actually never designed in embedded system to do that. Right? It looks like one of the easiest ways is instead of talking to like Twitter directly, or when these big API's actually make it, talk to it, a server basically, like a middleman server, and then that server pushes it up to wherever it needs to go. It's a lot that's it seems to be a lot more flexible. Sure. Not might change if I, you know, do more research about it. Yeah, yeah, probably. Yeah. And you actually just got back from the iron yard. What is that?
Host 3
Yeah. So so the iron ore yard is a it's actually a programming school. There's a handful of locations around around the states and one here local in Houston. And actually, a few weeks ago, I they actually sent me a message on Twitter. Yeah. I think you you
Unknown Speaker
basically said, Hey, these guys from the iron yard, or I can't remember what it was, sent me a message in. And I was like, yeah, just go do it. I had no idea what it was.
Host 3
Yeah. Well, basically, the I literally get this message from these guys that says, Hey, would you come to a guest lecturer at our place? I'm like, I don't even know who these guys are. So end up looking them up. And they do they do programming and they got a pretty cool facility. They focus mainly on JavaScript. And she, what's the other one starts with an our Ruby Rails? Oh, it's rails. Yeah, Rails is what should they focus on that those two?
Unknown Speaker
Ruby there is also Ruby on Rails? I wanted that is that,
Host 3
you know, I wouldn't be surprised. For those folks out there. I am not much of a programmer. So I've heard the word JavaScript, but I don't know a whole lot about it. And rails never heard of, but they Chris Chris
Unknown Speaker
was went there once for a lecture once right, you know, gave a lecture. Yeah, I
Host 3
believe he did. Yeah, Chris, the CEO of FAB went went there. But they actually had asked me to go there and do a lecture on hardware design. So I actually just before this podcast with the I was over there giving a lecture for them. And that turned out really well was a lot of fun.
Unknown Speaker
Yeah, um, how deep in detail, did you have to go about hardware? Or did you not go deep at all? Well,
Host 3
the lecture was 60 to 90 minutes, which honestly is pretty long for a general lecture.
Unknown Speaker
That's it. That's, I mean, that you basically gave a one class on on this stuff, right? Right.
Host 3
So that what I told everyone is, you know, how can you condense 30 years of information into an hour It's not possible. But But what I, what I really focused on. And what I thought was a good idea is instead of showing everyone, here's a battery, here's a light bulb, connect them together and something happens, it was more of the things that you don't really see on a regular basis. Because if you want to light a light bulb, go to Google and type it in, you'll, you'll figure it out in five minutes. Basically, I was showing them the tool sets that they need to know, what's an EDA tool? How do you what's the design production important things? Exactly? Well, and on top of that, the things that you don't really get told exactly, um, and I even showed actually the macro watch that you design Parker, I brought the macro watch. And I showed them the whole step from taking a schematic moving it into a PCB, taking the PCB, exporting the files, uploading them to macro fab and taking it all the way to ordering. So the whole shootin match. Yeah. And they really liked it. It was it was awesome.
Unknown Speaker
Cool. Well, hopefully, we get some of those guys, you know, building some
Host 3
hardware. It's it sounds like they were really interested in it. You know,
Unknown Speaker
did you suggest any kind of platforms or anything like that? I would just left that up to them. You know,
Host 3
we suggested Arduino and Raspberry Pi. They actually had a monitor there that was running some some video on Raspberry Pi. So some people were Yeah, they were a little bit familiar. But I can guarantee you now they're all they were asking a lot of questions and we're really interested in they're going to go research it.
Unknown Speaker
So there'd be like, what, 20 people buying Arduino starter kits tonight?
Host 3
And then spending a couple hours trying to make a light blink. Blinky,
Unknown Speaker
Blinky blinky.ai. No, yes. I think that's the right one. Yeah. Where did that come from? Dot inf O for Arduino? Ah, I don't know.
Host 3
But the wiring stuff. wiring. I bet you it came from wiring.
Unknown Speaker
Probably. We'll figure it out. We'll report back.
Host 3
Whenever we do, Arduino oh, gosh, the sound guy. Trump says he comes from the end of Arduino.
Unknown Speaker
Oh, we just we Yes. We pretend to be excellent. I
Host 3
quit. I'm done. Josh. You could take my spot here. Oh. Thank you. Thank you, Josh. Amazing. Shall we move on? Yeah.
Unknown Speaker
Um, and then I just finished writing up came out yesterday, or two days ago, since this podcast is gonna come out on Friday, the programming for PCB production article, which basically I just cover, you know, how do you program all your little tiny devices quickly and cheaply as possible, which is pretty important if you want your device to actually be usable in the field. So I went over like different methods. You know, Pogo pin, just started covering pogo pins. Yeah, dedicated programming connectors, that kind of stuff. We're going to be going over more in detail on each of these later. But yeah, it was a pretty, it was a pretty good overview article, I
Host 3
would say, super important for people to know. Yeah, very,
Unknown Speaker
very important information. It's actually one of the most asked questions we have here at macro fab about their products, because they, you know, most time they get to us, and they have a completed design. And they're like, how do I know this thing works? Yeah. And, and so, you know, you know, just like plugging, plugging in little tiny wires and or plugging into a breadboard for programming is fine if you're building one. But when you build 500 or 1000 of these things, yeah, that's game over. Yeah, some kind of method that all someone has to do is plug something in once and press one button. And it tells them whether or not it works or not. That's a pretty good in getting programming and testing down to that level is actually kind of difficult. If you have never done before.
Host 3
Oh, yeah. Yeah. The rule of thumb is just dummy proof it make sure that that it takes zero smarts to program it. Exactly.
Unknown Speaker
Host 3
yeah, that's pretty good.
Unknown Speaker
Um, we're gonna move on to our rapid fire opinions. And so last week, that RFO section wasn't really rapid fire at all. No. So we'll speed it up a little bit this week. Heathkit launched their new website this week. Okay. And so I've actually I knew they came back like a year ago. So I took a look at their website it looks like any other web 2.0 website, you know with like, you scroll and like the background slowly scrolls backwards. Yeah. All that kind of stuff. Focus on design. Yeah, it's really designing that whatever. But I was looking at what they actually build, they only have like two products and then plus, like, tons of parts for these products.
Host 3
Do they expect them to break?
Unknown Speaker
It? Maybe, maybe. But the big thing I looked at was, I was scrolling through and they have a am radio. Yeah, first I'm like, well, there's just am. So it's does am still exist. Yes, it exists. Okay. Yeah, I don't I don't have a I don't have XM or anything like that in my jeep. So okay, I just have FM and AM and the FM works most of the time. But this receiver kit is like 150 fucking dollars.
Host 3
Oh, gosh. 150 bucks. In 2016. I mean, sure. No, no, no, no, not only an am radio, a chore. It's an am radio that you
Unknown Speaker
have to build. You have to build the entire thing now. Sure. It looks really nice. It looks the part it looks like it's from the 50s Yeah, it's got wood, wood grain paneling out. Actually, I will say it's from the 70s. Okay, it's got wood grain. It's got brushed stainless. It's got really nice silkscreen for the logo. No logo looks awesome on it's like, from the 50s. Like, you know,
Host 3
did they keep the original Heathkit logo? When did they update?
Unknown Speaker
Oh, it definitely still looks like from the 50s where it's that like, nuclear rocket
Host 3
age kind of look. Yeah, the Jetsons kind of Yeah.
Unknown Speaker
But $150 for a kit that this does. Am radio. That's rough. I guess most of it's in the box enclosure?
Host 3
Pro you probably you know, he'd get used to have some really cool stuff. He could had like a build your own TV kit and to build your own TV back then when you were a kid. That's insane. Now and build your own oscilloscope. And actually, not that long ago, a guy walked into my repair shop. And he brought in two Heathkit guitar amps. And they sounded horrible. But he was still like, Hey, I like these things. Can you fix them up?
Unknown Speaker
What they were they too? Or? No, no, they were all solid state. What amps were they run? And then are you remember? Oh, sorry. Was it what amps were in there?
Host 3
No, I I can't even remember. I remember I had to replace a handful of transistors because they were just done when they were garbage.
Unknown Speaker
Actually imagined they put it in like HK 386.
Host 3
Yeah, exactly. No, no, they were just they were terrible. But he liked him. So it's all
Unknown Speaker
good. Yeah, that's all that matters. Yeah. Cool. So yeah, I guess, uh, I have no opinion about bow belt that both of us are kind of before the generation of Heath kids.
Host 3
So yeah, I don't I don't, I can't see them going very far with it with the fact that an Arduino which cost, you know, so much less and can do, you know, a million times more, is just gonna help beat it every time. And
Unknown Speaker
you say that too. And then you have the Raspberry Pi Zero, which costs theoretical $5 As you can find it? Yeah, right. Right. And that runs, you know, Python and all this other stuff. And also, more or less interfaces with the physical world pin wise.
Host 3
But come on, man, it doesn't come in a wooden box.
Unknown Speaker
So we need build a wooden box for the zero and 50 stainless strip on it. Yeah, and silkscreen. Some nice stuff and sell it for 150 bucks,
Host 3
we can call it the PI infinity, the PI infinity. You know, I
Unknown Speaker
betcha Raspberry Pi eventually come out with one of those like
Host 3
Unknown Speaker
And then, earlier this week, the little box challenge the winner was announced of that. And so what the little box challenge was, was a challenge that by Google to build a Pacey, a solar power, solar powered power converter or invert inverter. Yeah, so it turned the DC voltage from a solar panel array into AC voltage that you can plug your blender into. Right.
Host 3
Which is which is one of the weakest links in solar systems? Yeah,
Unknown Speaker
the weakest links in terms of Well, I think it's probably the second most inefficient part, right? The most inefficient part is the solar panel itself, right? And then the inverter would be number two. So the goal was to basically build a two kilowatt power inverter which those exists But the big thing was to get the power density about a factor greater than what already existed. Most power inverters are about point five to five watt per cubic inch conversion and they wanted 50
Host 3
Unknown Speaker
Yeah, that's, that's pretty big. And all this would have to go into a 40 inch cubic box. So it's it's pretty small. And other specs was like it needed it needed to convert 450 volt DC with the inline resistors so that was the simulation of a solar panel.
Host 3
And and you need to bring that down to 240 V AC.
Unknown Speaker
Yeah, they did 240 volts. Okay. I think 60 hertz, which makes sense. Yeah, that that's, that's what 240 volt AC, you can pretty much power at 60 hertz, you can pretty much power anything in the world on that. With the right converter on the back end. Yeah. So what
Host 3
were the results of the winning team?
Unknown Speaker
They did a let me see. Look at my sheet of paper. The winners were the CT power read electrical devils. That's a nice name. Yeah, really cool name. But they were able to do this all this stuff in 13.77 inches cubed, which was tiny. They held a converter and it was like, just in the palm of their hands.
Host 3
That's that's less than half the requirement. Yeah, less than half and they
Unknown Speaker
actually went at 143 watts per inch cubed. So they, what, that's 143 watts. Yes. So that was four times, almost four times. Acquired density
Host 3
three times almost three times because it was 50. Was the required? Oh, yeah, you're right. Three. Yeah. But still, that's that's ridiculous.
Unknown Speaker
Yeah. And the trick was that use these are the gallium nitride Yeah, gi n transistors. And so we gave a little look on on Mouser. Where can we buy these mythical transistors? At the end? So we searched Mouser for basically g a n transistors. Yeah. And they exist. Yeah. And actually, the company is called gallium nitride solutions.
Host 3
Unknown Speaker
Host 3
per in single and single in singles.
Unknown Speaker
But they have really, really awesome specs. So basically, like, they're highly optimized three power transistors basically. Yeah, they have huge drain source voltage breakdowns. They can do 60 amps. Continuous really low on voltage. They have a okay ish, you know, gate the source. So it's only about 10 volts to turn them fully on.
Host 3
We'll know that. I think that's that's the maximum because their turn on is the great gate to source threshold. Where's
Unknown Speaker
Host 3
Oh, there it is. See? That's 1.6 volts.
Unknown Speaker
Alright, so you can actually power this off an Arduino. or actually any microcontroller? Pretty much you can turn this thing on with Okada without level shifter on our cheat
Host 3
sheet here. There's the next the next statistic or, or thing here. characteristic parameter, whatever you want to call it. The gate charge. Yeah, this is ridiculous. What is he? nanofilms 13 Nano coolrooms. It takes nothing to turn these on. Yeah, no drive whatsoever.
Unknown Speaker
And then maximum operating temperatures. 150. C, which is also insane. Yeah.
Host 3
So they're just giants. They're just just beefy monster transistors.
Unknown Speaker
Yeah, but $61. I mean, Google didn't set a max price for like a volume run of these. Yeah. I want to bet you this box that they built probably had about a grand worth of parts.
Host 3
Well, and it just like any other technology, if they start manufacturing these they will drop. Yeah, I'm sure these these gallium nitride transistors are difficult to manufacture.
Unknown Speaker
Probably I bet they have, because they have a couple different versions on Mouser. But they all pretty much look the same. I betcha they're just binned is a build one. And then the best ones in the wafer. Are the higher tier ones.
Host 3
Yeah, that would make sense. And then they give all the rest of them to you and me, right?
Unknown Speaker
Yeah, the cheapest one was like 50 bucks. And it was not a lot different.
Host 3
Do they come in both flavors? Is it a NPN PNP kind of thing? I only saw NPN. Okay, maybe PMPs are just difficult to make or even not even when possible.
Unknown Speaker
PMPs are typically less efficient overall. Yeah, for given the same. Same structure. Yeah. or maybe Mouser didn't carry him. I didn't do any more research beyond looking at the datasheet that Mouser has.
Host 3
Yeah, and that means there's like, we only want to use positive voltage from here on out. So, where we're obsoleting all of our PNPs
Unknown Speaker
these are some crazy actually, one thing I wanted to look up I didn't was I want to know what the physical package is like on this thing. Oh, yeah, yeah. That I'd like to look that up.
Host 3
Yeah, we should. We should check that out and put that on the blog. Yeah. Cool.
Unknown Speaker
Well, I think that's gonna gonna do it for this week's podcast. I'm Chris Parker.
Host 3
And I'm Steven. Catch you later. Take it easy.
