- New Engineering Article
- Raspberry Pi Compute Module Motherboard
- Continuation from the Compute Module board from last year
- Added the PCM5122 audio DAC
- Learning about USB HUB design
- TPS2044 for overcurrent protection
- Only rated for 500mA
- USB spec says only 120uF cap per hub required but 100uF per port is recommended with low ESR
- USB DM/DP signals require around 227-ohmtermination resistors
- Optional filter of 22pF recommended for EMI suppression
- This creates a “low pass” filter with a cut off of 267.9 MHz
- 12M bit/sec full Speed USB 12MHz, well under this cut off
- Rapid Fire Opinion (R.F.O.)
- MacroFab is Hiring!
- Apply here
Visit our Slack Channel and join the conversation in between episodes!
If you have a cool idea, project or topic that you want Stephen and me to discuss, tweet at us @MacroFab or email us.
If you are not subscribed to the podcast yet, click that subscribe button. That way you get the latest MEP episode right when it releases. And please review us, wherever you listen (PodcastAddict, iTunes), it helps this show stay visible and helps new listeners find us. (:
Special thanks to whixr over at Tymkrs for the intro and outro!
About The Hosts
Parker Dillmann is MacroFab's Co-Founder, and Lead ECE with backgrounds in Embedded System Design, and Digital Signal Processing. He got his start in 2005 by hacking Nintendo consoles into portable gaming units. He also runs the blog, longhornengineer.com, where he posts his personal projects, technical guides, and appnotes about board layout design and components. Parker graduated with a BS in Electrical and Computer Engineering from the University of Texas.
Stephen Kraig began his electronics career by building musical oriented circuits in 2003. Stephen is an avid guitar player and, in his down time, manufactures audio electronics including guitar amplifiers, pedals, and pro audio gear. Stephen graduated with a BS in Electrical Engineering from Texas A&M University.
Host 3 00:10
Hello, and welcome to the macro fab engineering podcast. We are your host, Stephen Craig and Parker, Dolman. And this is episode 105. Yeah. So hey, this week, I found something pretty cool. So there's a character on YouTube, actually. So characters this real quick, this is a little bit more of like an announcement for someone else's thing. But I thought it was super cool that I just wanted to share it. Because it's something that potentially maybe we or maybe one of our listeners might be interested in actually doing. So there's a YouTube channel that's called Stand Up maths. Okay. So it's run by a guy named Matt Parker. And not me. No, he's an entertaining mathematician. And he definitely not me. He goes around the nation and just does like, well, not the nation goes actually around the world and give talks about mathematics. And most of the time, it's like, really just, it's goofy. But it's fun. And it makes math just kind of interesting. And it's really great background listening stuff when you're working on a project. So I've been watching this guy on YouTube for a while. And this week, earlier, I saw a video that he recently released about a three sided dice. And I'm totally going to steal this from his throne, as you tell you but not
Host 1 01:22
Like a normal d3, which has she has what four sides?
Host 3 01:27
Will a three sided dice. The way I've seen it before is like, take a triangle. Okay, and extrude it? Oh, okay. I've
Host 1 01:35
Side by sides. But it can only roll on three. You roll it like a? Yeah, okay, I see. So get
Host 3 01:40
This, and I'm stealing this directly from his YouTube. So if you go listen to it, it's going to be a repeat. So think of a quarter, right? Quarter has two sides. But it really kind of has three because it actually has a thickness to it.
Host 1 01:52
Yeah, it has the circumference plus depth. Right, right. Or times depth.
Host 3 01:56
So if you flip a quarter, you know, there's an unbelievably high probability that it's going to land on one of the faces, but it could land on an edge.
Host 1 02:04
Yeah, it's like it's like, it's like 4949. And then two? Well, actually, probably less than that. But way less,
Host 3 02:11
But yeah, like that. So take the other extreme, if you were to take a pen and flip a pen, it's never going to get heads or tails. It's always going to land on the side, right? Yep. So there's a cylinder that is somewhere between a quarter and a pen, where it's equally district distributed. That if you flip it, it could land on either of the faces or the side. And there's a chance for that
Host 1 02:34
Would not be just it's you said cylinder, right. Yeah, yeah, you have, right. Yeah,
Host 3 02:39
It's like a quarter that's extruded? Yep. So the cool thing
Host 1 02:44
Is, it's like the size of a suit pan.
Host 3 02:46
You know, the thing is, they're trying to figure it out. Oh, okay. nobody actually knows what the shape is. Yeah, we know what the shape is. But we don't know what the ratio of the thickness to the diameter is. And the way they're figuring it out is actually really fun. Cuz they're allowing the community to help out with this. And I wanted to bring this up, because it sort of relates to a little bit of what we've talked about in the past with a handful of topics. So they actually this guy, Matt Parker, got together with an engineer, and they tried to like, figure out some nice ratios that such that it works. And they actually 3d printed, or they laser cut some dice. And they tried it, they rolled them like 1000 times got all the distribution, got a statistician to come and look at the probabilities. And they were way off. Like their their dice is not a perfect three sided dice. So what they decided to do was create 3d printable dye that you can get off of Thingiverse,
Host 1 03:43
He wants to use my 3d printer.
Host 3 03:46
A lot of people to do this, they created these kits of dice that anyone can go, and 3d print them at different ratios. You can go and roll them and get a whole data set. And they have this website where anyone can upload their own data set for this dice. And they're going to take all of this data and they're going to do a massive community crunch, to find empirically what is an A real three sided dice? Because it's actually super hard to calculate. Yep. So instead of calculating it, they're just gonna have a bunch of people do. Yeah.
Host 1 04:17
So this got me thinking on like, maybe doing a AI learning program that has a physics model. Right, and hasn't Yeah, drop it 50 million times. Yeah. And try to find a cylinder that equally distributes that. And that way, you know, I wonder if you think they would have tried that, you know, or they're just trying a different way of doing it.
Host 3 04:41
Yeah, I mean, it sounds like they're just trying to, you know, they're probably trying to get YouTube views and things. Yeah, but I thought it was super cool to engage the community. And it's, this is one of those, like, things that seems like when I first heard about I was like, that sounds like something a mathematician would have done like 60 years ago. This sounds like a problem that, you know, these kind of simple and I put that in quotes problems. You know, a lot of them have been solved, but I guess a lot of them haven't. This is not something that really matters to day to day life. But at the same time, it's super cool that you can be a part of it. So we'll put up links to the websites where you can, well, the Thingiverse will put up a link to Matt's YouTube video. And the place where you can upload your data, I guess I'm going to print some dice will and at the same time for 20 bucks, all of them all the dice sets. 10 dice sets are available on Shapeways. Ah, so for 20 bucks, you can get them in various radiuses. Or you can get one set that's 10 dice and a bunch of different radiuses. Gotcha. And what and the radius are all ratios to their thickness. Yep. And sometimes, like they have a whole bunch of like goofy ones that are like, the ratio was he or the ratio was pi, or, you know, a lot of like real mathy kind of those things. I don't know, I just thought that was really cool. Not that it was worth sharing. I would love to see like a bunch of people give it a shot.
Host 1 06:06
Yeah, I'm I'm definitely that shot. That sounds cool. Yeah.
Host 3 06:09
And for you know, if you don't get a 3d printer, Shapeways got, you know, 20 bucks is not bad. partake in this. Yeah. And there is a video showing the official rules for this. And really, there's only one rule and it's how to roll the dice. Because they want everyone to have you put it in a cup. You put you put it in a cup, there's a certain like, number of shakes, and then there's a drop, and then you count, you know, heads tail sides. So yeah, they have like a whole system. And I don't know, I think I think that would just be really cool to be like, Yeah, I helped determine what the golden three die three sided die ratio is.
Host 1 06:42
You know, it got me thinking because we played Star Wars D 20. Last Friday. Yep. We did some streaming on that. But the I guess I can put a link to the stream as well. Oh, watch out. We got really drunk. You got really? Well, Ross was pretty drunk to one of our gaming friends. He got really drunk. But the, in the in this software we're using which sculpt? Roll 20? Yep. It has simulated 3d dice that you can roll. And I'm like, I doubt that's a physics engine. Oh, no. But it does make it look really like. It's really convincing how it rolls and bounces around. So maybe it is simulated. And it doesn't give you the result until after the role stops? To maybe it actually is a physics engine.
Host 3 07:39
Well, what about using Garry's Mod from the Source engine? Yeah, you can create basically anything you want in that and drop it from the sky. Yeah. So you could always use HalfLife physics engine? Who knows how accurate they are rolling? Yeah, I doubt it is.
Host 1 07:58
I'll take a look at the roll 20 API, I wonder if you can create your own dice sizes. I kind of doubt it. But be interesting to look into.
Host 3 08:06
That'd be that would be hilarious. If d&d solved a a mathematicians problem?
Host 1 08:12
Yeah. Because then you can then you can just simulate, I'm just saying, think of what's the fastest way to like, build up a physics engine and simulate the stuff like so you can do different ratios and stuff.
Host 3 08:24
Right. And and sort of the way they discussed it, because they gave some rules about 3d printing, because you could basically 3d print these days with a really thin infill, and that would skew the numbers. So if you were to, like accurately moderate model this, you would want a cylinder of uniform density. Yep. And complete, you know, no, 100 100%. And I would think you could completely ignore air resistance. So basically, just make a vacuum playground for this thing. And just calculate like, you know, moments and it's spinning and all the inertia and stuff like that. I'm sure it's not trivial. But I think it could be
Host 1 09:08
Done. You know, what we should do is we should tweet at that roll 20 company and see if they will go in on this.
Host 3 09:16
And bid on Saturday. Yeah. And put it in for three sided I think they just basically you roll a six sided and divide by two.
Host 1 09:24
No, no, I actually have three sided dice. A d3 is a four sided pyramid. So it's a well equal lateral pyramid. I think it's what they call it. And then each point, or is that default?
Host 1 09:39
That's a deeper
Host 1 09:40
Yeah, that's a deep I don't think there's a d3 then.
Host 3 09:42
Yeah, the way we used to do d3 is back when I was in high school playing d&d, we should just go out and buy erasers. Because a lot of times you can get a triangular eraser. Just write a number on each side. There you go, and that you already solved the problem. Yeah, there it is. There's a die three, although those have some You know, nonzero chance. Yeah. Landing on one of the other two. As long as that's a critical hit. Yes. Yeah. As long as that that not least the one that happens a lot isn't like one. Yeah, we you see, this is the exact this is exactly how the difference between an engineers and mathematicians mind work because like, to us, it's like, oh, yeah, well, that's there's, there's your solution. And to a mathematician, they're like, but there is a chance it could land on the side. And we're like, Yeah, whatever, I'll take my chance. So if you want to partake in that, the if you go to Mac fab.com, and go navigate to the podcast, and Episode 105, we'll have all the links up there when this episode releases. Yeah. Cool. Yep. So kind of on a little bit of a side note, completely different topic. I'm gonna segue here. I was actually kind of looking at some fun stuff. Last night, I was just seeing like, for an average home game, or just a guy who's got a little bench at home, like, what's the highest frequency signal that you can generate? Like, can you go make a
Host 1 11:06
I think you need to specify that too? Well, it's like, how, what do you say? Like, I mean, like, when you say a signal? Is it like, as a square wave? Is it a perfect sine wave? Because you do have, like, when you start getting up in speed, you start getting slew and overshoot and a bunch of other crazy things. So
Host 3 11:27
I, I didn't put the stipulation of the signal has to look a certain way. I just wanted to lately, I just wanted to go fast. Okay, gotta go fast. Yes.
Host 1 11:35
So you're looking for a sonic Yeah,
Host 3 11:37
Right. And so you know, just looking at chips available online, it's actually not that hard to get in, like the, you know, the 2.4 gigahertz range, like, there's a lot of chips that just do it, you know, you need a little bit of an amplifier. But most of the time, they have like, like a circuit description, where it'll show like, if you put this down, and you put this amp after it, and you use the right coax, then you'll get, you know, 2.4 gigahertz. But I was like, gotta go fast. I want to go faster than that. And I actually found that you can just straight up buy terahertz signal generators. Hmm, like, way fast kind
Host 1 12:14
Of stuff. Yeah, I wonder what you would use those for.
Host 3 12:17
That's exactly what I was looking at. Because I was like, What is the point of having something that goes that fast, and they're not cheap, obviously. And they look all like super, like, lab grade stuff, because that's probably the only place they ever exist. But like, what would you use a terahertz? Probably doing terahertz squarewave
Host 1 12:35
Probably doing timing. Like, you can probably pump that through a laser diode, do some kind of timing, with whatever you're doing kind of a strobe light yet, but
Host 3 12:44
What can count that
Host 1 12:47
Something that counts faster than terahertz.
Host 3 12:50
Exactly what that is, or, or I guess, you'd have to do some kind of like, I don't know, phase something, or there's something that counts a lot slower, but it catches every, like 10th pulse, or something of that sort, you know, some way to divide it down and you actually capture it.
Host 1 13:07
Probably some kind of FPGA stuff could do it.
Host 3 13:10
On Terror. Yeah. Wow. I don't know. So.
Host 1 13:14
So in the high end, FPGA stuff is pretty crazy fast. I mean, that's how they were capturing like, neutrino timings, which like go like 90 plus percent the speed of light. So
Host 3 13:25
Yeah, yeah. Something like that. Or like, you know, the I was watching a documentary earlier today on it. The the most sensitive equipment in the world is the gosh, we got three places. There's Louisiana, there's France. And there's a place in Washington where they detect gravitational waves. Oh, yeah, they have to detect the laser like shifting on to 90 degree axes. And they have to detect the wavelength shift by like, 10 to the negative like, 6 billion. I don't remember how many zeros but it's a it's a gazillion zeros or something. And they're able to do that. So somehow, there's probably some kind of terahertz generator in there just be yes, yeah. Yeah. Because it's like, I don't know. I thought that was fun. Yeah,
Host 1 14:09
If you know what you would use a terahertz generator for. Treat us.
Host 3 14:14
Yeah, please let us know or I will not sleep until you
Host 1 14:22
So why are you saying, I have to know I got a no,
Host 3 14:25
No. So that was my funky section
Host 1 14:28
Bark now. Alright, so I last week, we're talking about the article I was already which was the article for design, how to use like the Mac web interface design considerations. So that article will be live by the time this podcast comes out. So give that a read. It's pretty in depth. It's a lot about using how to like use pricing to drive your design so you can get your price point down easier. So instead of having to like go back and redesign stuff, because Oh, that's too expensive. It's like what's, where should you like take a stop in your design to go check pricing out? Go check out that article. And now with the pin hack, rev eight kind of slowing down, like on my end, the hardware is kind of done. So it's up to the software devs up in Benton, Wisconsin. Now to finish that out. I started working on that compute module, motherboard, Raspberry Pi compute module motherboard. So we built that LVTs board, which I won the bet from Steven on like episode like 80 or 70, or something like that.
Host 3 15:39
That's the one that I gave you a six pack Azima for right. No, our guests gave me his email. Well, they gave it to me to give to you. Okay, counts. Okay.
Host 1 15:50
I think those I was like last July. Yeah. Or something like that.
Host 3 15:54
That yeah, that while we were still we were still recording with Josh, because we took a picture of Zema underneath et Yep.
Host 1 16:01
And that was before the hurricane. So yeah. Man, only like, three months later, that hurricane will mess everything up. You know, it's crazy. So I started working on that again. So the lbds board was like, just enough to test the pin EC rev eight stuff. But I wanted to flush that out more like I want to add the HDMI connector. I want to put in a USB hub, basically make it a full fledge motherboard, or like a Raspberry Pi three compute module, laptop, motherboard. That's where I'm going with it. So I started doing a lot of USB hub. Like research, like how do you design a USB hub, all that stuff? The proper way to do it, because that's the only way to do it. So I started looking at the I basically went to mouser.com. And when typed in USB hub, and pick the first result. Yeah, always best way to go. Oh, like for like a chip that does it? Yeah. Well, you need like, it needs to be able to multiplex the signals. So usually use a dedicated chip to do that, that does like 90% of the functions. So I looked at the T USB 24 Six B, which is like a it's a TI part. It handle it's a four port, Chip, it handles all enumeration, all that good stuff. And also handles overcurrent limonene and all that good stuff. So the US overcurrent limiting us a like a you need to use like a power multiplexer that basically is monitoring all the lines that come out of it. And so I picked the TPS 2044 which is another ti part because in like it's in the application note.
Host 3 17:46
Does it talk back to the hub chip? Yeah, it
Host 1 17:48
Talks back by just pulling pins high and low. Like, oh, this channel is overcurrent. So pull pin high. Oh, okay. Yeah.
Host 3 17:56
So it's a supervisory power management IC.
Host 1 18:00
Yeah. So you can use I've actually used a chip that the smaller version that was like only two channels of this power advisory chip before doing power management on I can't remember what that project was. For some reason I had that part designed. That's what I will use them for. But it was designed in my library so
Host 3 18:22
It can it can supply 500 milliamps to each
Host 1 18:25
Channel channel. Yeah. Nice. Which is okay, I want to be able to support power. Was it bd 2.1? What's that? That's the USB 2.0 spec for battery charging. Oh, yeah. So that's like, you know, 2.1 amps over five volts, right? So it's like 10 Watts,
Host 3 18:48
Right? But but doesn't, whatever, whatever you connect to, it doesn't have to ask for that.
Host 1 18:51
It has to ask for that. But this chip won't be able to do that. So I was like, Okay, I kind of got stopped there with that design, cuz I'm like, Okay, it's kind of you know, not, it not doesn't do the power I want because I want to be able to handle like the high powered stuff. And, and then I looked into in the TTY USB 2046 hub champion, which is only like 12 megabyte of megabits per second. So it's kind of slow. Because I think the Raspberry Pi three it handles high speed, which is like 480 megabits per second or something like that fast. Yeah. And then I started looking into the like, what requirements you have to put on to like two USB lines. So the USB spec says 120 microfarads. I think it's yeah, 120 microfarads cap per hub. So you need like a capacitor basically in their panel surge. You mean like right on the five volt line? Yeah, right on the five volt line the handle surge when you plug a device in? Yeah. CIR recommends actually using 100 uef microfarads per port, though, that's pretty beefy. That's pretty beefy. And they also recommend using low ESR. And they say tantalum, you can probably get away with a ginormous ceramic cap to
Host 3 20:09
Well, I mean, but but if you're talking about five volts, you can get a small five volt ceramic. Or well, I mean, the lowest you go is like, what? 10 To 16 volt? Something like that. Yeah, they're not that expensive.
Host 1 20:22
No, no. So you can you can probably get away with a ceramic. It's usually pretty low ESR. And so you can probably get away with 100. us like it probably would be what? 1210? or bigger?
Host 3 20:33
Yeah, you're probably not going to get lower than Oh, 805 Yeah, lucky. Oh, no,
Host 1 20:37
No, you won't be able get away to 500. That's a lot of gas. It's in the small cap, you would get one of those in like those crappy dielectrics. But yeah, yeah. So yeah, you need so good design practice 100 uef per pin or per port. And then the differential signaling, usually put about 27 ohms. of termination. It's depends on your design, but usually 27 ohms. is in the ballpark. That's 27 to ground. No, no answers. Oh, terminate my bed my bed. Yeah, yeah. And but to ground on the chip side of the port. And the chip side of these resistors is usually put a 22, Pico farad cap to help suppress. They say, EMF or EMI stuff. And so I'm like, well, that basically you created a low pass filter there,
Host 3 21:32
We so that the 22 peak of hair goes right at the port or denote on the chip at the
Host 1 21:39
Chip. So at the pin of the chip to ground is what you where you want this, this.
Host 3 21:44
So dumps through the cap instead of blasting out and yeah, space,
Host 1 21:48
Or actually it's from coming in, or it's probably EMI coming in through the cable. Oh, conductive. Conductive EMI, which is interesting is this is supposed to be most of the times you put EMI stuff close to the port. But this one's kind of weird, because that way, but regardless, or maybe it's just on has to be on the other side of that 27 ohm termination resistor will if you put it if you put it on the chip side, it is on the chip side. So it goes the pin on the chip. Yeah. Then this 22 pico for cap, then there is in the in series 27 Ohm resistor, and then the port. So I bet you what it is, is that the resistor needs to be as closest to the port. Yep. And this cap needs to be as close to the port. Because you limit how much right energies coming on to that line. I did look up because this, this creates a I put in quotes low pass filter because it's a low pass setup. But it's like 267 megahertz is this low pass filter.
Host 3 22:50
The thing is it creates okay. So it creates a low pass if you're looking into the port. Yeah. But you also have to take into consideration what the impedance of the pin of the chip is. Yep. So that's going to shift that. Yeah, frequency all over the place. But you're also seeing a two pole low pass filter. If you're if pretend you're the chip looking out. Yep, you're you have some output impedance. Yep. Then you see the cap, then you see the resistor, then you see anything else attached to that? So, you know, I guess those two components are probably they've probably been found empirically. I doubt that they're calculated.
Host 1 23:26
Yeah, no. But yeah, so that's like it. So it does create this like I Yeah, low pass filter, which basically probably cuts off any kind of EMI noise and stuff. takes the edge off. Yeah, it takes the edge off a little bit. You could
Host 3 23:39
Also if you don't have that resistor right at the port, I bet you could put a ferrite bead. Yeah, you can port you can and have that thing set way beyond the transmission speed.
Host 1 23:51
Yeah. Which is at for this chip, that's 12 megabits a second, I've calculated this up, which is 12 megahertz, right? So way below that,
Host 3 23:57
You could probably put like a 50 megahertz bead on the end, and you'd be fine. Yeah. Yeah.
Host 1 24:03
But given the fact that it's only full speed this chip, I'm probably have to go with the TTY USB 4041, which is full, which is a high speed, USB 2.0 or 2.1 device.
Host 3 24:15
Is it just the same thing, but high speed?
Host 1 24:17
No ish. It's a bigger chip, but a lot of the pins are unused. I had to do a lot more research on that. Okay. So hopefully next week, I'll have a I'm going to design a board that is a USB hub. That's gonna be the cheapest way to go to prototype this and I will probably build the world's most expensive USB hub.
Host 3 24:38
Are you willing to share that? Yeah, because I would love to see it myself. I'm not much of a you know, I don't know a whole lot about USB. So I'd love to see how you accomplish.
Host 1 24:48
Yes. So what I really want to add to this stuff is it needs to have a beefy five volt line, because I want to handle the BD 2.1 spec so Okay, and I think this T USB 4041 Hand supports that, too. It supports that that.
Host 3 25:08
Does it need any heat sinking? Because I mean if it has a big thermal pad, okay, yeah, cuz if you're, if you've got four point ports and they're all dropping 2.1 amps, yeah, I mean, that thing's gonna get hot. Well, no,
Host 1 25:18
The power doesn't go through the hub. Oh, it has a you have an auxiliary power distribution stuff.
Host 3 25:25
Will okay, but it's going through that chip, right? Or does it have offboard MOSFET or something?
Host 1 25:29
I think those the power distribution hub like the TSB, 2044 has built in MOSFET. And they're like, like, 20? Mila, home. Okay, so they don't get real don't get really hot. They're designed to handle the power all of it. Yeah. Yeah.
Host 3 25:46
All right. That's cool. Yeah, hopefully next week,
Host 1 25:49
I'll have that design. And we can talk about that design. And I'll probably talking about the next phase, which is gonna be some other section of this motherboard. Basically, I was gonna do like, every week, like alternate. So we'll talk about the design and then go over the design of the hub. And then we get to that we'll go off the next section, which might be you know, the lithium battery charging management's. You know,
Host 3 26:13
I think, eventually, all of your pinball stuff, and all of your technology is going to end up on one, like massive board that incorporates like, every possible form of technology, like ever, maybe. And it's one of those things where if you ever, if you ever switch jobs or go for a resume, all you need to do is just bring in that board. I built this. Yeah. This has everything out. Just point. Yeah.
Host 1 26:40
Yeah, definitely on this resume? Well, it's like, how do you make a complete system? So it's all self sufficient? Is what this raspberry pi cm board is going to be like? You just connect
Host 3 26:51
The dots. And it works. Yep.
Host 1 26:56
Because I already have the lbds stuff down. So display is done. Yeah. HDMI is just hooking up to HDMI. So it's not it's not anything crazy. You just got to make sure the differential pairs are the same. And yeah,
Host 3 27:10
Does a does Eagle have differential pair? Handling? Yep. Nice. Cool.
Host 1 27:15
Yeah. So that's gonna be a fun project. I can't wait to get that one done. It's gonna take a long time, though. So cuz especially as I'm building out like this, like in modules, like I'll do the hub, and then just copy the hub over. And then I'll do like the lithium battery production. Copy that over. Like, but you gotta admit, that's the best way to do it. Oh, yeah. For prototyping, especially, especially as I don't have like a pressing timeline. It's definitely the cheapest way and easiest way to go about it.
Host 3 27:47
Well, and each module you since they can effectively run independently, you can test it, make sure it's perfect. And then it becomes a plug in module. Yep. Yeah, that's so much easier.
Host 1 27:59
So the RFO. So we got to this week, well, kind of two and a half. All right. The first one's nice and quick. Is Mouser is now an authorized distributor of expressive? I think I pronounce that right. Expressive. Expressive is the manufacturer of all those really cool, cheap Wi Fi chips like the ESP 80 266. Cool, basically, what made the IoT stuff kind of explode in the last couple of years?
Host 3 28:30
So yeah, now you can go on Mouser and just buy the stuff that you used to have to, like, outsource to like AliExpress, like the full on modules. Yep. And the I think they're going to carry all the ICS too. So if you want to, like super embed their stuff you can. That is great. Yeah. So I can't wait to see that show up on. I'm very curious to see what the price is going to be on that. Yeah, that they're just unbelievably cheap on eBay and stuff. Yeah. But it's eBay, electronics on eBay as always iffy. Yeah, very iffy. But and so I wonder, I wonder how well mousers I mean, it's Mouser. They're usually pretty good about doing their homework. So they pick they pick good, valuable vendors?
Host 1 29:13
Well, they're going directly to the manufacturer, and that manufacturer expressive will be providing the parts to them. So as long as of course, that manufacturers supply chain hasn't been compromised. Yeah. Should be good.
Host 3 29:27
Hmm, that's cool. I'm really I'm, I'm curious to look at the data sheets. And yeah, dig through all that. Yeah, we're gonna have to
Host 1 29:33
Wait to see the stuff that drop and then see what the price differences are. From like buying from Aliexpress will
Host 3 29:43
Just sort of as like a default, it's probably going to be more expensive, but it's trusted. You get it faster.
Host 1 29:50
And I wonder what the price differences will be between Mouser in like all the small guys like Adafruit and SparkFun that we're carrying these parts as kind of like they would buy a real on Ollie Express and then test some of them and make sure they were okay. And wonder what their price difference is going to be. You know,
Host 3 30:08
And and it's interesting because Mazur carries a huge portion of Spark funds. Catalog. Yep. That kind of I wonder how that's working? How that will work for SparkFun? Because does that mean like SparkFun is now undercut on all their ESP stuff? You know, I guess if you want to read PCB from SparkFun, you could still pay more for that. Yeah. Just get the ESP, the Mouser, ESP or whatever it is. That's cool. Wait, would you hear about that? That was
Host 1 30:35
On electronics weekly. Awesome. Yeah. So look forward to that. And that's the best thing about that. It's going to be when you search, you can search for the ESP parts in the McWrap. Bomb, and you can select them now. Oh, because we get information from Mouser.
Host 3 30:53
Oh, that's super cool. So that's like a just a flat out to finable party getting Yeah,
Host 1 30:57
Instead of having be cept that well, we use that the consignment because you couldn't buy it from a authorized distributor.
Host 3 31:03
Right? Because going to Amazon or eBay or Ali, whatever. Yeah. It's kind of like a good sketch. Yeah, real sketch.
Host 1 31:12
Cool. And then the next topic is some FPGA dev boards that kind of come down the pipeline. So the first one I saw was, arrow, makes a $40 FPGA board now, okay, that runs off a Altera Max 10. Actually, no, it's $30. As I wrote down the wrong thing, $30 even cheaper. It's got built in USB programmer, and all this other good stuff. It's Hackaday did like an article about it. It's like, oh, that's kind of cool. And it's pretty inexpensive. And it's not super beefy. But it's like you can do like every single starter project, and you can do so much stuff on it.
Host 3 31:53
Could you still chunk a microcontroller core into it?
Host 1 31:56
Small one, maybe it's what's more like a max 10 is not really a FPGA. It's a CPL, D, okay, but CPL, ds now are kind of pretty hefty devices. Now. It's the 80 Tiny of the FPGA world, kind of what used to be CPL, ds were like, super cut down and really limited and what they have, but nowadays, especially the Max 10s, are like, basically just mini FPGAs. Now, right? They're just, they just don't have as much logic elements in IO now. So and then, so I was like, okay, you know, that's pretty cool. And then I was, you know, browsing for topics and stuff, and then saw this article on E web that just came out. And it's this product called Tiny FPGA that uses lattice FPGAs. And then you're talking troll bucks for an FPGA dev board. Wow, from this guy starting to come down cheap. Yeah. And you have to buy his $9 programmer, but you can reuse it. And I think he sells like on one that has the USB built in for like 2530 bucks. Okay, so go check those out.
Host 3 33:01
That's awesome. Yeah, in fact, gosh, over the Christmas break, you actually gave me a lattice? FPGA. Am I, too? Yeah, the lights too. Yeah, Digi key gave to me for free. Oh, nice. Thank you, Digi key I've been send me more free stuff I've been meaning to plug it in and dig into it. But I've been on a bunch of other stuff, I really want to get into digging through some FPGA stuff, just kind of get my feet wet on so much fun. It seems like it and from what I've what I've really kind of found out about FPGAs is that they were supposed to be in a way they were supposed to be the microcontrollers of the microcontroller world, they were supposed to, like, surpass microcontrollers, many, many years ago, or at least that was like the intent, but they just didn't because they they're expensive, and a lot of other limitations that come power consumption will and difficulty. Yeah, at the same time. And so the thing is, like, it would be cool to microcontrollers, that are not necessarily microcontrollers, that just pre programmed FPGAs in a way, where they just are presented to you as a microcontroller. Oh, you mean like ASICs? Basically. Yeah. It would be cool to see a bit more of that and and I wouldn't be surprised if we do start seeing something like that. So
Host 1 34:18
Cool. Awesome. So that will wrap up this episode, the Mac fab engineering podcast episode 105.
Host 3 34:26
So we were your host, Stephen Couric and Parker don't
Host 1 34:29
Take it easy guys later. Thank you. Yes, you our listener for downloading our show. If you have a cool idea, or project or topic that you want Steven and I to discuss, tweet us at McWrap or email us at podcast at my cred.com Also check out our Slack channel. If you're not subscribed to that podcast yet, click that subscribe button. That way you get the latest episode right when it releases And please review us wherever you listen. Whoa that changed podcast addict or iTunes. It helps the show stay visible and helps new listeners find us. Macro is also hiring apparently still
Transcribed by https://otter.ai