 and welcome to Descalada. Hey everybody, and welcome to a Descalada, this warm summer, July, Sunday night. We're back last week, we took a little bit of a break. We did throw you a great search, a quickie, but we took a break from Descalada because we were out of town and then I also got my hair done. So, we're back on our bullshit, which is awesome. All right, well, Mr. Lady, do we have any news or updates? You want to just- No, let's just jump right in. Week of shows ahead. Last chance to sign up for Adabox, go to Adabox.com. Last, last chance. We're shipping starting next week. And as each day and week and month and year and soon decade goes by, there's only one independent electronics company out there that can get you all this cool stuff and I believe it's us now. So, please support us, buy some stuff on the website. Okay. Adafruit.com, okay. All right, sweet. So, let's chat about, okay, so I basically have two things I can talk about. I sent out some prototypes before we left. So, let's go quickly look at one of the prototypes that I did, that I did get on the overhead. So, we chatted, you know, a while ago. I'm going real slow on this, but I'm slowly but surely adding the AT Tiny. You can see it's like barely, I'm not gonna unplug this, maybe more visible. The AT Tiny 817 or 807. This is a cute little, I think it's a 28 or 24 QFN chip with the AT Tiny series that, you know, new, you know, Mega Zero, Mega One, TV Zero, TV Zero One series of AVR 8 bit chips. I'm looking to redo all my seesaw boards with this chip. Why? Because I can't get San Dio 9s for the next 52 weeks or more, some of them have been 80 weeks. So, I thought, you know what, let's just redesign these, I've been meaning to, you know, we've covered this for a couple of months, you know, a month or two. So, finally, I designed a breakout for this board, similar to my San Dio 9 breakout. I haven't soldered all these pins yet, I'll get to that. I usually don't solder them all in in case I have to remove the header for some reason, but I basically put semi-QT connectors on each end, which is kind of cute. This is good for prototyping I squared C devices, chip in the middle, regulator, although this runs on either three or five volts, so the regulator's kind of optional, I put it on there anyways. There's a power LED that's green, there's a red blinking LED, and then some resistors, and that's it. It's actually a very simple chip, you only need like one pull-up on the UPDI line, that's like the reset programming line, and you know, a power LED, and that's it. There's no crystal acquired, it uses an internal oscillator, it doesn't have external flash memory usage, like the RP2040, it's internal, and this chip is like 40, 50 cents. So it's a good deal. So I thought, you know, let's make a little break out with semi-QT connectors for me, and also stuck in the shop. So I'm very familiar with the AVR series, so that's easy, but there is something different about this family of chips, and that's how they're programmed. So traditionally, you have something like, you know, I've got here a Sam-D09, or Sam-D10, this is Sam-D09 chip. So this chip, you need to program it with two pads here, actually three pads reset, SWD and SCL, SWC, sorry. So SW, single wire debug, you need a clock and data line, and then a reset line, such a three wire programming setup, and then, you know, for, you don't have an Arduino, you can get an Arduino handy here, but if you look at a classic Arduino, they have the ICSP pins, a six pin connector with reset, clock, data in, data out, right? So it's got like, obviously, SPI and the reset line, when the reset line is pulled low, you can program the chip over the SPI pins. And with the Mega Series, sorry, the new Mega Zero, Teeny Zero, whatever, this new generation of chips, they do something a little different. They have a single wire, and like really, like, you know, the other, like SWD is called, like single wire, okay, this is really single wire. There's literally just one wire, right? There isn't a reset pin, even. It's one wire, and it's used for debugging and programming the chip. And they did something I thought was kind of neat. I think they did a really good job, which is, instead of coming up with SWD, which is a weird ass protocol, they actually said, huh, let's not invent something new, let's use something that already exists, which is, they did asynchronous serial. So the way to program the chip and debug it is done using basically a UART on one pin, so it either writes or reads, and it kind of flips the direction of whether it's reading or writing, but it's one pin using serial. And I believe in autobods, but I pretty much am using, I'm using 115.2 kilobot, but I think it can use a variety of different bot rates, but 115k, of course, is kind of a typical standard one. And so to program it, you can actually, it's kind of neat, you can use a standard, this is a CP 2104 FTDI cable, and call these console cables on the end. I'm gonna wrap this. On the end of this cable, you've got PowerGround TX and RX, and then you can wire this up with a resistor between the TX and RX lines, and so that way you don't have the TX length. When the direction flips, the TX out of the serial converter doesn't overwhelm what the chip is sending, so you basically have like a one, this is a 1k resistor, one to 10k resistor. And then you can just use a, the UR interface on that cable, I'm even powering it through that cable as well, and then if we go to my computer, we go fast, we can use a program called MC, I don't think I can go with this. I can use a program called PyMC Prog, it's available on PyPI, it's a Python program, and you tell it, you want to use UR, you tell it, let me hit flux, you tell it the COM port in this cave COM three, you tell the device that you think you're gonna connect to, and I'm just doing verbose printing and then ping, and it will connect to the device and it will say, okay, yes, I found the serial number, and I found the device ID, this is the fingerprint of the ID, and up here, I have a lot of debug output, so yeah, I found the teeny AVR chip, so this is all the data sending back and forth, and you can see it starts with OX55, which people who write code for UR to know that's a great auto bod byte because it has, it's a square wave, and then you can store and load data from the register, like a programming register. So it's an interesting model for programming, it's a good idea, I kind of wish they'd start with this instead of SPL because what's neat is you can program these chips, you don't need any kind of weird funky program or you literally use any USB to serial converter chip that you've got, including one that's built into your microcontroller, for example. So I've got here a printout of the data you're sending, it basically opens the COM port, and then you can see it stores data to registers ST and loads data, and then it kind of goes back and forth and it can transmit all the data and receive it, you can program chips, you can erase chips, you can, the whole interface, and then if you want to debug as well, you could also debug, I haven't done debugging with this chip, but apparently you can do step debugging as well using this interface. Only thing that I'm a little sad about is at least on the zero and one series, the UPDI pin, there's again, it's one pin, it's the reset pin and so there's no reset pin, like it's not reset, it's UPDI. So I think, I don't believe, I mean maybe you can use it also as a reset pin but I think you can only use it as one or the other and if you want to not use it as a programming pin and it's like a high voltage programming, it gets a little complicated anyways. So I wanted to add this to the AVR programming library that I use for in-house programming because that's like doing the hardware is easy but like how do I actually test the hardware? So this is, Codeye wrote a couple of years ago, it looks like two years ago and the thing it does is it basically just makes it very easy to from within and Arduino, a completely standalone program, program and AVR chip. We also have one for SWD for programming our SAMD chips with a bootloader but it's really handy to have a library that kind of does all the work and then I can, I like to store the Codem programming either in a header file just as like binary data if it's small enough or if it's large, you put on an SD card and then you can load the hex file and program it and basically not involve a computer at all in the programming process because that's what makes it fast and easy for us and reliable. You don't have to worry about a file gain deleted or corrupted or the computer updating operating systems or the COM port disappearing with USB dying, it's very standalone and so that's how I program all of our boards. So what I've been doing is I'm taking so PyMCU Prog is the tool that microchip actually publishes. It's from microchip pick AVR tools which I'm assuming is an official thing because it has a big logo and it's how you can use again the USB to serial converter with Python, very handy and then this person Brandon Lane ported it to, they have a product or project called Portaprog and it said UPDI and I was like, oh, maybe I can reuse some of their code instead of porting that Python code myself because I don't want to do the whole port. Like UPDI, it's a lot more dense, it's more complicated than the AVR programming system which was like, it had maybe a half a dozen commands. This is more intense and because it's asynchronous, you are, you have to read the data while you're writing it anyway. What's really nice is they published this under MIT license and it's really long and detailed and it supports all these chips and what I've been doing is I've been porting it over and so far with some success. So this is a QT pie that I've got wired up to an AT Tiny 817 and you can kind of see that I start the program mode and I look for an echo when I check for the UPDI setup and you can see there's all these commands. This one's kind of cute. This is a key, I guess you type in NVM, PROG and return and then it lets you program the nonvolta memory, that's the key that they use and you can see all the commands and details and getting the chip signature and then the next step for me is I have to, now that I've gotten the chip signature read, which means I'm communicating with the chip from the QT pie, writing Arduino, the next step is for me to kind of mush the chip erase, flash write and maybe even eProm write, eProm read into the library and so hopefully I'd be able to use the AVR programming library as like my single interface for programming these chips and that way as I design the chips, it's like I have the method by which I'm gonna program them in-house for manufacturer setup and I wanna do that before I start manufacturing boards because in case I have to change something with breakouts or pins or voltages, if I have to do a high voltage program or something, I wanna have that all tested and ready to go before I order the boards. This is like, once I'm used to a chip, I can start making boards without design the tester first, but in this case, I really have to get that tester code going. So that's kind of what I did this weekend and it worked really well. It was easy to port this code from port-a-prog over. It's in a branch and so far just getting it to respond and saying, hey, yes, I'm an AT2817 is a big deal because the toughest part is just getting that communication. I took a couple tries. There's a little bit of weirdness, for example, you know, you want a parity bit and it's two-stop bits. So it's not the standard eight and one UR, it's 8E2, which actually I think makes a difference. So yeah, great success. So this is gonna be hopefully merged into the AVR programming library very shortly and then I can go ahead and make tons of board of this chip knowing that I can deploy to production very easily. Okay, let's do some questions and then we'll go to the great search. Yeah. All right, hello, Aide for Loving Products. Have you ever considered adding planetary geared steppers? Not yet, we do have internal geared steppers, the little mini cheap steppers we have, they have internal one to 64 gearing. That works pretty well. All right, so when asked, can we broadcast directly into our live chat Discord channel so they can watch it on their cell phone. So Discord doesn't have the same video capabilities but we do post a link to the live YouTube video. So you could theoretically watch it live inside. YouTube, yeah. If Discord for your phone supports it to watch the YouTube. Oh, that's right, they have YouTube sharing. Yeah, you can watch videos, watch with friends. Yeah, and as the video platforms evolve, we'll continue to figure out more places we can broadcast. Okay. Does the Sam D09s become available earlier than expected? I don't think that's gonna happen and I'll show you why, but regardless, these chips are less expensive and totally capable. And I like that they're three or five volts and they have an EEPROM. So in a sense, it actually is a better option for me. So I'd like to, you know, I would have done this eventually but my hand was sort of forced a little bit. I'd like the power of the Sam D09, but you know, in hindsight, I don't actually need that much power for the stuff I'm doing it. It's just like, when I first did CSI, I don't think the 817, I don't think these chips existed yet. And you know, the Sam D09, it's like, it's a wonderful chip. It's 48 megahertz, it's got 2K of RAM, it's wonderful. This only has like a half K RAM, but it's still good enough. All right. And so it's amazing, no fancy programmer needed. How did it take them so long to do this kind of thing? I don't know, but I think that there has definitely been an evolution in like programming modes for chips. Like this is not the first UART chip, like I saw also the 8051 Bumblebee from SILabs, the BB series that they have, is also I think programmed over UART. And then, you know, we're seeing the RP2040, which allows programming over USB. I think people are getting, I think there is this desire for like, can you make chips that are easy to program and don't require special weird ass tools and even better, you can do it over USB. That's just not USB, so it doesn't make sense. But I really think, and I know that they have for the, at mega series like the 328, there was debug wire, but I never saw a single person use it because it was like under NDA and like weird. And this is just a simple asynchronous UART system with auto bot detection. It's just like, it's really trivial. So in my, you know, my sketch for programming it, so I've got like this UPDI test code that I'm writing. I'm just using serial one. You know, I just used the hardware you are. It's really nice. So you can't bit bang it, but like there's, you know, for a lot of my boards, I have an extra hardware you are available. We'll do one more question and then we're gonna do the great search. Will there ever be a trinket with BLE? Maybe, but I would go with the itsy bitsy. It's like, we really like these modules and it's as small as you're gonna be able to make it really and still have, you know, all the things you need and broken out. All right. Okay. That is the question so far, but we're gonna do the great search, ready? Yeah, we're gonna transition into it. Where are you? The great search brought to you by Digikey and Adafruit. Thank you so much. Digikey, every single week, lady it uses our power of engineering, smarts and others focus to show you how to find stuff on Digikey. And this is a series that we started way before the chip shortage. So if you watch this from beginning to end, you'll probably be able to find that thing you're looking for, because everything is hard to find now. It's true. When I do the chip shortage chip today, but we're using something that isn't a chip shortage, although I'll show this trend thing. Since we've never had, and I'll show what I've been hacking on. So I've been making some more boards. And two of the boards that I designed, and I sent off before the holiday break and just got back, are these real-time clock breakout boards. So these are STEMIQT boards with my two favorite RTCs. This is the PCF 8523, and this is the DS3231. This is a temperature compensated and more expensive RTC. This is a three or five volt non-compensated, and it's still pretty good quality, easy to use, inexpensive, just to pop on a 32 kilohertz crystal, and you're going to be able, this one has the crystal built in. And for real-time clocks, you really want to have a battery for battery backup. The whole point of the RTC is that it has a battery backup and keeps track of time, even when the rest of your system is depowered. So you usually have a separate coin cell battery. So let's look at that. Nice little coin cell battery. So there's a couple of different ways to do it. There are pick and placeable, non-rechargeable batteries, primary cells, but I do like these CR1220s. I find these to be great little coin cells. They're extremely common. You can get them kind of anywhere. They're three volts, so they're perfect for real-time clock, real-time clocks like running off of these. But again, they're available and they last like seven years. But if they do get damaged, you can replace them unlike the ones that are stored onto the PCB. So I do kind of like having these removable, replaceable coin cell batteries. This is a lithium cell. And you can see barely on there, it's a CR1220. So that means it's 12 millimeters diameter and two millimeters thick. Watch out. There's also a 1216, 1.6 millimeters thick like maybe even 1212, 1.2, very thin, but they might slip out of the coin cell holder, just make sure. So you pop it in the back like this. You see there's a plus symbol here, plus symbol here, you match it up, you push it in. And I kind of like this style and design and you can always kind of get in here and pull out the battery with your fingernail as well. There's also these kinds of coin cell holders. These are the kind of the two that I see the most. Like there's more, but there's the from the side ones and there's from the top ones, from the top ones. The battery kind of snaps in like this. Which one to use? Well, I'll tell you, if I have side access, I kind of prefer these. It's just my personal preference. These are a little bit thicker, they're a little bit bigger, but they're not too hard to use to remove the battery. Just kind of push open and the coin cell pops up. But for some reason, I sort of prefer these. I think they're a little bit more secure, although these are also very secure. But they're two options. So let's show on the great search how to find these two coin cell holders. But first up, I want to show you, this is from Jepler Point of the Sells. This is like, did you get a lot of hidden gems? This is a hidden gem. Lead time trends. And this is very handy for you to know, what are you likely to get some chips? So here's some, whoa, pardon me. Went down too fast. Some things. So one thing you'll notice, I've noticed this as well, crystals. You know, timing crystals. The lead times are up to 42 weeks from Epson, 24 weeks from Abercon and ECS. NDK, if they got them, they're three weeks, so buy from them. Ceramic capacitors, you know, also 24-week lead time. Still, this is a little intense, but still, you know, they're available in the market, although I'm seeing, you know, crystals be difficult to get. Inductors, you know, not too bad, 16 weeks. Ribbon contacts, not too bad, you know, 10 weeks for Rolex and Amphenol. You know, some of these are a little bit longer. JSD, JSD always has sort of long lead times. You know, Hardware connects. Phoenix has only seven weeks, Harding 4.3. So not too bad here, but what you're gonna see, the T.E. seems like as a company is having some long lead times. If they ever appear, they're always longer, like headers are. Their headers are 24-week lead time. But what you'll notice is where it gets really, a little bonkers is chips down here, semi-conductors. Op-Amp's aren't too bad, you know, 20-week lead time, but switching regulators. So MPS, a common regulator company, 50 weeks. Microcontroller, ARM Cortex, STMs, 45 weeks. NXPs, 51 weeks. Microchips, 41 weeks. TI, 20 weeks. I think that's a little, it depends though, right? Because TI doesn't make that many Cortex chips. I think they make mostly MSP430s. And then Xilinx, you know, 52 weeks. So you're gonna see really long lead time still for semi-conductors. This is good to test. Also they have by company. So, you know, common companies, if you look at, you know, microchip, you can look at their 8-bit micros, 40 weeks. Arms, again, 41 weeks. E-Proms, 40 weeks. 32-bits, 40 weeks. FPGAs, 20 weeks. So yeah, you want FPGAs, good to go. Looks like TI actually is not too bad. 14 to 20 weeks. And all devices also not too bad. ST, you know, everyone knows that the ST microcontrollers right now are just impossible to get. So it doesn't look like it's going to get much better for the next 30, 40 weeks. So hopefully find some alternatives or maybe you've got some stock. Okay, so but back to coin cell holders. So let's look at coin cell holder because that's what we're here for. So you can get, you know, you can get batteries as well. Even though this says search for holders, don't forget, you can also get your batteries here. Although note, you can't ship them air. You'll ship ground. So just be aware of that. Your cell has a little, like, happy family here. Yay! Okay, happy years. Yay, I'm taller than you. You can't have this battery. I know what's going on. It's teasing the person, it looks like. It's a little mean. All sorts of batteries. But let's go back to, but again, what I like about the CR1220 is you can actually get them at a grocery store. They're that common. All right, so clips and contacts. All right, so the first thing you'll want to do is of course, let's only get active ones. And we're going to look for the ones for our CR1220. Don't forget the 12 is the 12 millimeters diameter. So let's look for 12 millimeter diameter coin holders which is battery cell size. You can see that. Okay, let's go down. 10, 11.6, 12, 12.5. I'm going to grab 11.6, 12 and 12.5. You know, I don't ever see a 12.5 millimeter battery, but you never know. Maybe the holder can have both types. All right, so we only have about 65 options left and you can see, you know, we're starting to see some of the ones. This is a two hole one and this is again that surface mount one. One thing I've noticed about coin cell holders is like, they're kind of jelly bean. If you like this style, again, the top press in style, you can get them from like 10 different fabrication companies. Let's go for the surface mount because we want it to be pick and place of all. And then you can select different companies. There's a battery series. Again, there is a little bit of a difference between there's the 12.16s, 12.20 and 12.25 that's the thickness. It may affect your holder. So just be aware, they're not necessarily, I have been bit by, I got a holder and it was for 12.25. I tried to put a 12.16 and just like slid right out. So just be aware of that. Okay, so there's a couple of different kinds. Again, there's this snap top kind, a couple of different companies make them. They're very common. There's also, this kind is also a sort of snap top type. And this one, which is like really big. It looks a little bit like a spaceship. But I rather like the, I really do like these bent metal ones. Links and MPD both make ones with a smaller tabs. But I really like the chunky tabs. Oh, also you can get ones with little holders. You're probably wondering, why would you ever want that little plastic piece? If you want to hold two batteries in, you can't have metal that will contact the outside because it'll short the middle battery together. So you have to have something protecting it. If you think about it, you know, the outer ring is a pole, right? It's not like a double A battery where like you have it on the ends. And so you can have the batteries next to each other, not going to short with these. If you have a bunch of batteries stacked up and the metal is against the side, it'll short the batteries together. So you would need that plastic piece if you had multiple batteries. And then there's some simple ones. These, I don't like these as much. These are a little bit too simple because there's no mechanical, you know, thing holding them together. Again, I like these because they've got that stop. You know, the battery can't push too far back. So let's look at ones that are in stock and let's sort by price. And it's my favorite, the AtomTech. So this is the one I like. This is the one you saw on the overhead. Thank you. Again, these come in 20 millimeter size and like maybe 24 millimeter. They come in various sizes, but for real-time clocks, I really like these. It can't be the price, like a 10 cents for a battery holder and super easy to solder, super easy to pick in place and they're really near indestructible. I've never seen them break, you know, unless they get completely smashed, they'll hold your battery in place for a very long time and it's very hard for the battery to shake loose. So this is my pick for the red search, coin cell battery holders. And that's a great search. Alrighty. Okay, 30 minutes, we're doing good. Yeah. On the dot. All right. All right, any last questions before we roll out? Yeah, we do electronics. People are asking if we do electronics, so. Yeah, we do electronics, really. Yeah, this is what I do. This is pretty much what we do. That's pretty much what I do. All right, we'll see everybody during the week. We've got a lot of cool stuff going on. Thank you so much, everyone's supporting the only woman-owned electronics company in the world. No, I think there's other women-owned ones. Kittenbots women-owned. Yeah, in the U.S., I don't know, maybe. I don't know. Look, man. We have a weekly live show, yes. Look, I'm just saying that someone will correct me if I'm wrong on Twitter. Great, that's fine. So that's why I'm just putting it out there. Go for it. Okay, see you everybody next week. Bye.