 All right, so every single week, everyone's talking about the chip shortage. Everyone's saying, oh my gosh, I gotta find out where I can get stuff from Digi-Key, substitutions, part replacements, you name it, and there's only one weekly show that has a weekly segment. The Great Search, brought to you by Digi-Key Native, where Lady Aida uses all her powers of good to find the things you're looking for at digikey.com. So. Only powers of good. What are you doing this week? Okay, so let's go to the overhead and I'm gonna show this board off. So this is a board that is, it's okay for now. This is my little iSquared C to Rotary Encoder converter. And in the middle here, I've got my SAM-D09. The SAM-D09 is a very low-cost ARM Cortex M0 chip that I got great pricing for and I use that to read the Rotary Encoder and then I act as a iSquared C peripheral so you can query the peripheral about the Rotary Encoder and that's not having to do the Rotary Encoder math and pin toggling measurements and, you know, pin changes and neopixel controlling. That stuff's all handled by this little co-processor that's controlled over iSquared C, wonderful. And I designed like a whole bunch of different boards that were like it. Like I designed this board that has like four Rotary Encoders. This one here, which is still brand new when I designed one that has a slide potentiometer some folks have seen over the last few shows I've designed them and I designed them all with a SAM-D09. And that's when the SAM-D09 got kind of hit by this part shortage. And so, since these boards were just designed, I don't feel too bad about swapping chips before I order PCBs because if I can't get a chip, there's no point in ordering the PCBs on a LARC. I should get something that I can actually acquire in the next year or so. So I wanted to, there's also a couple things I wanted to improve on this design. This ordering code by the way is like a three year old design. I finally got out the door, which I'm very happy about. But there's some things that can improve. For example, the SAM-D09, now it's a wonderful chip, but it's three volts only, which means that for it to work over STEM at QT, the iSquared C, I want it to be three or five volt compatible. So I have to have a level shifter MOSFET over here and a regulator over here for level shifting and regulating the possibly five volt power and logic down to 3.3 volts. So one of the things I would like to find is an alternative. If I'm going to get an alternative, I really want to save 15 cents a board and skip those parts because I need fewer capacitors that way, fewer diodes, fewer placements. The boards can be smaller. You can see it's kind of packed in here, but I still wanted to be able to do all the things that I was having the SAM-D09 deal. And finding a chip that can do that, I thought I found a good chip and then I'll show you, it turns out that it was not actually a very good option because I can't get it in any future universe that I can see. But I want to find basically my controller that is physically about the same size as this. It's not going to be pin compatible. I'm not expecting that. Three to five volt compatible has eight to 16K of flash. This has 16K of flash memory, but I don't actually need all 16. I think I can do with half as much. I do need a bunch of RAM. I need at least like 256 bytes of RAM. It doesn't sound like a lot. The iSquirt C buffer itself, I'd like to be, you know, it's going to be 32 bytes in and 32 bytes out. And then I want to control Neopixels and that can be another 64 bytes or so. So it's going to add up quickly. I need 256 bytes at least. I wanted to have an internal crystal. One of the things that I really like about this is you notice there's no crystal or resonator. Those are 20 cents. So I want to cut that out too. What I really like about this is that other than the regulator, there's only like, you know, one or two resistors and capacitors needed to get the SAMD-09 up and running. So let's get, let's find something that is similarly very easy to get up and running. Needs very minimal additional hardware because I don't care if the timing is a little bit slow because it's going to be 100 kilohertz iSquirt C peripheral. Doesn't need to be perfect timing. General question. Why is the processor rotated 45 degrees on the board? Because it otherwise it would have gone in the way of the pads. If you look at it there's, if this was not rotated 45 degrees, it wouldn't fit and wanted to fit. There you go. And yeah, it just, it wouldn't fit. Okay, so let's go to Digikey and let's search for this microcontroller. And I'm going to be really flexible about the microcontroller because, so I'm just going to search for microcontroller which I usually don't, right? That's usually- Billions. It's like, you know, yeah, 900,000 results. So there's a couple of different options here but I really just want embedded. I want an embedded microcontroller. So let's pick that. So yeah, 90,000 options. So let's zoom in a little bit. So we're going to pick only active parts because I want something that I can get in the next, again, year or so. I don't want something, if this continued now, isn't ever coming back. I'm not going to pick stuff that's in stock but I'm going to pick something that's normally stocking, right? Because obviously there's a lot of stuff that's not in stock but might be in the next few months. That's okay. You know, if I can get samples I can always scrounge up some samples. I can get the designs developed and then by the time the PCBs show up in a month or two, I'll be able to go into manufacture. So this already cuts it down by three quarters. So next up, we're going to actually start picking up specifications. So for GPIO, let's pick that first because it's a very easy thing. I want at least 12 GPIO. I actually probably want more but the reason I want at least 12 is, look, I need two just for the programming pans. There's a couple of programming pans. There's a reset pan and then I want it to be able to control a couple of, like four buttons and maybe four neopixels and then I want to have address selection pans and the I squared C pans and then maybe an IRQ pan. Basically you do the math and it's like, I need maybe an activity LED. I need basically 12 GPIO minimum because I don't want to have a different chip for each of these boards. Like one for the rotary encoder and one for the one with four keys. I want to have one chip problem. So I'm going to pick 12 minimum. So I'm going to use the Min Max boxes down here. Great. Next up, I definitely only want to have a surface map part. I am not going to do the whole list, so surface map. For oscillator type, I'm going to want, I don't know what dash means, but I definitely want not external only, like external or internal spine as long as internal is available. For program memory type, there's a lot of these types that basically I just want flash. I don't want EEPROM or UV or OTP or Masgram or FRAM. Those are all going to be very expensive. Flash is the least expensive, most common. So it doesn't actually cut away a lot of things, but just gets rid of the ones I'm definitely not going to use. Okay, so next we can start doing program memory size. So remember I said I need at least 8K. Now, of course, it depends on the compiler and the optimization and the core and how many bits per instruction, but 8 to 16K is kind of what I want, right? Because also if you think about it, you know the original Arduinos were 8K and then there's the Amiga 8 and then there's the Amiga 16. And the 32 is wonderful. I mean, I can keep going, but it's going to get more expensive because what you're paying for when you buy a microcontroller is IO pins. It's not really peripherals because the peripherals are the peripherals. You're paying for flash, rim, and pins. The more you want, the more it's going to cost. But let's just go to 32K. So now we really cut it down because we've gotten rid of a bunch. Okay, so finally, RAM. Well, all these are, you know, these all have a fair amount of RAM. EEPROM, you know, if it has EEPROM, great. So the next thing is I'm going to do that power supply because remember I was kind of picky about the power supply. I really want it to run from three to five volts. In this case, there isn't like a min max. So I'm just going to pick out the ones that go up to five. So here's five, and this goes up to five, 5.5, 5.2. The 5.5, so it's going to kind of go through and pick up the ones that you want. Oh, as you're doing this, you have not been checking the exclude marketplace and all these, is that to add possibilities during the shortages? Yeah, I'm not going to be picky about that yet. I get picky about marketplace later and also usually when, again, I need something in stock right now and I need it like tomorrow. If this is something that I'm planning for long term, I'm going to pick normally stocking because that means it might be out of stock today but it would be in stock tomorrow. And marketplace, you know, I'm seeing more and more stuff showing up in the marketplace. It's not a bad idea and the pricing can be better in the marketplace. So the oscillator type, what's the difference between external, internal and internal slash external? It means that you can have either. It can support, some chips don't support any external oscillator. Like you can't put a crystal on them even if you wanted to. So I don't care if it allows me to, I just won't connect pins. I won't connect a crystal to it. Okay, so I picked up all the ones that cover three to five. There's more here, but of course these go above 3.5 or 3.3, so I don't want them. Okay, so now we're down to about 3,000. Okay, so next up, I'm actually not gonna select a core processor yet and I'll show you why. It turns out it doesn't really matter. I also don't care about the speed so much because all of these are well, they're basically eight megahertz or above, which is plenty fast for what I'm doing. All of these, I do like that these all have data converters. I'm going to pick only the ones that have ADCs. Some of these have like ADCs and DAX and that's wonderful. They're all kind of a little different. Okay, so selected all of them, except for the one that has a dash because I don't want any that don't have ADCs. Almost all microcontrollers these days have ADCs. And then finally I'm gonna do a supplier package. So if you remember that board, I'm actually very picky about the package. I really, really want something that's in QFN like 16 to 32 QFN because those are very small. The yield is very good. They're easy to work with and they're gonna come in tape and wheel and I just like them. I don't want SOIC or SOP, they're gonna be huge. I want something very compact. And I found that QFNs, these are sort of my favorite packages. So I'm going to pick out from the supplier device. I'm gonna skip the SOICs, TESOPs and BGAs. You know, depending on your soldering, sometimes people are like, no, I only want SOICs. I want them very easy to solder. Great, go for it. I want something very compact. So I'm gonna pick out the QFNs, QFN, and then you say like a VQFN and VQFN, that's like the density, but I don't really care about those. So there's quite a few of them. That's okay, let's go get through them all. QFN, QFN, not QFP, not QFP. And then a couple more QFNs here. Although now I'm getting into like, once I'm getting into this many pins, I don't need something with this many pins. 32 is actually kind of the max. Although 33 usually just means the ground plane. So I'm gonna, I'll add that in. Okay, so now we've really cut it down. So now we're down to like 400 options. So we're not gonna get all the way down to like two. Right, that's not gonna happen. However, we can start now. Let's look at prices at 1,000 pieces because that's what I care about. So there's a couple options here and I'm not surprised by the options. So the first one is the MS-51, which nothing beats how inexpensive it is. However, when I went to put it in my cart because it's like, well, it doesn't have the lead time here, but I do a little trick where I go, I put it in my cart and then I click the check the lead time here for 1,000 pieces. And it says it's not gonna show up here till 2023, which I'm not even sure I'm gonna be alive then. I have no idea what my life is gonna look like in two years. So that's kind of scary to me. So that's why I did get a dev board for this before. I saw the lead time, but it's a little bit too scary. For the rest of these, a lot of these were really good. I will say I selected, yeah, normally stocking. Yeah, so all these looked pretty good. There was the, a lot of these had EEPROM, which I kind of liked. And once I sort of saw that there were ones with EEPROM, I sort of started thinking like, well, I kind of want to optimize for those. I want to have EEPROM because it is kind of nice if I can save the I-squared-C address or configuration in it. So I started like, once I saw that there were ones that were with EEPROM, I said, okay, well, let's filter out the EEPROMs ones. So the next one, so that actually got rid of the MS ones. So there's the ST8S. This one also had a pretty scary lead time. But what I liked is that there were a bunch of AT-TINY 806 and then 816s. So these were the ones that were available and were about like 60 cents in quantity before I even got a quote. And given that I had a choice between basically for the same, it wasn't that much different in price, the STM8 series and the AT-TINY series, I actually ended up saying, okay, well, I'm probably gonna try the AT-TINY series just because I know that there is an Arduino core for them that I can use so I don't have to write as much code for them. The AT-TINY 816 in particular, I liked it had both capacitive touch and a DAC. So I ended up getting that one to start because I do use a capacitive touch for the soil sensor and I'm probably gonna have to replace that one. I'm probably not gonna be able to get that chip either eventually. So I ended up getting, this is a sample and what was kind of nice is for the, in my cart, one second, go to my cart. When I checked the lead time, it was like, well, it's gonna be August, which was kind of like, which is actually pretty good because a lot of chips are not available till next year. So the fact that they're gonna have it in a couple months was kind of promising to me, I liked that idea and they did have SOIC versions of this chip and stock which I ordered. So I got a dev kit for this board. So you can show it on the overhead real fast for this family. And this is the 817, which is a little bit bigger it has the debugger built in and has capacitive touch which I thought again, it was just kind of a neat thing. But then all the GPIOs and this even has like an Arduino E pin out. So I can get started with this and there's the X mega core which uses this chip to give it like an Arduino compatible core. So I can use things like the wire peripheral code. I don't have to do a lot of porting from the existing seesaw code I have. I can port it into Arduino which I'm very comfortable with. I should be able to do that in about a month and then I can kind of redesign all the boards that I just did into this new chip set. And I like it. It's a very compact chip, very inexpensive. It's about 50 cents per has everything. And then to program it actually uses a one wire program which is asynchronous UR. So I don't even need to have all six SPI ISP pins like most, if you remember most at megas or at T tiny's you have all six, you need to expense clock boasting me so and reset. You don't need those. You only need the reset pin to program it. So this is my great search. I think this is, I think this chip is the right chip. So I'm going to get some. And if you're looking for a low cost microcontroller I think this one's going to be in stock. So check it out. And if you're not looking now, no worry. You will be soon. Okay. That's the great search tonight. All right. Where in the world is that great search