 Hey everybody and welcome to Desk of Lady Eda. It's me Lady Eda with me and Mr. Lady Eda on camera control. I don't think any of these are updates. Of course, Eda Fruit is semi-closed today. We had a full holiday. Yeah, so Juneteenth was Sunday, but Eda Fruit has Monday's official day off. So we're doing the show tonight. Yesterday we saw some art, checked out our blog and more. And normal show schedule this week, Wednesday show and tell, some 30. And Pedro will be doing their 3D hangout show, 11am, JP on Thursday. JP's back. Liz is back. I'm with Tim on Fridays and I'm an Ascan engineer on Wednesday night today. So we're doing this show tonight and we have some cool stuff to show. It's also a great search. 30 minutes of fun, great search as well. So let's kick it off with some samples. Originally this wasn't in the plan, but these came in yesterday evening. So I got these cool samples of knurled knobs. These are anodized aluminum, colorfully chromed knurled knobs. So five colors. You get gold, black, silver, blue. I like the blue a lot and this kind of deep red jewel tone. JP requested some more knobs because we have a lot of touch amateurs. And these are set screw, which is not unusual because it's easier to machine. It means machine on a lathe obviously. That's how you do knurling and how you get such even smooth shapes. Which means it's harder to get a D-shaft or T18 style mold inside. Instead at the end of the process you just drill and tap a hole in and then you can use this 2mm hex wrench. So let's say you have a D-shaft potentiometer. You put it in, slides in, but then you can tighten it with the set screw. And you've got this really cool knob. These are often used in very nice synthesizers and drum machines and other devices. You can also use a T18 knob. It's a little bit of a press fit, but it does press fit in. And then of course you can put the set screw on if you want to have it tightened so it doesn't slip. And I like the indicator because that means you can use it with pots or with the watering codes. I like to have a little bit of an indicator with knobs. So I don't know, people like this knob, these knobs? Oh yeah. They like them? Great. We'll stock them. So these are the samples I got. These seem good, so we'll get them into the shops. That's some samples. What else you got? Okay, so next up there's this part shortage. And one of the parts that's being affected right now is we asked about this on our weekly... 10 minutes of begging called part shortage where we ask companies, hey, can you please ship us the stuff that we ordered? We're still following up on that, but we are unable to get at Mega 328s, at Mega 328Ps, famous because they're used in Arduino Unos, like the most popular mic controllers in the world, the debt boards. We also use them in the Metro Mini, the Metro, and the Feather 328P. We both use a TQFP package and TQFN package. We don't really use a DIP package. We did for the Boarduino, but we just continued the Boarduino many years ago and now we just have the Metro Mini. The Metro Mini uses the QFN, the Metro, which is kind of Arduino Uno compatible. We made that basically when we were like, we want to have a microcontroller that has logic level shifting. You can switch logic level and also it wasn't clear if we were able to get Arduino during a very tumultuous two years in Arduino history. Those were just TQFPs, but we're unable to get both. Both were ordered about a year ago and the lead time at the time was 52 weeks. I remember everybody was like, 52 weeks, 52 weeks. I was like, yeah, sure. Okay, here's how many we'll need in a year. We ordered them back in March 2021 and it's been a year and a quarter and they got bumped to 2023, basically, or maybe even later. We're not going to get them anytime soon. This is a little bit of an issue because we actually ordered what we needed with the expectation that we did this thing called scheduling it out where we say, hey, we don't need them immediately, but we need them in a year because we'll run out by then, years gone by, no chips. It's an issue because we still have some Metro Minis in stock and we're probably going to have a couple enough for a few months. The Feather 328P and the Metro, both were out of chips and we were expecting them so we can't manufacture these two products, which is a little sad. I was sort of tasked with, hey, can you find a way around this? I'm like, yeah, okay, that's a bit of a challenge, but I'll see what I can do. One of the things that I found when I looked around is Digikey actually had some Atmega 328, not the P-Type, but the Plain 328 in 28QFN package, the MMH package, and they had like a couple hundred in stock, 500 in stock, and I'm like, well, at least keep us going for a few months. And what's interesting is Mr. Lady 8 was like, that's really weird. I thought that there weren't any chips. Why were you able to buy these? Why were they available? And I think the answer is a lot of people don't use the 28QFN package. So let's go to the computer because it's a little bit of an oddity. Basically, I grabbed them and then I was like, wait, what did I just buy? So let's see if I can, I don't think I can hide this side thing. So this is what a lot of people purchased. They purchased the Dip, which is 28 pins in a Dip format, or they get the 32QFN. This is very common. This is the most popular package. And then the other one that's popular is the TQFP, which has longer legs than the shows, but also 32 pins, same as the QFN, honestly. The difference between the two packages is this one has the two extra ADC pins. Anyone who's played with it at Mega328 knows there's these two kind of weirdo, like not GPIO, they're ADC only input pins, ADC 6 and 7. And I think there's an extra VCC pin, an extra ground pin. So these two pins are duplicated and then these two pins are extra. Otherwise, it's actually almost the same pin out. But because it wasn't the P type, which basically a lower power version, I think people didn't notice that these were available. And so I was able to snag some. So let me hold on. Actually, I left them over there. Let me just grab them real fast. Yeah, I remember you brought them home. I brought them home, but I put them away because I forgot that we were doing a video later. So I wrapped them in tape, which is not really static. So let's go to the overhead and I can show these chips. So this is the, on here, you can see the 32 QFN. So this is again, like everyone used this package because it was the most available. It was the cheapest, etc. And then this is the 28 QFN. So I don't know if you can see, but it's a little bit smaller. It's a little bit smaller. It's a finer pitch. It's, here I'll put it upside down so you can really see it. So you can see that the package fits, you know, when you can see all the pads, fits on top. It is about a millimeter smaller on each side and the pitch is 0.45 millimeter, not 0.5 millimeter. So it's a little bit, you know, finer pitch. But basically, I'm like, well, you know what? I, you know, I can't get the 32 QFN and it's smaller. And, you know, it's easy to revise a board if the new package is smaller. But what's a little sad is I don't believe, like I couldn't get it so you could have either or because the package is only like a teeny bit smaller and also like it's not one-to-one pins. Like there is a little bit of like a few pins kind of go side to side. So it is a kind of new board spin. This chip would fit inside of a TQFP but it won't fit inside TQFN. So it's too small. So let's show the do dev board. So the first one I started with is the Metro Mini because, and this is, you know, people are like, well, why, you know, oh, if it's you're swapping the part, why is this too big deal? Like why can't you just do it and why does it take so long? The reason it takes so long is it's like, you know, if you're going to go to a board revision, I'm like, well, here's all these things that I've been meaning to do. And I might as well just get it done now because I'm going to do a board revision. Like just do everything you want to do. So one thing is I'm not quite done with this design is, you know, updating this Metro Mini to use a USB-C, not a Micro-B. I changed a lot of the parts to be, let's see, it's a little clearer. I changed the parts to be 0603 rather than 0805. So there's a couple parts that are smaller, which actually gave me a little bit of space to move this 28 QFN over to the left because like these LEDs got smaller and allowed these capacitors that were like 0.1 microfarad got smaller. And that actually gave me space to move the reset button instead of centered, it kind of shifted to the side. And then I put a STEMI QT connector on there. So I'm not quite done with this design. As you can see, I'm still in the process of routing out the data lines for the USB-C. But, you know, I figured like, look, if I'm going to revise this, I might as well, you know, do an update. I always really wanted to have a Metro Mini with a STEMI QT port at the end. I thought that would be really useful. So, you know, that's why the revisions take longer. It's like, if it was just this part swap, I'd be done in 10 minutes. But it's like, you know, once you start cleaning out the fridge, like, oh, wow, look at this thing in the back of the fridge. Oh, you know, I should really like bake some pancakes right now. I don't know. Anyways, that's where I'm at. So that's the revision. So taking advantage, basically, of like, if I can't get a part and I find an alternative, you know, why not go in and fix it up? So this revision, I'm going to wrap it up. And then, you know, I'll, I want to get, of course, I'm not going to manufacture the Metro Mini immediately, but it's a good test bed because it's small. And then the Metro is actually the board I'm going to do the bigger revision on. The Metro, this is very tough because again, I can, there's a TQFP, I can fit a QFN in, but I can't fit both the 32 and 28. So I have to pick which one, which is like a little bit infuriating. It's like, I would prefer to have all three options because if it's going to be a year and a half to get more chips, I don't know what I'm going to get first. You know, a year ago, it was interesting, like a year ago, you know, when I needed some parts, once in a while, I'd email the company and be like, look, do you have anything? And one company, which was my name list was like, well, you know, we don't have the, like the B subtype, we have the A subtype of this chip and I was like, what's the difference? We're like, oh, there's this timing, they're basically a silicon fix. And they're like, as long as the silicon, if you don't need the new version of the silicon, because you don't care about low power, you don't care about some timing bug. They said, as long as you don't initialize it in this bad way and like, annoy the chip, you're good to go. And I'm like, this is not a problem. I'm just going to make sure I initialize it correctly. I'll work around the silicon bug, send me the 4,000 chips, send them to me, send them to me. And so I got them. But that's kind of over. You know, everybody's looked under their couches in the closet. There's no more just like mystery chips available. So that's the part shortage I'm dealing with this week. Very exciting. Okay. Any questions before we move on to... Nope. Want to do some great search? Yeah. Let's do some great search. All right. Here we go. The Great Search brought to you by... Did you know there's a part shortage? I did. I did. Do you do? Okay. Well, today is another adventure. And we're going to do some great search. Part shortage. I do. I do. Okay. Well, today is another adventure in part shortage land. So this week I'm looking for some alternatives to make, to use for the very popular MCP 23017 GPIO expander. It's a 16 pin GPIO expander over I squared C, which I really like. I've been using it for like over a decade. But it's totally unavailable. We have some on order. Not sure what I'm going to get them. Again, a lot of stuff from Microchip placed orders a year ago, and they've been pushed out. So let's go to the overhead, sorry, the computer, and I'll show what I would like to replace. Okay. So what's interesting is you can get the DIP version, but what I really want is this is a very popular breakout, which is great because it has 16 GPIO expansion, address, interrupt output. It's a very easy low cost breakout that over I squared C you can add 16 GPIOs. And here's the things that I like about the MCP 23017 and 23018. Sorry, 23008. But specifically the 017, 16 bits, sweet IRQ. So you don't have to constantly pull it when a button is pressed. Built-in pull-up resistors, which is big because there's the other GPIO expander that we stock, the AW9523. It doesn't have built-in pull-ups, and it's a little bit annoying. And I like that it runs on 3 volts or 5 volts, which is very handy because you can kind of use it a little bit as a level shifter, a teeny bit. You know, you can use it with 3 volts, but use it to power LEDs or send signals to 5 volt servos. So, sorry, not servos. People use it for rotary encoders. So I do like the MCP 23017. If I can get more, I would just use that unavailable. So let's find I squared C, 3 or 5 volt power and logic, 16 GPIO, I squared C expander with interrupt output capability, and ideally a small package like this, so I can make a similar breakout board, maybe one that's even dropping pin-out compatible that people can swap out while I wait for the 23017 to appear. Okay, so I did want to show, yeah, you know, there's no 23017. This is the chip I would normally purchase, and it's like, you know, 2023, maybe I'll get some in 2022, but I need them sooner than that. So let's go to IOXpander. So I just went to the same topic because I'm like, I might as well be there. And then this look for our standard, you know, we want active products, and I want 16 IO, although I might do a little bit more because, you know, look, it's 20, it's probably still fine. I'm going to not select this like eight output only input only, like I want them to be all GPIOs. Okay, so I'm going to apply. Next up, the interface. I want I squared C, you know, I squared C sometimes has multiple names. Sometimes it's called TWI to wire interface, to wire serial SM bus. It's all going to be the same. And then voltage supply. I'll pick that. I want to want the surface mount. That's important to me. It has to be easy for me to pick in place. And I do want interrupt outputs. And then the voltage supply, you know, I want to remember how to select, sorry, option click to select all the ones I wanted to make sure it goes from three volts to five volts. I'm going to pick these and I'm going to skip this one because I want it to also cover three volts. And let's see what we got. Okay. So actually there's a lot of options and there's quite a few that are in stock, which is cool. So let's look at only ones that are in stock now. And there's a quite few from NXP. I've used NXP expanders. They're great. Analog maxim. Also great. Let's, I want to keep, get something that's within the price range. Remember the MCP 2317 was like $1.25 per person. I want to try to keep it at that range. So I like to view prices at a thousand because the prices for one piece is always like, it doesn't really give you a good idea of how much it's going to cost. Some chip companies like the cost of one is basically the same as the cost of the thousand. Some you get pretty big discounts. So go to a thousand and then we'll see what's up. There's a couple of good options. The PCA 9535 is available. Comes in a couple of different packages. But the first one that came up was actually, looks pretty good. So this is the Essex 1503 16 channel. It's got 28 QFN. So I know it's about the same size, which is kind of nice. The pricing is really good. It's about 80 cents a piece. And then I looked up the datasheet and, you know, yeah, they're like, you can use this with three volt logic and then controlling five volt LEDs. There's a reset pin as well as an interrupt pin. There's only one interrupt compared to the MCP 2317. But I think that's okay. Lots of GPIO. They have versions with eight channel, four channel, you know, you want to have more pins, less pins. And another thing that I wanted to check for is pull up capability. Because that was like the thing that I really wanted. Yes. So programmable, true bi-directional style IO. So it means input and output and programmable pull ups and pull down. And all 5.5 volt compatible iOS. This is actually a really good, really good option. It basically has even more capability than the MCP 2317 because it actually can do pull downs as well. The MCP can only do pull ups. And I'm willing to kind of look beyond the fact that there is only one IRQ. I think it's not a big deal. And then I don't know if there's address. Yeah, there's one address pin. So you can have two of them, not as many as the MCP. But again, you know, it's available and I'm willing. Oh wait. So that's the, that's a one three. Yeah. So there's, oh, I might not be able to change the address. Hold on a second. Because the 15, the SX 1503, let's see if I can find the pinout for that one. Lock and diagram. Okay, so this is the four channel, eight channel, 16 channel. So it looks like there's GPIOs and there's a reset but there's no address. So this is a fixed address. So yeah, that's the one downside. You only get one address pin on 0x20 hex. But you know, I think it still might be good enough for most people if you don't need to chain multiples and then maybe like this could be useful for people until I can get more of those MCP chips in. You know, you need more GPIO expansion. I don't know. We'll get a grand central. Well, you can't get grand central because I can't get those chips either. And that's what it's like living in a chip shortage. But I'm still going to make a breakout with this. I think it'll still be useful for 75, 80% of people who don't need to chain multiple boards together. And we'll see it native food shop in a couple months. And that's this week's great search. Okay. Okay. So here's a question. Yeah. The seesaw is a really good concept. Why not use AT Tiny to replace the MCP 230X fixed function iSquared C devices? It's good, but it's not as fast. And it doesn't, you know, it doesn't have things like five volt IO compatibility. It does require a little bit of clock stretching. It's not quite as fast. There are side effects. The seesaw is very good, but I wouldn't use it as a plain GPIO expander, especially for 80 cents. It's a really good deal. Also, a lot of these have much more durable outputs. They can control LEDs, multiple LEDs. Where's the AT Tiny, which is also by the way hard to get and about the same price. You know, there's trade-offs, but I would, I tend to use the seesaw when I need an analog input or like some other stuff. It's really hard to beat a pure ROM silicon GPIO expander for speed and usability across like a wide range of platforms, especially when it's like, you know, clock sensitivity, clock stretching and GPIO strength. Just you're not going to be able to compete with a specialized chip. Okay. Also, I don't think the AT Tiny series has pull-downs. That is it. So that's our Desk of Lady Aida. We'll be on our regular schedule soon. I think for the Fourth of July weekend, we probably won't be doing a Desk of Lady Aida, but we'll be posting content and stuff. But we'll keep everyone posted. That's a couple weeks away, but just to give you a heads up. All right. Thanks so much, everybody. Love. See you soon. Bye, everybody.