 And welcome to Skully Data. Everybody, and welcome to my desk. So this week, I did not do as much hardware as I usually do, because I was working a lot on some internal tooling for Adafruit, because I design all this stuff, but I also have to make sure that we have it in the store. And so one of my goals this week was to get our in-stock numbers much better, now that the part shortage is really over and I can get parts. We've been at around 85% to 90% in stock, but I'm trying to get us up to 95%. I've been working a lot on process improvement and stuff, which is cool and great, but uses all our internal tools, so can't really show it off. So the next stuff I've been working on is, first up, I've been bringing up the floppy board. So let's go to the overhead and I'll show off this prototype. Okay, so I've got the Proto for Flopsy or my floppy drive board. And this is based off of we did this floppy featherwing for folks who remember, so this is the floppy featherwing. We did this a couple of years ago. I couldn't get all the parts for the power supply that I did for the floppy board. So we just did the featherwing to start and to be like, let's just get this going and get the library live and then we'll deal later. So level shifters, because floppy drives use five volt logic and feathers are 3.3 volt logic. So up shifters, a little five volt logic, bias generator, so it takes three to five volts to give you a nice clean five volts for the voltage level, not for powering the board. This won't be able to power a floppy, but it will at least get you like a clean signal. And then a write enable switch. So you can manually disable the ability to write to the floppy disk. So you use your experimenting with floppy integration and stuff, if you wanna make sure there's absolutely no way you could accidentally overwrite a floppy, just like literally disables the data and the write enable pin on a floppy. And then on the other side, the two by 13 headers. So this board is more integrated. So I've got the USB type C data going to an RP2040. Going to an RP2040. And then this is the power supply. This is a buck power supply. So if you're using large disk drives, like this one, like this one, this is a five and a quarter, not a three and a half, you need to provide 12 volts as well as five. These smaller ones, this is a classic three and a half. Use only five volts about like an amp-ish of power, but the five and a quarter and the eight inch require 12 volts. And so if you wanna connect to those, you'll need to provide the external 12 volts because there's no easy way to get like 12 volts, one amp out of a five volt power supply, unless you use USB PD, but then not all computers support it. And this is like, I went back and forth and was like, you know what, just get a six external 12 volt DC power jack, super easy. But also you can buck down from 12 to five. So if you were powering externally with 12 volts, you could also have the 12 volt, generate the five volt cleanly. And then here is your kind of standard Molex connector. And then you can just use like a Molex splitter cable to get it into the right format for the floppy drive. And then SD card for saving or if you're emulating, maybe you would have the images on here that it would emulate 16 megabytes of flash. So even if you don't have a micro SD card, you could still save or read images from the internal flash and also for, you know, program memory. This is a power monitor. So it'll monitor the five and 12 volt DC supplies because one thing I definitely learned as doing floppy stuff is that the five volt power supply is kind of sensitive. Like if it dips, you'll actually get bad reads. Like the floppy driver will still work, it'll just give you bad data. And you'll be like, oh my God, my floppy disk is bad, but turned out it's like, no, you need a really nice, strong power supply. Because think about it, most floppy drives were in towers and they had those really chunky, you know, IDX, I don't remember the name of them. There's like the name for the big power supplies. They were easily able to give you like 10 amps at five volts. And so you'd have a nice, strong power supply and that wouldn't be as much of an issue. Semi-QT connector for like connecting other sensors or displays or, you know, whatever, temperature sensors, I don't know. And then also laptop floppies are supported. I do have a couple of laptop floppy drives. So even though like, you know, they're not as common, this connector is like pretty cheap. It's only like 30 cents. And I think it's worth it to support that pinout as well. So that's just an FPC version of this. And then I still have the right enable switch so you can allow enables right now. I don't allow them. And then for the shift, the level shifters, these are, you can see that this one has a couple extra pins. I don't know if you can kind of barely see they're a little bit longer or a little bit wider than these. These are bi-directional. So there's another pin for floppy direction. So let's go to the computer and I'll show that off. So over here, you've got the three-volt signal and the five-volt signal. Side by side. And then you've got like pull-ups or pull-downs because these are open drain. I don't remember if this is the right direction. Yeah, I think so. Yeah, because normally you want it to be high. And then so that way, even if it's not an output, it will be, it'll have a valid signal. Oh, this is an underline, that's why. Okay. And then over here is the five-volt level. And then this is a dual supply. So you have VCCA, VCCB, and it will, using the floppy durpin, I can determine whether I want to like read from the floppy, or read or write to the floppy disk in the drive or that connector could be used to plug into a computer and then it would emulate a floppy drive. Now I don't have that code like at all working, but like I was like, hey, like technically, like it could maybe sort of kind of work. So like let's add it in and see what happens. So what I'm doing right now is like, you know, this is a rev A. Rev A is me just like barely trying to get stuff kind of sort of working. So for example, I do have like graphics tests. So let me upload it. So one of the things I like to do is I like to have in my sketchbook I have what I call like all my, you know, my board support tests. So it's like, I try to test every little element of the board. So like I have a test just for the NeoPixel of a test just for the SD card. And this is a test just for the TFT. And that way as I do revisions, I go back and I just test each element individually, not necessarily together, because if there is some like weird like, oh, I accidentally, I swapped pin names or I moved a pin around and I forgot, I can just like debug things one at a time. I think that's like one, definitely one skill I try to teach people, especially when they come to me for help with a project. I'm like, well, can you like verify that each individual part of the hardware works? And they're like, well, one gigantic program and it just doesn't work. And I'm like, well, like who knows, right? So this is the graphics test. So you have the overhead again. Oops, yeah, sorry. Can you go to your head? Thanks. So this is just a graphics test. So you can see like, okay, great. You know, I know the backlight works, I know the TFT works and that's just doing a little inversion test. I did the same thing with the SD card, make sure it can read the SD card, make sure that I can read a sensor from I squared C. And then, so the next thing is of course getting the floppy drive to work. So I haven't gotten it working yet, but that's okay, right? This is, I'm a hardware developer. I'm developing actively. So I've got this old code that definitely worked on the floppy featherwing. But there's a couple of things. First off, I could have made a mistake with the pin ordering or the connection or whatever or like something isn't wired right or I misnamed it. Another thing is it's been a couple of years since I looked at this code, at least like a one, I think it's been like one year I looked at this code recently. And there could be board support package stuff that's changed. So if you go to the computer real fast, you know, and I can hear it. I can hear the floppy drive is indexing and it's moving, but the data isn't right. And so you can see it's like, I couldn't read any sectors. One, my disc could be bad. And two, the drive could be bad. But I think it's unlikely. I mean, this has been in storage, but I don't see any reason why every track would be bad. It's capturing zero sectors, but it is getting an index pulse. So it got indexed pulse, which means motor is working, direction is working, index in is working. So like I got like data out and data in is working. So like the direction pin is correct, but I'm not reading the data. It's actually quite likely that something changed, maybe in the board support package with the PIO assignment, because I'm using PIO to read the floppy pulses in. So, you know, that's part of the hardware system. So we'll get this up and running soon, promise-ish. Okay, so next up, something I'm kind of starting on, but not ready yet is in the shop, we have Phona boards. And especially the Phona 800, I really loved it because it was a very small, easy to use cellular module. Cellular module, it did GSM, 2G, and it was really inexpensive compared to like, you know, 4G or 5G modules, or sorry, 3G or 4G modules. Promise, as of like a year ago, you really cannot connect to 2G and US. Now it was discontinued in like, you know, whatever, 2010 or 2005 or something. And it ended up 2010, but you know, for like 10 years, you can kind of still connect to it, especially like in New York City, there were still at least a couple connections. But really lately it's been impossible, like nobody will even let you turn on a SIM card for 2G. And the 2G spectrum is being reassigned to LTE. So, you know, it's time to revise these and update them. And I want to skip ahead, like, we do have a 3G Phona, but honestly, like, it kind of sucks. I mean, it's massive, it's extremely power hungry, and like 3G is also disappearing. And so, I thought I'd skip ahead to LTE, N-B-I-O-T. So this is cat M, LTE. And so I like SIMCOM, I've used them a lot, the SIM 800 and the SIM 808 and the SIM 3G, whatever it was. All are pretty good. So, you know, we had a friend of the fruit email us and be like, hey, can you like, re-design the phone out? I'm like, yeah, I should get on that. So this module, one of the things I like about it is it's very simple, it's got USB for power. You still need a battery to like handle the two amp burst, but there is a power supply, you could set up for five volts buck conversion. It still has some things I'm used to, like the power key, the ADC, there's a USB interface, multiple URs, one for the GPS unit built in, one for the cellular AT commands. I think USB could also communicate that way. SIM card, that's a net light. And then that's a little bummer is that audio, one of the things I've loved about the SIM 800 is it had speaker and microphone, like literally, like you could just hook up an electric mic and you could have voice. And one of the things I wanna make sure is that if I do have a cellular module that you can use it for audio, because data is cool, but like, I really miss being able to do, like it's nice to be able to do audio calls as well. So you can actually make it your real cell phone. So the only thing is this doesn't, this has PCM and I'm not a hundred percent sure that this can make phone calls, but it might be able to, there's a PCM interface, which is like, I think it's an I2S, you know, it's basically I2S implementation, data in, so that would be microphone, data out, which would be speaker. And I'm assuming it's stereo because there's also sync, which I think might be like left, right sync and clock. So like, you know, hopefully I'll get a chip, maybe I'll find a chip that can do the PCM into audio and hopefully I'll be able to figure out a way to make it do audio calls. If not, it's still worth doing because I think for IoT projects, the whole point was that it's a low-cost way of doing MQTT or HTTPS requests over cellular network. But I do, I do really love audio stuff. So the only thing I have to do is, one of the things I have to do is follow this hardware design document to add all and implement all the little things that they're like, oh, you need to do this in order to implement the like power key. So this is the reference set up for each input. And that leads us into the great search. The great search brought to you by Digikey. This is the time of the week, every single week the lady at user powered of engineering to help you, yes you, with your Vision Pro meat edition to find the things on the digikey.com site, lady. And what are you looking for? The Vision Pro just had like the Digikey site, like all the, no, I don't know. All right, so one of the projects I'm working on right now is I want to make a board using a LTE module SIM7080 which is an NBIOT slash LTE cat M cellular module to replace the old 2G and 3G phono boards because these are way past their expiration date and during the part shortage I can get them and then of course now they're totally discontinued. So moving to a new module, you know, I'm familiar with this family of modules and so a lot of the stuff isn't too new. One of the things I have to do is there is a reference schematic, maybe I'll go down and I'll find it. There's reference schematics for, you know, every implementation of, you know, the hardware. So it's like, oh, here's how you hook up the, you know, external PCM setup. Here's how you can do I squared C, you can do SPI for connecting to like SD card or something. But one thing that is interesting about these modules is they tend to have like, you know, fairly advanced, you know, whatever, not narrow band, but like a net band processor, whatever it is that communicates with the cellular network, that chip usually has an IO voltage of like 2.2 or 2.8 volts. It's not 3.3. So if you want to have it, for example, light up an LED, especially if one that's blue, the 2.8 volts won't be able to do it. Now, if you happen to only want red LEDs, like maybe it would kind of work, but the chip is not intended to have high output, high pin strength outputs. And so in order to turn on this LED for like network status, they're like, hey, you know, here's really how you should do it. You should have this transistor, which is biased and, you know, when this turns on, this transistor sinks current, it turns on the LED powered from VBAT, which is a 3.8 volt or 3.4 volt, you know, battery input. Why don't they use N-channel FETs? You know, I'm assuming it's because maybe the IO voltage could be even lower sometimes. Maybe it's like 1.8 volts. And so like, we want to make sure that you can turn on a transistor. Transistors turn on at 0.7 volts in some FETs. Maybe they don't turn on to like 1.5. I don't know really why. This is the reference circuit. We follow the reference circuit because it's probably for a good reason. So the only thing that's annoying is it's like, okay, well, I want to turn on the LED and like before you know it, now I need one, two, three extra components in addition to the resistor and the LED itself. And I want to keep this board small. And one thing I learned is there is such a thing as a pre-biased BJT transistor, which means that you get one three-pin component that has all of these parts built in together. So very handy, especially if you're, well, you're doing cellular modules where you want this reference circuit. But really anytime you're like, oh man, I have like a large number of transistors. I need like, by the way, six of these pre-biased transistors because there's one for each LED and then there's one for like another input or an output or the power key. Like there's a bunch of places that they use these transistors that are pre-biased in the reference circuit. So, you know, I don't mind picking up maybe an array later, but I definitely singled on ones of these to save me from having to also have all the resistors just like less components to pick. And also I don't have to worry about the components being wrong because it's like, it's built into the transistor itself. So let's look for transistors and then we just click on, yeah, here you go. So there's single bipolar transistors or standard MMBT222, PN2907, whatever, but they have a whole category just for pre-biased bipolar transistors. So I don't know where else these are used. I mean, I've definitely like, know that they're used a lot but I've only really seen them referenced in the circuit, the data sheets, the hardware design data sheets for cellular modules. So let's check it out. First off, we're only gonna go for active. We also don't need a diode. I'm assuming a diode is maybe a flyback diode. So if you're using it to like power a motor or something, but we just want it pre-biased, NPN. Let's verify that, hold on. Okay, I clicked away, hold on, it was net light. There you go. Yeah, so NPN transistor and then let's make sure that this is surface mount. I don't wanna through a whole version in stock and I'm gonna exclude marketplace for now just to make it super easy. All right, so the next thing is you get to pick the collector emitter breakdown. So all of these transistors are gonna run at no higher than like four or five volts. So I don't care about this because all of them are higher than my maximum. Next up, I will pick the resistor base and the resistor gain. So this is the base. So the base resistor is 4.7K. So let's go here, base resistor 4.7K and then base emitter is 47K. So it gives us a gain of 10, right? Maybe not, no, I don't remember. Okay, 47 and then apply. Okay, cool, great. So now we've got 51 options. So it looks like there's quite a few families. So the only thing I think I would watch for is, hopefully they all have the same pinouts that they're like cross compatible. Make sure that that's true before you like start secondary sourcing because I'm not convinced that these are all gonna be the same pinout. Nothing else here is that important. I mean, I needed to be able to handle like 10 milliamps but all of them handle 10 milliamps. These are the gain but the gain, the gain isn't relevant for, and these are the saturation also not relevant. Okay, so let's look at pricing. So let's say I'm gonna get 1,000 a piece. It looks like the MMUN 2223 is kind of the cheapest but there's actually quite a few which is nice. One thing I'll notice is that there's a few in this family, LT1G and T1G. You don't know the difference between those. Probably just packaging though. And then yeah, so this is making sure this is set up the right way. Looks good and looks like it hasn't, they actually have it in as little as a SOT 1123. So you can get them quite tiny. I might actually, you know, what's funny is that normally I would say go for a SOT 23 but if they have it in SC70, let's see what packages they've got. They've got SC70, SC75, that's teeny. So let's maybe look at one of these. I think the rest of these are a little bit too small. SOT 70 is a little bit small in the SOT 23. SC75 looks okay but these are, I think these are a little too tiny for me to deal with. So let's look at these. Okay, cool. Yeah, we've got the SMUN 5233. Also got this DTC, same family, just smaller package size. All easily, yeah, handled 100 milliamps easily. So yeah, these are some good options. So I think, you know, if I'm going with a SOT 23, oh, it looks like the naming determines the package size. So yeah, the SMUN is the SOT 323 SC70. The DTC is the name for the SC75, but they're all the same group. So yeah, either one of these would be good. Check this one out. This should save me a lot of space because I'm going to need like six or seven of these. So I'm going to pick these and make sure that I get them laid out on the board. Maybe by next week I'll show you the board layout for my new cellular module. And that's the great search. All right, thanks for joining us this week, everybody. We'll see everybody throughout the week. We've got a bunch of shows and more. Thanks for spending your Sunday night with us. We will see you next week and all throughout the week. We've got shows just about every single week. Check the socials, check our website. And one last note, if you're a native box subscriber, we're shipping again. So log in and check your payment information and address. Don't, don't, it's been a long, it's been a long time. It's been a long time. It's been a long time. It's not a super secret, but still. Yeah, keep it, keep it secret because there's some young folks out there that this is a surprise for them. So that is this lady. Thanks, everybody.