Well, we're live! We had a bit of a hiccup at the start, but hello! Hopefully, this is coming through fine now. I'm just making sure we've got our comments connected properly. There we go, kicked off the YouTube ones. Not connecting, of course, right at the time of a live stream. It says, "Live chat you were trying to enter cannot be found." Oh, interesting. If it was easy, everybody would be doing it. Just going to have a quick look at these configs for the chat. There we go. Alright, should be coming through on YouTube. There we go. Thanks, Seon. Video and audio are good. Great. Thanks for the heads up. I'll move over a little bit, give Michael a bit of room. It's been a while since our last live episode, hasn't it? Well, this is my first one. Yeah, it is. My fourth. But we basically geared up for live earlier in the year when we were getting ready for these moves with some internal projects around production and some changes with the way we're doing things. Great preparation, but it's been a few months since we've been live. Sorry about the delay there. This is pretty exciting for us. We've been doing a fair amount of work over the last six months, 12 months. Michael's back from his trip around the world. So if you are not familiar with Michael, he's been with us for how long now? A few years. Like early 2017, I think I joined. Yep. Had a bit of a hiatus in the middle there and has come back in to work on some projects. We get along great. It's awesome.

This episode is a bit of an introduction to what's changed since people sort of know who we are, what we're up to. A few things have changed in that time. So we'll kick off with a new segment on our live stream, and that is the factory. The front of our building has been completely transformed into a big workspace. It's more of like a workshop environment. By the way, we're seeing comments that come through at the same time. So if you have any questions along the way, I'll check in every now and then and happy to answer them on the fly. The factory itself, and I'll just switch over to this view of what we've been up to. Let me get the picture up. This is classic. We were setting up the buttons to switch through the pictures nice and quickly, and they didn't work last minute. So instead, we'll bring up our pictures manually. If you've ever picked up an order from the Core Electronics warehouse and maybe stuck your head around the corner, this might be quite different from what you remember. So this is the new workspace, the factory, and it's in the screenshot that was there a moment ago. You saw our support team, and we'll come back to that. But this is the new workspace. We have our pick-and-place SMT line. We have our 3D printing farms, a few others, our Ultimaker material station, basically the full system for the Ultimaker, and the Trotec Speedy 400. Love that machine. Love all the machines. But particularly the SMT line, that's what's new, and it's a big deal for us.

Here at Core Electronics, there's actually 19 of us in the team now, and while we're not all full-time, a good portion of us are. There's a lot more knowledge and creativity in the team to fill gaps, to solve problems, and also to add some resilience to our business. Coronavirus, everybody gets it. It's been a hard time with logistics, with supply, with manufacturing, and it was great timing for us to get in front of this and explore manufacturing here in Australia, in Newcastle, and also the knowledge that comes from that. Both Michael and I and Co. are working on these projects, upskilling ourselves with this process that is manufacturing, and it's just been such an excellent experience to have. There's not a lot of guidance in how to bootstrap electronics manufacturing, so we are just learning all the time. So much experimenting. Well, let me see if I can get these socials working. There's a couple of comments here. Hello, Dave Hines from YouTube. Thanks for dropping in. And we have Colin as well. Awesome. And of course, already, thank you, Sion, for letting us know that the video and audio are all good. I appreciate it. Yeah, it's great to have somebody out there that can just give us that direct feedback.

This is quite new, and for anyone that's dropped in to pick up orders from us, that's seen how to walk around, because I invite people to come down to do a tour, or I've done a workshop here. We've had over 2,000 Newcastle citizens join in on a workshop, learn something new, learn how to program a microcontroller or how to laser cut or how to 3D print and 3D model. I don't want you to think that the support team is just gone, because yes, of course, back in the day of this video, which is the Meet Our Support Team video, this opportunistic shot here will actually show that. Let's just go over to this picture. Oh, wrong monitor. There we go. This is what it is now, and the support team is just down the other side of the building. We're in the middle of a transformation, so we have our support office, our creative team, our product studio, and, of course, way down the back there, you can just see the Make a Verse sign on the door. That's our main workshop area, and further up this corridor, there's another two offices where IT and accounts are. We're definitely at the brim here. Every single square meter of this whole building is in use. In preparation for setting up the factory, it's like the whole building was tipped upside down. Everybody was shook out and then rearranged so that we could set up this space. Oh, wow. We've got some international hellos. Hello from Canada, from Live Art, and we have a – oh, thanks, Cathy, for retweeting the feed. Really appreciate it.

With that, obviously, we've got this new machine. This new machine is a Neoden K1830. That's the model number of that machine. It's got some nice specs. Up here on the gantry – I'm not sure if I can zoom in easily with this viewer, but up here on the top, we have the gantry. That's got eight nozzles, and so those eight independent nozzles can move across, pick up components in simultaneous pick, which I've had some issues with. It's better just to pick them up individually really quickly, like within a few milliseconds, rather than a simultaneous drop and pick, but it's still very quick. They claim 16,000 component placements per hour. I feel like for a reliable process, it's closer to 4,000 to 8,000, depending on how well you can optimize those pick and placements, and it's a big deal for us because it means we can manufacture printed circuit boards, assemble components on the printed circuit boards here in Newcastle relatively quickly. We have 66 feeder locations, so there are 33 per side. Now, in the ideal world, you'll have 66 components that you can fit to this that are all completely unique and that adds a lot of diversity to your options, but you end up with systems like this, the vibration feeder, which takes up eight slots and only allows five tubes to be fitted. We've also got some wider feeders here that take up two spots. Well, it's not the full two spots. It still means one spot is, I guess, impeded upon, but 66 is more than enough for us to get a lot of work done right here. At the moment, this is a little bit in the past, but there were, at one point, four unique projects loaded into all those slots on the machine. There's still probably room for another four projects, so you can fit, as a working volume, quite a number of projects into the pick-and-place machine. We have two vision systems, which is great because it means, as the components are picked up, it can use either camera to more, it's just more optimal to do the pick on one side, the vision alignment of those components, making the little adjustments for the X and Y rotation, and then it zips over and does a placement.

It's quite amazing to witness those nozzles picking up components. Because they all occupy a small window in the tray, they can shuffle around and get bumped around a small amount. As the nozzle picks it up, it quickly swipes over the camera, and every component gets a picture. The alignment system can then correct for any small deviations in translation or rotation. That's what we've been up to with the factory transformation. We still intend on holding workshops in that space, though with fewer numbers—probably four to six instead of six to eight. We'll resume those once the dust settles with the current climate. We're very excited to continue and have just pressed pause due to current conditions. Moving on, we have a question from a YouTube user asking if we have any videos of the machine in action. Absolutely, you'll be seeing plenty of those. We have lots of content coming up around this process and will share our intentions for the new segment titled "The Factory." Core Electronics Live, the factory, will be a part of this segment.

We also have a question from Tim in the chat about experimenting with different things for reflow soldering. You might have noticed our hotplate video, which we use mostly for quick in-house prototyping. Our production line has a reflow oven, but it takes quite a process to get up to temperature. That hotplate is really for quick, on-the-bench scale projects. Let me show you a quick intro; the audio won't be great, but it's just the first few seconds of what will be a recurring weekly series to share the whole process of what we're up to. We want to share the way we're going about the manufacturing process. This is just one of the episodes. Today, we're applying solder paste to a PCB panel using a solder paste stencil. If none of that means anything to you, we have a video that introduces the PCB manufacturing process. There's some video of it in action, and there will be plenty more coming up. We'll share that process with you, and any questions along the way, please ask. We're happy to explore the process from your perspective. We're doing so much experimenting and learning that we aim to capture as much as we can for others who pick up a machine like this.

Sion from Unexpected Maker says it's incredible how fast it photographs every component on the eight nozzles. For those new to live streaming, Sion is a seasoned live streamer for Maker hardware and the Maker spirit. He is the founder and inventor of TinyPICO and other products. Go follow Unexpected Maker if you love live content. He's also got a Neoden K1830, and we've been helping and sharing ideas, workarounds, and tips along the way. When you get a machine like this, there are going to be problems. Sometimes the problems aren't with the machine but with the interpretation of a workflow. Having someone accessible during the same hours of the day is handy. Thanks, Sion, for your help. I hope we've helped a little bit as well. We'll jump straight into the next segment. PiicoDev, the overarching project, represents three major markets in the Maker community: zero to Maker, Makers, and the industrial side.

The zero to Maker segment includes people just getting started, ignited by an idea. We've been at this for 14 years and have seen all sorts of entry points to the Maker industry. Classically, it's a project that inspires you to get started and explore how it happened. Then there are Makers who know their way around hardware and technology, solving problems that matter to them or inventing cool things. Finally, there's the industrial side, with people building commercial, scalable, affordable sub-assemblies and parts or complete products. The ecosystem is well-segmented. It's unusual for someone in the zero to Maker segment to be using industrial Maker to Market products. We're breaking up our projects and product design with these ideas in mind, starting in the middle where our core audience is. We need to pick a lane and follow it, and the Maker electronics sphere is what I'm most familiar with. We're building an ecosystem that will form a development kit celebrating the Maker journey.

You'll be following this whole process of the experience and the things we're designing. It's all open source, so you can contribute, learn from, reuse, and remix however you like. We'll make them accessible via Core Electronics. If you have any questions along the way, let us know. Let's bring out some hardware. We put it all on the floor to avoid cluttering the bench. Frank asks if the guy in the black shirt is Graham Mitchell. Yes, Frank, I am Graham Mitchell. Thanks for asking. Sion has already answered, and it's been a great help. We've covered a lot of ground, and if we were working with OEM support alone, we'd get there, but not as fast. A system like this is all about the moment, solving something within minutes. Being able to Skype someone and ask a quick question is valuable. Let's go to the top cam. We'll start with Glowbit and the Rainbow. After purchasing the production line, we thought we'd best have some hardware to fire it up with. We landed on this simple output device as a learning experience for using the production line. It's a great hello world project for electronics manufacturing. We have a long list of ideas and needed to pick something that stood out. It's important for us as a commercial business to find creative solutions, particularly for zero to Maker.

This project is designed for use with the micro:bit, and while it's our introductory venture into PCB assembly, we wanted to delve into ideas around version control systems (VCS) and how we manage these projects internally. Choosing simple projects allowed us the time and energy to refine our versioning and collaborative processes. This is our first project, featuring the latest WS2812 version 5 technology. Unlike classic WS2812B LEDs, this version is advanced and clean, requiring no external components. Typically, circuit boards with these LEDs have numerous decoupling capacitors and sometimes resistors for input protection. However, the new LEDs integrate decoupling capacitors, making them unnecessary and extremely robust against reverse polarity and overvoltage. I tested them up to 12 volts, and they handled it effortlessly, dissipating the current as heat. This integration allows for a cleaner design without the clutter of additional components, enhancing the visual appeal of the design.

In this project, we explored connection systems with great interest, leading to the development of an excellent test jig. Quality is paramount to us, and we aim to electrically test everything from our factory to ensure it functions as intended. We'll share our jig design and manufacturing process in upcoming factory episodes. The jig began as an Arduino shield and evolved into a production test jig, simple to use. Each device has test points, and experienced makers can use headers for permanent wiring. The jig uses standoffs and springy pogo pins for connection, effectively serving as an Arduino shield, which is stable and doesn't require fasteners. To test a device, you simply press it down, and a test pattern runs, showing red, green, and blue in sequence. An onboard LED ensures data is correctly outputted for daisy-chaining. Testing is quick; once all LEDs work, you move to the next device. This was one of our early projects, designed before our machine arrived. We ordered the machine in June, expecting it in early August, but it arrived in November. This delay allowed us to work on designs and branding, giving us time to discuss our future projects and strategies.

This LED us to the maker's market, aligning with our experiences. We aim to integrate our latest maker guides and tutorials into our products, focusing on PiicoDev. PiicoDev is a suite of chainable input and output devices using standard headers, designed for beginners to explore different transducers and output devices quickly. It's also meant for professionals to integrate into their designs. We're leveraging open source, using quick connection systems to ensure compatibility with other guides using quick connectors. Our libraries facilitate quick integration into projects, working towards a PiicoDev kit for makers, including our own dev board. We're planning this journey and will share it, featuring upcoming sensors and manufacturing process videos. Our dev boards can produce about 60 units per panel, with some panels taking 12 to 15 minutes to manufacture. This capacity allows us to produce many panels in Newcastle, enhancing our team's skills and knowledge. This project is a long-term design for manufacture, with standard connectors and components on every board, providing insights into DFM and loading projects into the K1830.

Starting Core Electronics in 2007 followed digital DIY in 2003, writing guides for the maker community before major online platforms existed. We created tutorials for using modules and parts that were hard to find at stores like Dick Smith's and Jaycar. Core Electronics began as a maker retail venture, with aspirations for a pick-and-place machine. However, I was in the Air Force, balancing work and helping with projects. By 2016, I couldn't manage both, and now, after leaving the Air Force, joining the team full-time has been rewarding. We're thrilled with our progress and future plans. Frank in the chat asked about programming, specifically MicroPython or Arduino. We plan to stick with MicroPython, a technology-agnostic environment that suits our needs. Damian George from Melbourne, an Aussie and Cambridge graduate, developed MicroPython, creating a great ecosystem for our guides and software support. As modern embedded maker projects evolve, MicroPython remains central to our strategy.

I started off with Pickaxe and then moved on to Arduino. You often feel kind of locked into the language of the technology you're using, and MicroPython is really a game changer for that. It's about how fast you can access it and then iterate. Because it's Python, you can share code on a forum, and it's more interpretable, making it easier to solve issues. This ease of sharing and collaboration is invaluable. When you get down to assembly and similar languages, it becomes much harder to share and collaborate on problems and ideas. MicroPython ticks all those boxes. Now, let's talk about panelization. Electronics aren't produced one board at a time; it's more efficient to produce them on a panel and send the whole panel through the machine as a batch. This is two opposite ends of the spectrum: the rainbow you saw before came directly off one of these panels, with routing around the perimeter and breakaway tabs. On the other end is a higher-density switch mode power supply module. You can see the breakaway tabs in the panel of rainbows, but here we have a single slot routed down the side of the devices, allowing them to be easily snapped apart into single units.

PCB colour is another interesting topic. Consistency is key, and different colours can require different reflow profiles due to how they absorb infrared light. For example, black PCBs might need a different profile than white ones. This difference can impact the reflow process, and choosing a PCB colour is tricky, also limited by the manufacturer. We've experimented with colours, and for our core electronics, blue seems appropriate. For our Zero to Maker range, where aesthetics are important, black and gold might be suitable. The difference between gold and tin finishes is also something to consider. For industrial applications, green is a safe choice due to its ease of visual checking and process compatibility. Green boards are common because they provide good contrast for visual inspections, making them easier to work with. As for manufacturing custom user design boards or open-source boards, everything from us will be open source. We'll have a dedicated repository for projects where you can contribute feedback or gain insights. Currently, we're focusing on our own designs to ensure our process is solid, from parts management to quality control. Manufacturing bespoke parts isn't feasible for us right now, as we're building up our inventory and focusing on make-a-module designs.

Lastly, let's do a brief show-and-tell of one of our earlier designs: the logic-level converters. These are essential, and testing them quickly is crucial. We have a test jig for this, using a Teensy for its abundance of I/O and voltage tolerance. The jig allows us to test a panel of logic-level converters efficiently. If you march down the row of converters, everything works until you reach the sabotaged one, which gives no feedback. This principle ensures we don't send out confusing information; a device is either good or yet-to-be-proven good. This extensive testing process is appreciated by those who delve into the details of test jigs. We've received a question about developing LoRa protocols, which is an interesting area to explore.

No, not yet. We do have MicroGateway, which is a past project for a low-cost, Pi Zero-based MicroGateway. It was quite popular, but LoRa and LoRaStands move very quickly, so our team is best skilled in this domain at the moment. We'll definitely, in our pipeline, consider having a dev board with Wi-Fi and Bluetooth, but maybe not LoRa out of the box. The guys at Pycom do a great job at that, and it's at a price point where you need to offer OEM modules. That's what makes LoRa more appropriate in scale. People using LoRa are often industrial types solving industrial problems. While a dev board solves a problem as a rapid prototype, you often need an OEM module placed straight onto the board, available in trays of 1,000, and that's a big deal. The guys at Pycom have done a great job at solving that problem.

Sion mentioned a fancy camera setup, and yes, the gear acquisition syndrome is real. It's about curiosity and making things better. We're trying to present content in the best way possible with what we have. We've created over 400 guides on Core Electronics and over 300 edited videos on YouTube, with five playlists of full workshop experiences. Our creative team and curiosity around content creation have LED us to this point. Someone from Unexpected Maker commented on the fancy camera setup, which was nice. They also mentioned the challenge of setting up a test jig and doing an exhaustive test program, which takes longer than making a four-channel logic-level converter. It's a satisfying part of the process, as you're looking at all possible scenarios and use cases of how things shouldn't work. You're making a device that tests whether something works, and if it doesn't, your device must be robust enough to not break or break the device you're testing. It's an interesting challenge.

I'm curious if Sion is still on, as I remember seeing a screenshot of TinyPico and the test jig that said it was 40-something seconds to do the test. Is it still 40-something seconds, or do you get the firmware preloaded? It seems like a lot of waiting, so you might need a battery of these jigs. That's another part of it, isn't it? The scale. Let's jump straight into something worth a read. We ended up going to manual mode way at the beginning, if anyone remembers, due to an issue with automating some videos. I'll open up some links. Let's kick off with this one, as it's a big deal for us: Python workshop for beginners. It's 24 videos, and you can see how small the scroll bar is. This isn't geared towards MicroPython or Raspberry Pi, but Python in general. You can use any Python editor and follow this journey. Tim covers setting up a Python editor and following the journey through to working with APIs. We feel it's essential to have some coding experience, and Python is a great way to start. It's a precursor to MicroPython, as most Python code can be ported over with minor changes.

If you have questions during this process, it's supported by us, Core Electronics. You can post a question at the bottom of the guide, and we'll help you through any issues. This is also on YouTube as a playlist, but it's better suited on the platform for copying and pasting code. Sion got back to us, saying the process includes erasing and flashing the MPy before testing, which is slow over serial. The Feather S2 takes 50 seconds with a 16 megaflash. Writing to flash via serial is slow, and parallel might be the solution. Thanks for sharing that, Sion. It's a problem we'll have to deal with someday. Let's jump into the next guide by Michael. Automating test equipment with Python is very on-topic. This tutorial came out of necessity, as I needed to run an automated test to collect data for technical documentation. You can have your whole lab plugged into your computer via USB cables and automate a battery of tests. Your only limit is the test equipment you're using.

Nice. I've got to say, that is one impressive rat's nest of wire. I wonder if it's coming through the... Maybe we can open the image, copy the image location. Look at that. So we went through pretty much every scenario curve with this that was meaningful. Min, middle, max, overnight or multi-day tests just to really make sure that this is an industrial thing. We want to make sure that it complies with what we claim it does. Testing that whole range of operation is a big deal if you're going to put this in a product and rely on it to work. Which is probably not as needed for, say, CloudBit Rainbow. But CloudBit Rainbow is more of a functional test. You're relying on upstream quality and standards. But when you're working with something like this, there are actually lots of choices between the inductor, the capacitors, and you've got the thermal ratings of parts. All the way down to how you draw your power polygons when you're laying up the PCB artwork. It's actually an interesting topic on its own, just that, the thermal specs of the parts. If you buy cheap parts, it will absolutely affect the performance of this module in some way. It will still be functional, but it will compromise some element of its outer limits. Even if all it affects is efficiency, that's such an easy-to-understand metric of how suitable the device is for your needs, particularly if you're doing something battery-powered.

Nice. And we'll head over to Ask a Maker. I was just going through the forum a moment ago seeing what unanswered questions there are. Great to see a vibrant community, and everything's answered except for one. It's quite hard to find something unanswered. So Ross has asked on our forum that they've bought items to build an Octoprint from us. They have read something on the Creality forums that the 5-volt wire on the USB cable might need a USB isolator, like in the link below. So I'd clicked on the link and had a quick read of this. So this product looks like an inline USB system. Looks like three of the wires are connected, and one is not. And indeed, even though it's called Power Bluffer... Bluffer? Bluffer. A play on words there. Why is it useful? Because they mentioned that some 3D printers... I haven't seen one that runs into this issue personally, but some 3D printers, if you power off the printer but the Raspberry Pi remains powered, it may power the 5-volt system on the printer, which may illuminate the LCD backlight or something. So it's more of an aesthetics issue. Doesn't sound like you're going to break anything, but it sounds like there's a bit of an aesthetics issue there. So I'm assuming that's where Ross is coming from, that the printer that he's using has this symptom.

And is it necessary? Well, I don't feel like it's necessary, but this makes a lot of sense. This sounds very functional. This solution is to simply leave the two data lines and the ground line connected. Ground is essential so that you have a common earth between these two devices. It'll make data more reliable, even though it's all coming from the same PowerPoint in most cases. It's great to have that in any communication system, unless it's like RS-485 or something, I guess. But in this case, probably needed. And I guess if you've got it disconnected, the voltage on the USB, then there's nothing to go back and power the 5V bus if for some reason the printer design has that connected. We use OctoPrint on our LulzBots and haven't seen that issue. But yeah, I'd be curious to know what Creality, we've got some Creality printers. If it is something that we stock, I'd love to know if it's an issue and if this did solve it because it's always great to add some content to projects like this. Just another thing in the bag of tricks. And Sion's chimed in saying that it's pronounced blocker. Blocker. Gotcha. Nice. Yes, a play on words. Great. Thanks, Sion. Appreciate it.

There was one more. It was more of just a general product question, but somebody's asked a question about this device, an MPLAB-compatible USB PIC programmer. If you ask a question at the bottom of one of our product pages, a helpful bot called Core, and with a bot logo, will provision a forum topic and then post the question underneath into our forum so that it's an opportunity for us to answer or other people to answer. In this case, Martin has asked, who owns one of these themselves? I don't have this specific model, Martin, but I have used a range of programmers that are MPLAB-compatible, particularly Picket. Not Picket 4, but definitely Picket 2, Picket 3. But if you have any direct questions about functionality or workflow, please add to your comment, and happy to help. And I'd just like to give a quick shout-out as well to Robert. So Robert has well over 20 years of experience in engineering and has just been such a champion helping out on the forum. So thanks, Robert. Really appreciate it. It's great to see the Maker community come together, work through issues and solve problems and just build stuff. It's very, very refreshing. Appreciate you, Rob.

Well, that's... We did have other stuff, but coming up on the hour. So it's Saturday, it's 11 a.m. here in Newcastle anyway. Covered a lot of ground. You know, we're really excited. I don't want to take away from the upcoming episodes of The Factory, which we shared that preview for earlier in the video. Michael's content will come out weekly, and the segment will be called The Factory, so watch out for that on YouTube. It's quite different than the classic Maker education guides, which will still be produced on a regular basis. And we're just going to share it. We're going to show you the whole process and we'd love for you to just be a part of it. So any questions along the way, reach out, ask. Drop in for a workshop when they're back on again. And likewise, if you see something where it's like, hang on, I know what's going on here. I can help out with this. Absolutely get in touch with us. Oh, please do. Yeah, that'd be amazing. That's really where open source begins to champion is that community approach to not just, you know, for interest or for fun projects, but commercial endeavours as well. It's great to see the community come together this way.

Well, that about wraps it up. And what do you reckon, Michael? We've got a couple of little things, but it is the end of the day. So we'll call it quits there. Thanks to everybody that... Oh, Luke. Got to get him in on the mix. Luke is the reason why the production quality in Core Electronics videos has jumped so high in the past year or so. Luke edits all of the videos that you see. So thanks, Luke. Thanks, Luke. Appreciate you. It worked. The live stream worked. I hope it worked anyway. We'll go back and watch it afterwards. See what all the glitches were. But, yeah, that's a wrap, guys. And thanks for dropping in today. Until next time.