Welcome back to The Factory. This week, we have the Makerverse SMD Prototyping Board, now available in gold. We also have an aesthetic upgrade for Glowbit. Let's get started.

Last week, we discussed the Makerverse Prototyping Board for surface mount components, specifically the SOIC Prototyping Board. However, the initial prototype did not provide a satisfying experience. The hot air leveled solder finish made it too easy to create unwanted bridges. We were pushing the limits of the technology and the clearances involved. So, we decided to try a different approach and opted for a gold finish. Here is the revised prototype, now featuring an Enig Gold finish. It looks incredibly shiny and the user experience has greatly improved with this surface finish. In the test footage, you can see us soldering a SOT23 transistor, a SOIC component, an 0603 capacitor, and a big 5050 LED. We also recommend adding some solder to one of the through holes to ensure there are no unintended connections on the rear side ground plane. The clearance is appropriate, as the solder domes nicely on the pad without spreading onto the ground plane.

Currently, Brenton is using a microscope to assemble a high-speed analog circuit so we can thoroughly test the board's capabilities. Let's see what we can discover. On our SOIC breadboard, we have an inverting op-amp circuit with a gain of 10. On one side, we have a signal generator, and on the other side, we have an oscilloscope monitoring the input and output of our circuit.Amplifier. On the left side here, we just have our power input. It's a standard bipolar power supply, ground minus V and plus V, where the positive power supply comes in on this wire, the negative on this wire on the back here. And for the ground connection, we are using the solid ground on the back of the SOIC breadboard.

Our signal is coming from a signal generator here. It's a hundred millivolts peak to peak at the moment at about 400 kilohertz. So on the oscilloscope, we've got our input signal in blue and the output signal in yellow. And at the moment, the channels just have the same gain. If we crank up the scale here, there we go. We can see that there's a gain of 10 where the signals are equal and we have a factor of 10 difference in these channel sensitivities. So this is only at 400 kilohertz at reasonably low frequency. So the really nice thing about building these sorts of high frequency analog circuits on a SOIC breadboard like this is that they still behave quite nicely at high frequency. This particular circuit with this particular amplifier in theory has what's called a gain bandwidth product of about 12 megahertz, which means that at a gain of 10, we end up with a bandwidth in theory of about one megahertz.

So we'll turn up the frequency and see when the output signal starts to drop when our gain starts to die off. So if we slowly turn up our frequency, we're just looking at the yellow trace here. If it's gonna start to drop and there it starts to drop. So at 1.6 megahertz, it's basically the same amplitude. And by the time we get to about here, about 2.1 megahertz, it's dropped a little bit. At about 2.4 megahertz, this is actuallyWhat's called our minus three dB point where the output is dropped by a factor of 0.7. The chip is overperforming from the spec in the datasheet basically. The datasheet has a gain bandwidth spec of 12 megahertz, which would be a minimum guaranteed. And this particular chip happens to have almost double that and we're realizing that performance.

So now that we've measured the performance of our op-amp circuit on the solid breadboard, we're gonna simulate having built this circuit on a breadboard and then measure the bandwidth of the circuit again. Basically, by simulating it, what I mean is we're gonna plug a couple of points in our circuit into a breadboard. And what you do there is you introduce a capacitance between two breadboard rows. And that capacitance can actually be quite significant when you were talking about high frequencies and so forth.

Well, the main thing is that we put some capacitance on the feedback resistor because that's when it's most crucial. I've just gone to the soldering bench and I've soldered on a couple of pins, just a standard pin header so we can plug the SOIC circuit into a breadboard. And straight away, even before we've plugged it into the breadboard, we already have a small decrease in the output amplitude. This has gone down from about 700 millivolts to about 580 or so.

So we'll plug this into the breadboard and we'll straight away watch the output drop. There we go. All I've done is plugged those two pins into a breadboard. Those two pins are across the feedback resistor. So that's introduced somewhere around 10 to 15 picofarads of capacitance in parallel with the feedback resistor. So we can now wind down. The frequency and try to find where that minus three dB point has gone. I'm just winding down the frequency and watching the peak-to-peak voltage drop. It needs to be around 700, needs to be around 700 millivolts. Going the wrong way. But there, you see it there at about 700 there. Now that's at 580 kilohertz.

So the bandwidth of our circuit has dropped from 2.4 meg down to 580 kilohertz just by building it on a breadboard instead of a specialist high-frequency prototyping board. We've talked about the new Glowbit modules a fair bit recently and you might be wondering why we're not just getting on and making the thing. Like, come on, it's just a bunch of LEDs on a board. What's the holdup?

Well, we wanted to make Glowbit as beautiful as possible for creative makers and we just couldn't say no to the opportunity of getting in some black Glowbit LEDs. Since we started up Glowbit about a year ago, every time I've been interacting with the manufacturer, I've been like, hey, are the black LEDs ready yet? Are the black LEDs ready yet? And I'm happy to say that we now have a whole box of these rolls of Glowbit LEDs now in black. Let's take a look. I haven't opened one of these yet. This is the best bit.

Okay. May not seem like much, but that is going to really drastically reduce the visual impact of the white squares on these Glowbit modules. Oh, it's a good thing we've got 90,000 more.

Same LED technology, same WS2812 version five technology, but now with a black epoxy package instead of a white package. And so now the visual impact of each pixel, instead of being a white square, you'll still have the white of the dye material, but it's just going to be aBit more regular looking, a bit more omnidirectional looking rather than these big squares. And that, I think, is gonna make a nice difference. So that little change is all that we've been waiting for to begin manufacturing the new Glowbit modules.

Keep an eye out on the socials for when we assemble another prototype now in black. Yeah, just think about how good something like this is gonna look with a black package LED instead of these white squares. On the production side of town, the production line is commissioned. The stencil machine, we've figured out a good workflow for the stencilling machine. The oven is commissioned with a useful reflow profile. The oilers are oiled. The, everything's ready to go. And you know, it's just as well because we have a few designs yet to be manufactured that are backing up.

In addition to our regular production schedule, making existing Core Electronics originals, we also have in the lineup, we need to make the first run of the PiicoDev capacitive touch sensor, the PiicoDev colour sensor, the new OLED module, and the RGB LED module. That's also in addition to all the Makerverse hardware that we've been talking about for the past couple of weeks, but the panels for those should be arriving today. You might've noticed this stack of boxes behind me. For our new automatic stencilling machine, we needed to upgrade our stencils to use Fiducials. This is what one of our production stencils looks like. And the reason for the upgrade is a detail on the back. This tiny little, it looks like it's a laser engraved dot, is the only upgrade that we needed to make.

Our automatic stencil machine has an up camera and a down camera.The down camera looks for the three Fiducials that are on the rail of our PCB panels, and the up camera looks for that same Fiducial, but on the stencil. So when you first load a stencil, the camera flies around, looks for the three Fiducials on both the panel and the stencil, and then it'll move the panel around to perfectly match the position of the stencil, so that all the pads line up perfectly with all the apertures.

Now that there's nothing left holding our production, we'd best get to it. Thanks for joining me on this week's episode of The Factory. If there's anything you wanna see a little bit closer, or if you just have some questions, leave a comment below or open a thread on the Core Electronics forums. Until next time, thanks for watching.