Welcome back to the factory. We've got a new prototype to show you today. We'll talk a little bit about BOM optimization and of course it wouldn't be a factory episode if we didn't talk about some jigging as well. Let's get started.

Last week on social media we gave you a sneak peek of what Peter's been working on and I can show you a prototype today. This is the PiicoDev OLED display module and I've had this running overnight. You can see the uptime counter is steadily ticking away just as a little burn-in test for this prototype so it's quite pleasing to see that everything's going well on that side of town.

Just like every PiicoDev module, two PiicoDev connectors so you can daisy-chain through it and we've been testing this with a half meter lead just to see how it goes hanging off the end of a pretty long I2C bus. No worries.

Taking a look at the back of the display board there's a lot of passives hanging off this thing and they're all really just analog supporting components for the driver IC to drive the display. Now when you look at typical applications of these there's a lot of different implementations. You get quite a widespread of capacitances chosen and often the implementation uses a lot of unique components like there might be two or even three unique capacitor values on that board and this was a little bit of an experiment.

We tried to optimize the BOM a little bit just to see if we could and it looks like the experiments paid off. I'm really pleased to see this because if you look closely every capacitor here is a 2.2 microfarad cap and there's even two in parallel here to make about 4.7 and there's only two unique resistors. Everything's a 4k7 except for this current reference which you can't really get away with. From what I can see, it looks like this experiment has been successful. When designing a product, it's important to consider the availability and management of each component. By removing just one part, we can save ourselves the trouble of setting up systems to handle that specific component.

The assembly process for this board will be similar to our other boards. We will panelise it and send it through the pick and place machine to place all the necessary parts. However, we won't include the OLED module in this step. The OLED module cannot withstand the heat of the reflow oven, so we will have to add it later. We have a few ideas on how to do this.

If you take a closer look at the ribbon, you'll notice two holes. We have included two non-plated through holes on the circuit board to align with these holes. Our plan is to design a bare FR4 PCB with two fine pitch pins that can fit through the ribbon and into the circuit board. We will also include some kind of tongue to hold the ribbon down. This way, we can use an alignment jig to hold the pins in place while we solder them onto the ribbon. We are currently waiting for special soldering tips for soldering flat flex cables to arrive.

Last week, we were working on ideas for a programming and test jig, and it has evolved into what you see now. The basic concept is the same as last week, with two PiicoDev connectors to test both connectors simultaneously. We have also included a hole for a pogo pin, which will be used for programmable modules to program into a UPDI pin.

I hope this gives you a better understanding of our process and plans. If you have any further questions, feel free to ask. With a motion sensor board because I don't have a programmable board assembled. Load the board in connect it with the left connector that bottoms out against some stops and then slide the carriage in and now you've got two independent connections so we can test these two connectors with separate I2C buses and this is being held against a pogo pin coming in the back which will do the programming.

I mentioned that I might have a bar to hold this carriage down you can see that's not here. On one of the internal layers the carriage has two like ears that stick out that run inside slots on an inside layer so this is now completely retained you can't pull it off the end and it won't fall out of the jig.

This is an earlier version where I was experimenting with that idea of having a glued carriage and I just found that was way too rigid and just left no room for error. When you glue up this part if the connector is just out of alignment a tiny bit you have no hope of correcting it the thing just doesn't work. You can see here if I were to bring that carriage in that's being glued a little bit too low and so this part is just completely useless.

By going with a design that uses fasteners there's a little bit of play so you can massage things into place tighten up the faster and the jig is now tuned. It also makes it probably a little bit easier to replace parts you just undo the fasteners put the new part in you're good to go. So rather than replace the whole carriage assembly like in a glued model I would just undo this plate and replace the cable.

To eject the device under test you just pull that carriage away and some stops catch the corners of the device under test and help pull it off that connector just for a little.Bit of a quality of life enhancement and if I tune these stops on the left hand side of the jig to basically limit how far this left connector can engage then that means that the right connector may have a stronger holding force so that when you slide this carriage away it's more likely to detach from the fixed cable and then self-eject off the carriage using those end stops from before.

Now I designed this jig using illustrator for no other reason than I'm pretty comfortable using it. You know you can quickly make a single panel graphic design that you can laser cut out very quickly with more complex designs like this that have many layers that interact with each other and sliding fits it would probably make more sense to use a a dedicated like 3d CAD package. I know there are CAD packages out there that are good at generating files for the laser cutter I just haven't taken the plunge. If there are any laser cutting aficionados watching let me know your thoughts I'd love to I'd love to use something a little more parametric for projects like this.

I suppose Fusion 360 could work but if I was going to go with Autodesk we'd be using Eagle instead of KiCAD. Let's keep it open source. Looking forward to using this I'll probably grow the bottom layer so it can fixture onto a programming and test PCB and we're not too far from that because the prototype RGB LED module has arrived so we'll be able to prototype these in parallel.

In any case that's all I have for you this week let us know your recommendations for free laser cutting CAD software and until next time thanks for watching.