June 27, 2023
Raspberry Pi Pico Schematic in 20 minutes using Flux Copilot
To put this new feature to the test, we revisited a recent project where we designed a Raspberry-Pi-Pico-like schematic using only AI. Initially, it took us 30 minutes to complete the design using Flux Copilot's advice. With Copilot taking the lead on making connections, the same schematic design was completed in just 20 minutes!
In this tutorial, we'll walk you through the important workflows and example prompts to help you design a Raspberry-Pi-Pico-like board in 20 minutes. Curious about the end result? Take a look at the finished project here.
Using generative AI for PCB design means working at the very edge of current possibilities. This requires a few workflow adjustments to get the most out of it. These are some of the principles we found work particularly well, but we can't wait to hear what you come up with. Let us know in Slack!
Your input and interactions play a significant role in getting the most out of Copilot. Be clear and precise with your objectives, don't hesitate to explore different suggestions, and feel free to give feedback with a thumbs up. We compiled a list of useful prompts if you want to learn more about what Copilot can do.
Go from general to specific. It's much easier to get precise responses when Copilot is provided with more details. You can get those details by first asking more broad questions and use Copilot's response to ask a more detailed question. When you're confident about Copilot's reponse, use the "Take action"button to have Copilot wire the schematics for you.
Use Copilot as a teammate. Depending on the question, it might take Copilot a few seconds to reply. Use that time to focus on other areas of the design, Copilot can answer several questions in parallel.
Without further ado, let's dive into the prompts that made this project happen:.
At the start of your project, you'll need to identify the necessary components. This is a good opportunity to ask more general questions at first and start narrowing down the scope of the project.
For a Raspberry Pi Pico-like board, your interaction with Copilot might look like this:
@copilot what is the minimum set of components I need to make a Raspberry Pi Pico-like board?
Having to sift through datasheets consumes a significant amount of time.
When you find a part in the library that you think might for your design, ask Copilot to verify:
@copilot does this LDO work for powering the RP2040?
If you're not familiar with some of the ICs in your design, you can ask Copilot what other components are required or what a pin is used for:
@copilot how should I connect the shield pin?@copilot what else do I need for the oscillator?@copilot what decoupling capacitors do I need for the RP2040?
Copilot can do the wiring for you and save you precious time. Before you ask Copilot to take action, make sure you provide it with as much information as you can. If a component can be connected in multiple ways (as a flash memory to the RP2040), state your goals clearly.
@copilot how do I connect this flash memory to the RP2040? I want to make sure I can boot from it.
If you take a close look at the final project, you'll notice that these patterns repeat over and over on the different elements. Once you get familiar with this workflow, you'll be designing at a speed only big teams were able to accomplish in the past.
It's an exciting time for PCB design. With this new feature of Flux Copilot, we're not just accelerating the design process but also opening up new possibilities for creativity and innovation. It's like having a skilled coworker by your side, ready to tackle the complex parts so you can focus on the big picture.
We're eager to see what amazing designs you'll create with Flux Copilot. This is just the beginning of a fascinating journey toward a future where AI assists in creating, imagining, and realizing incredible hardware designs. And we're thrilled to be on this journey with you.
Happy designing!
A practical guide to calculating PCB trace resistance, covering the core formula, how geometry affects resistance, worked examples, and design tips to minimize voltage drop and heat.
A practical guide to diagnosing and fixing PCB failures, covering common symptoms, a step-by-step debugging workflow, essential tools (multimeter, oscilloscope, logic analyzer, thermal camera), a pre-power-up checklist, and the design mistakes that most often lead to broken boards.
A practical guide to PCB impedance control, covering why it matters for signal integrity, the four physical variables that shape trace impedance, and how to enforce impedance targets from stackup planning through routing and fabrication.
A practical guide to reducing EMI in PCB design through grounding, return path control, shielding, and layout best practices. Covers EMC compliance with CISPR 32 and FCC Part 15.
A practical guide to PCB grounding techniques — ground planes, return paths, star grounding, and analog/digital partitioning — with best practices for reducing noise and improving signal stability.
A step-by-step guide to designing accurate PCB footprints — covering pads, silkscreen, courtyards, IPC-7351 density levels, origin setup, and common mistakes to avoid.
A practical guide to designing multilayer PCB stackups for signal integrity, EMI control, and stable power delivery. Covers layer types, controlled impedance, common mistakes, and how modern tools simplify the process.
A look at how AI is reshaping PCB design by automating routing, placement, and signal integrity checks so engineers can focus on architecture and higher-level decisions.
