I am working on (another yet) microcontroller expansion card based on RP2350A to provide GPIO for Framework PCs. I have seen others created similar cards based on RP2040 and this chip, and this one is my take. I will attend to Maker Faire Rome with this one and some other cards I will make. Here is the GitHub repo. Its appreciated if you can take a look and make suggestions.
13 Likes
How is deepComputing related to this?
1 Like
They will sponsor our booth and we will make RISC-V based expansion cards.
2 Likes
This is awesome- glad to see something new microcontroller wise with so many rp2040 card designs out there
One suggestion is add some onboard indicators/actuators like RGB LEDs, tiny piezo, haptic motor, etc and maybe a hotswap socket on top for a standard cherry switch
1 Like
All board errata are fixed (no violations on DRC anymore) and its ready for production.
3 Likes
Ok… me wanty, very much me wanty, that’s awesome.
Any chance of an update on the readme.md to also show the backside? (rp2350gpiocard-back.png)
Also any chance of a pin table in that readme?
Guessing its something like:
Front:
| Label |
Name |
SPI |
UART |
I2C |
ADC |
| 0 |
GPIO0 |
SPI0 RX |
UART0 TX |
I2C0 SDA |
|
| 1 |
GPIO1 |
SPI0 CSn |
UART0 RX |
I2C0 SCL |
|
| 2 |
GPIO2 |
SPI0 SCK |
|
I2C1 SDA |
|
| 3 |
GPIO3 |
SPI0 TX |
|
I2C1 SCL |
|
| 4 |
GPIO4 |
SPI0 RX |
UART1 TX |
I2C0 SDA |
|
| 5 |
GPIO5 |
SPI0 CSn |
UART1 RX |
I2C0 SCL |
|
| 6 |
GPIO6 |
SPI0 SCK |
|
I2C1 SDA |
|
| 7 |
GPIO7 |
SPI0 TX |
|
I2C1 SCL |
|
| VCC |
3.3v or 5v? |
|
|
|
|
| R |
Reset |
|
|
|
|
Back:
| Label |
Name |
SPI |
UART |
I2C |
ADC |
| 29 |
GPIO29 |
|
|
|
ADC3 |
| 28 |
GPIO28 |
|
|
|
ADC2 |
| 27 |
GPIO27 |
|
|
I2C1 SDL |
ADC1 |
| G |
Ground |
|
|
|
|
| 13 |
GPIO13 |
SPI1 CSN |
UART0 RX |
I2C0 SCL |
|
| 12 |
GPIO12 |
SPI1 RX |
UART0 TX |
I2C0 SDA |
|
| 11 |
GPIO11 |
SPI1 TX |
|
I2C1 SCL |
|
| 10 |
GPIO10 |
SPI1 SCK |
|
I2C1 SDA |
|
| 9 |
GPIO9 |
SPI1 CSn |
UART1 RX |
I2C0 SCL |
|
| B |
Boot |
|
|
|
|
Talking about the pins, what’s the pitch of them, is that a standard 2.54mm (0.1 inches), so can a normal header fit? Would be great to see a render of how that fits in the Framework 3D Printed Expansion Card housing so people (by which I mean me… 
)can think about how they can use this.
1 Like
Thank you for the table and nice comments! VCC is 3.3v. Yes, pins will be generic 2.54mm soldered horizontally on the pads front and back.
I will get it produced soon but I can look into 3d rendering (I have zero experience with 3d)
I changed the board a bit before sending the gerbers to NextPCB, our PCB prototyping sponsor. I realized I missed several important stuff to consider written in hardware design with rp2350 pdf, such as inductor’s magnetic field and the need to put decoupling caps near to the chip. Other than fixes, I got the board certified as open source hardware by OSHWA.
If any issue comes to your eyes let me know!
4 Likes
I consulted some PCB experts and they told me to mind impedance matching at USB, making them nearby and not using vias. They also told me to try putting decoupling caps nearby to the chip, and leave inductor area unfilled to reduce magnification effects on the 3v plane. Also I was about to forget PCB should be 0.80mm for MOLEX 1054440011 to fit in USB Type-C footprint as said in the datasheet. Here is the ultimate design before assembly by NextPCB.
4 Likes
Hi, thank you! Sorry for late reply. I will definitely add more stuff after I get the first batch of PCBs from NextPCB. I want to see it functioning first, then this can serve as a baseplate for future designs with gadgets embedded on it rather than having only GPIO exposed.
1 Like
MJ1
14
Ohhhh, that would be sweet! I was thinking that.
I don’t know how much room you could get by packing components closer, but it would be nice if you could fit a something like a 0.5mm pitch ffc connector for connecting a daughter board.
1 Like
It’s usb1.1, you can get away with a lot there but still good practice to do it right.
1 Like
yeah, i think thats mostly when dealing with usb 3.0+ (i’ve never seen it cause problems with usb <= 2)
Still a good idea to do at least rudimentary length matching but from personal experience usb1.1 can deal with some really crappy wiring.
1 Like
NextPCB is not very happy about male Type-C footprint. I will go with Suggestion 2. They are GND pads so there shouldn’t be major issues (hopefully). I have also decided to switch to solid connect from thermal relief so the connections to the planes are more steady.
1 Like
fair point, not like it adds that much design complexity either
just curious, would you recommend more precise matching measures for 2? i’d always just done basic length for 2, and never worked with 3 because of complexity that i dont need for most projects
Haven’t used usb2 in pcbs all that much myself but the few times I did I got away with not even properly length matching it. But I am probably not the best person to ask about best practices there XD.
1 Like
