Custom RP2040 PCB: Design, Turn-On, and Debug - How Hard Could It Be?

[@dru6809]

No sarcasm. That is a beautiful Rhode Schwartz scope. I hope I can get one someday. There seems to be a big change happening with scopes. This movement towards much larger displays and flatter boxes is welcome. Thx for making this video. I like how you showed us a more minimal alternative to the RP2040 recommendations for support circuitry.

I have just ordered my first RP2040 PCB design. I ordered it with top side assembled in order to avoid soldering issues. The PCB is the size of a DIP40 EPROM so I hope I did not make too many design mistakes! Thanks for sharing your problems!

[@HL65536]

I fixed a fried pi pico once, soldering a new RP2040 onto it. That was my first try at soldering components without (at least stubby) legs. I used an upside down clothes iron as a hot plate and a one of these torch lighters as a hot air tool. I applied solder to the pads and the chip using a soldering iron with lots of solder on it (of course, using lots of extra flux). I managed to get all contacts working at only the second try. The key to success was pressing the chip down onto the board, connecting all the pads and squeezing out the excess solder. Verification of all pins being connected is also possible without it booting up, even while that thing is still hot. Just use a multimeter on diode mode and test all pins (positive probe on GND). Connected pins should read somewhere around 0.3-0.8V (also the debug pins and the usb data pins). The other GND pins should of course measure 0V.

[@TheGmr140]

Nice boards and design 😊

[@andywest5773]

Great project and very informative! I'm a big fan of the RP2040, but this would definitely be a "hard" difficulty rating for me.

[@bald_engineer]

Hah, well the idea behind my qualified "it is easy if..." is that now it should be easier for everyone to debug/turn-on an RP2040 design! I'm sure yours would have far fewer problems compared to mine. ;)

[@danblankenship5744]

Getting the correct amount of solder paste will always plague the home hobbyist. I like using an IR under-heater with hot air on the top. I then will use an iron to reflow those QFN contacts.

[@darrenslab5537]

I have used the RP2040 in a project of mine and have found that hotplate soldering gets them down first try

[@bald_engineer]

I really need to get a hotplate!

[@Tenniszogger]

Can you recommend a UF2 converter? Currently I have some trouble with the conversion from bin to UF2, Because of that the Chip does not Jump into application…

[@dawnofgods]

10% pad reduction works fine for me on on a 100 um SMD stencil

[@bald_engineer]

That's a good point. My friend has been cutting many of my past stencils for me. We worked out some reductions based on pad sizes that work well. I used OSH Stencil for this stencil and just left everything on default in KiCad. Next time I order from them, I might re-submit this one with a reduction and compare it to see if it works better. The real problem is that my paste was over a year past its expiration date. So, it didn't hold its shape well.

[@vijaysulakhe5605]

Many Thanks. You are showing what practical problems one will face, so one will be ready.

[@MAYERMAKES]

Do you have to list everything I did wrong on my first RP2040 design....uughghghg

[@RottnRobbie]

11:38 - Say what? Too much IPA damages pushbuttons? How? [ "How", as in both: - what's the mechanism of damage? (It's a _button_ !) and - how can there be 'too much' IPA? ]

[@bald_engineer]

It was just a passing thought. It's more likely damage from spending too much time under the heat gun.

[@RottnRobbie]

@@bald_engineer OK, thanks for clarifying. I was afraid I might've hurt my computer mouse when I used IPA to clean the circuit board yesterday...

[@bald_engineer]

@@RottnRobbie As I think more about why I said that, I remembered I had damaged an SMT microphone around the same time. And I was researching ultrasonic cleaners and surprised how many components are damaged by those. So, washing failures were on my mind. But it’s more likely flux got into the switches and melted something.

[@vijaysulakhe5605]

Hope someone develops ZIF socket

[@winsrrow8125]

if u know what do look for AND U HAVE THE TOOLS... i dont have an osciloscope with memory nor a logic analizer, i have the boards i did with the just reload to bootloader problem, i didnt asembled it, was factory asembled so... and i couldnt make use of debugger, my pc setup isn't working and idk why or how to solve it.

[@jeremiahbullfrog9288]

But how would you have troubleshot this without access to a $20,000 scope?

[@MAYERMAKES]

With a basic scope. He used only functions that are presentvon even the cheapest real scopes.😮

[@bald_engineer]

I can’t figure out your joke’s punchline. So I have to ask… how?

[@HL65536]

multimeter in diode mode. Positive probe on GND. Test every pin that comes from the RP2040 using the negative probe. Every GPIO pin should read the same value (within ~5mV of each other). gpios shorted together should read ~40mV less, GND (and stuff shorted to it) should read 0V. If it reads open, it's not soldered on. Also test the special pins (usb data pins, debug pins).

[@jeremiahbullfrog9288]

@@HL65536 Thank you for providing a useful procedure