Disaster!
I finally got around to putting some components on the latest version of my Dual Temperature PCB. Unfortunately, I found a huge mistake!
The Module
I started assembling the board with the MinewSemi nRF54L15 module.
I made a better first pass with this than I did with the DevKit. Plenty of flux!
This was a huge improvement. I just went over the connections again to remove the bridges. A dab of solder on the remaining pin and it looked good!
SMD parts are small. Damn small!
For this version of the board, I’m using SMD components. I knew they’d be small, but bloody hell.
I had ordered a set of tweezers just for this occasion. I opted to just start with the power components. Two capacitors and an LDO. I opted for the solder paste and hot plate to solder these. It’s pretty horrific, but again, represented an improvement. Might not look good, but the connections were sound.
Trouble with the voltage
I added in the programming header (visible in the bottom right of the photo above) and applied 5V.
I expected my LDO to output 2.2V, but instead it output 4.7V. This was too high for the module, which I may have fried. I couldn’t understand it. The MCP1700 should dropped the voltage.
Maybe I had a short under the LDO? I removed the power and tested the resistance between the battery terminal and the power input to the module.
No resistance reported, which led me to believe there was no short.
Did I mess up?
At this point, I went back to KiCad. Had I really messed my PCB layout? The schematic looked fine, assuming that VI was voltage in and VO was voltage out.
When I checked the PCB (which I’d already checked 100 times) it looked fine.
I decided to check the MCP-1700’s datasheet.
Look upon my works, ye mighty, and despair.
According to this, pin 3 was the voltage in. Damn.
Verifying the mistake
It was indeed the wrong way around, I figured I should be able to verify this. The GND pin was in the right place, so I just needed to swap the power input.
On my board, I had allowed for a VDD header pin to act as a reference for my J-LINK programmer. I applied my 3.3V to that pin and measured the voltage.
It was 2.2V as expected. This also highlighted the fact I’d ordered the wrong LDO: 220 instead of 330 🤣🤣🤣
Where did I go wrong?
So, what did I get wrong? Both the schematic and the PCB looked fine. I posted my question to Reddit and got an immediate flurry of responses.
My mistake was a mismatch between the symbol and the footprint. My footprint was correct, but I was using the wrong symbol. The selected value of 330xxTO was wrong. I should have used 330xxTT.
This has the correct PIN assignment.
Back to the drawing board
Not the end of the world.
I learned a lesson, which is the important thing. I’ll update the PCB layout and order another batch.
In the meantime, I’ll continue testing the PCB I have (hopefully my module has survived). I still have to attempt to flash code onto it! Fingers crossed it will work.
It’s all part of the journey I suppose 🙂
Support
If you found this blog post useful and want to show your appreciation, you can always buy me a coffee. Thanks!!
Leave a comment