My first – very noob – soldering of a PCB with hot air is now behind me. Hopefully it won’t haunt me for too long!
Here is my process …
I cleaned the PCB and checked continuity of all of the traces. One was marginal and one was open so I made a note to fix those issues when I was soldering.
I printed out the PCB from KiCAD with the component footprints and part values.
I picked each component out of my “parts books” starting with the smallest and simplest and working my way up – resistors, capacitors, power regulator IC, switches, and the ESP8266. I placed them on the printout.
I used the solder paste dispenser to apply the smallest possible dot of paste to each component contact pad. this skill needs practice
I transferred each component from the printout to the PCB using tweezers, tapping them slightly to press them into the paste.
I soldered them in place using a hot air reflow station set to 320C and a very low airflow setting (to avoid blowing the components off the board).
I fixed the two bad traces – each with a single strand of wire and my soldering iron.
I inspected the PCB using a microscope and a multimeter.
How did I do? Well, not perfect.
The view under the microscope revealed I had not heated a few of the pads on the ESP8266 enough. Easy fix. The test of my soldering revealed I had a short on a trace I had put under a resistor (between its pads).
Let me describe how crazy that was. I am using 0805 resistors which means there is a less than 0.05” of space between the pads and I put a 0.015” trace down the middle of that space. The space between the trace and the pad is no bigger than the trace itself. It’s a DIY etched board which means it has a raw surface and does not have a masking layer of paint to cover up the traces. Also the traces are anything but perfect. Yeah, the odds of me soldering those pads and not dribbling onto that trace was a Las Vegas Bookie’s dream.
It took three attempts to reflow that resistor.
Does the board work? I don’t know. I still need to drill and solder the pin connectors – especially the 5-pin programming connector along the bottom.
Other lessons learned?
I need to buy 10K resistors in bulk. These projects use a lot of them. The component books are working out well. The 3D printed solder paste dispenser is indispensable – giggle. I’ll be glad when I have this board working because I’ll have more confidence to start using online PBC fabrication services and abandon the DIY etching process.
Postmortem: the board failed to program. After dissecting, it appears the ESP-12 got scrambled at some point.
Post Postmortem: a new board design has been completed and programmed. So at least there’s that.