PCB etching for MCP1624 boost converter

I'd previously attempted to hand solder a MCP1624 boost converter, but for some unknown reason it failed to work. This time I'll try making an etched PCB for it.

MCP1624 test circuit schematic

I designed the board in KiCad. Component values are based on what I have available, with the output voltage set at 3.3V for testing purposes.

Partial board layout

The DFN package is inconvenient for a single layer design, so I need two layers. Because it's very difficult to align the layers accurately with my current equipment one of those layers will be a solid ground plane. If I had a better fabrication process I'd put the feedback loop on the ground plane layer, but instead I'm using an air wire. There's a recommended layout for the SOT-23 package, but it's not much help in this case. I attempted to keep the current loops as small as possible.

I've slightly adjusted some pad sizes from the datasheet recommendations to compensate for the lack of precision of my toner transfer process and help solder it without a solder mask. I drew one trace in Inkscape instead of KiCad because it needed to be routed around these modifications.

Toner on copper clad board

I transfered the toner using a cheap laminator. It doesn't run hot enough so I preheat the toner with a hot air gun. This gives a good clean transfer.

Applying toner reactive foil

Toner isn't a perfect etch resist, so I applied toner reactive foil using the laminator. This messes up the edges slightly but it's still good enough after some minor cleanup with a scriber. I'm not 100% convinced the foil actually improves the quality of the final board.

Toner reactive foil applied

I colored in the ground plane with permanent marker as etch resist:

Ground plane colored with permanent marker

I tied some wire to the board so I can fish it out of the etchant. I used ferric chloride + citric acid (Edinburgh Etch). The citric acid avoids the need for agitation.

Finished etching

After about 30 minutes (it's winter, so room temperature is cold) the etching is complete. The cyanoacrylate glue I used to hold things in place also acts as an etch resist and it spread over the board too much so I needed to scratch some more copper away. Soldering will be challenging, I'll post an update after attempting it.