Even after modifying my switches with silicone and dental floss, I still wasn’t satisfied with the noise of my Ergodox EZ. The dampened upstrokes were still causing a reverberation that I determined was coming from the Ergodox EZ case/PCB itself.
The Ergodox EZ CIY case has an integrated plate-mount design (where the “plate” is just part of the ABS upper shell), which is a design that’s notorious for producing reverb. The case is also very hollow, with 1–3mm gaps above the PCB and 2–4mm gaps below it. This all contributed to a mid-range “thonk” around 1kHz on upstrokes.
I successfully dampened the “thonk” sound by adding neoprene foam and rubber. This cost less than $10 (AUD) in materials and took about 90 minutes.
There’s about 1.4mm clearance between the bottom housings of the switches and the PCB. I initially tried to insert (~1.5mm) O-rings around the bottom of the switch stems (i.e. the outside of the switch housing), but they were too thick and made it difficult/impossible to “click” the switches into the plate (more so for the Kailh than the Cherry MX housings). More surprisingly, it also didn’t seem to help with the noise.
The neoprene I inserted between columns in the top housing is 1.5mm thick, which isn’t thick enough to be truly “sandwiched” (it doesn’t contact the plate and PCB simultaneously). Anecdotally I do find that it helps reduce noise, but I’m not sure why (it doesn’t seem like enough mass to have a deadening effect, and lack of pressure means it shouldn’t be undergoing deformation). At a guess, I think it’s loose enough that high amplitude vibrations of the PCB might cause it to lift off, and it absorbs energy elastically when falling back down. It’s probably also working a bit as a generic reflection surface / gap-filler. It would be interesting to test whether the less-dense and thicker neoprene foam is more effective.
The 2mm-thick neoprene foam in the bottom case is probably just acting as a gap-filler, with its porous surface diffusing rather than reflecting sound. Again, it doesn’t have much mass, but there’s likely some damping due to it being loose enough to “flap” in the air. It does exert pressure against the PCB in places, namely where it’s sandwiched between the PCB and the ridges on the inside of the bottom casing, which would have a damping effect on vibrations of the PCB.
The measured peak loudness didn’t change much from this mod (still ~50dB for normal typing), but the duration of reverb was reduced. Qualitatively this changed the metallic “thonk” ringing to a more plastic-sounding “thock”.
The main downside to this mod is voiding ZSA’s warranty by opening the case. It seems low-risk to me in terms of damaging the keyboard, although I was mindful of electrostatic discharges while handling the PCB. Ongoing concerns may be risks to the ICs on the PCB due to heat buildup (the neoprene acting as a thermal insulator) or electrostatic discharge (if the neoprene were to pick up a charge from rubbing against other parts). Again, these seem low risk to me.