Since I tested a lot of switches lately...
I wondered how much the different stem designs influence ingress of dirt/dust into the switch.
So, I conducted some tests...
Why Even Care?Dirt, dust, hair, etc. is something that hits every mechanical keyboard eventually.
If it finds its way into the switches, it can have a negative impact (scratchiness, malfunction).
No matter how careful you are, the open nature of classic MX switches will lead to contamination over time.
I’ve had some malfunctioning switches in the past that looked perfectly clean from the outside.
But after opening them, I found a tiny hair, or gritty lube inside.
How Stuff Gets In?The classic MX switch has a small gap all around the stem.
Small, but big enough to allow fine dust to slip trough.
When the stem travels down, the situation gets worse.
Cherry MX Grey with the stem pressed down:
Even with keycaps installed, dust will (WILL!) find it's way onto the switch.
Fine dust floats around in the room and moves with the air currents.
It will get into every place that is not properly sealed off.
Then there’s is also the classic move a lot of people do...
Flipping the keyboard upside down and giving it a good shake to remove debris.
Most falls straight out, but some gets caught in the inner side (cup) of the keycaps.
When you set the board back down, those particles drop right down onto the switch/stem.
What Household Dust Is Actually Made Of?For the most part, tiny fibers, and dead skin...
In rural/green areas, it might also contain things like pollen.
In dry/hot parts of the world, it might contain sand/soil (dust from outside).
But even within the same environment, the exact composition depends on the household.
In urban areas, dead human skin cells make up roughly 50% of typical indoor dust.
Broken-off fabric fibers (from clothing, carpets, bed-sheets, etc.) about 40%.
In a carpet-heavy home you’ll see a higher percentage of fabric fibers.
A home with hardwood floors will have a higher percentage of dead skin.
In hot/dry environments, most of the dust (up to 90%) might come from outside.
Why is this a problem?Over time, dust that gets into the switch, mixes with the lube, and creates a gritty paste.
Fibers that get in might wrap around the contacts or spring and cause binding or chatter.
Smooth travel becomes scratchy, tactile switches stick, keypressed dont register.
I even had some Cherry MX Blue switches in the past that stopped clicking.
Closed-Box style stems (like on Kailh Box switches) seem to add a protective wall around the stem.
This design significantly reduces the gap you get on the classic design when the key is pressed.
Kailh Box Brown V2 with the stem pressed down:
Same goes for closed-cylinder style stems (Gateron, Kailh) and other closed shapes.
Notched Box designs (AKKO, BSUN, OUTEMU, JWICK, TTC...) are (in theory) somewhere in between.
What did I test, and how did I do it?Basically, I did a accelerated contamination test...
But as this is just a "curious hobbyist" test, take it with a grain of salt.
1. Abrasive contaminants (dust and fibers) were generated from Scotch-Brite pads
2. Switches were exposed under (sorta) controlled conditions
3. Switches got actuated 10 times after the dust exposure
4. Switches got cleaned (outer side) with a static brush
5. Switches got evaluated for scratchiness, sticking...
Materials, Methods, ProcessI used key-switches from major manufacturers (Cherry, Kailh, Gateron...)
The Switches were chosen to represent a range of stem types and housing designs.
Prior to contamination, all switches were tested (in stock condition).
Multiple identical units of each model were used to enable statistical replication.
Each switch was mounted on a switch tester and actuated many times.
Actuation feel was documented (smoothness/scratchiness/stickiness).
I felt that dust from abrasives would be best, as it's ingress into the switch would be more noticeable.
I decided to go with Scotch-Brite pads which I rubbed against each other.
This not only creates scratchy dust but also small fibers.
The switch tester was securely fixed at the bottom of a small sealed container.
A "precisely measured" (not) amount of the contaminant mixture was placed into the container.
The container was sealed and vigorously shaken for 30 seconds to achieve a random/even distribution.
After shaking, the container was left undisturbed for 1 minute to allow airborne particles to settle.
Once settling was complete, the chamber was carefully opened under minimal airflow.
Each key-switch was manually actuated 10 times in succession.
Post-actuation, the switches were carefully cleaned with a soft static brush.
Performance degradation was evaluated (comparison vs stock-condition).
The Result?Box-style switches showed significantly lower contamination effects compared to standard MX-style switches.
Sticking only occurred in tactile switches, mostly the ones that had a weak return to begin with.
Designs with tighter housing tolerances seem to have a bit less dust inside of them.
Notched VS Closed Box did not seem to make much of a difference.
Lubed switches worsened more than un-lubed switches.
What next?Future work could focus on housing tolerances...
Just "Dusting -> Cleaning -> Evaluating" without the actuation before the cleaning.
Also I'd like to get (lend...) a microscope to inspect the inside of the switches visually.
What is your experience with dirt/dust/fiber ingress into switches?
How do you clean/de-dust your keyboards?
Recent Posts
