Switch Lubing: Where the Change Shows Up in the Trace

Two people lubricate the same switch and report opposite results. One finds it transformed, the other finds it barely different and slightly mushy. Both outcomes are real, and the difference usually comes down to where the lubricant went and how much of it was applied rather than whether lubrication works at all.

We measured switches before and after several treatments, and the traces showed exactly what changed. Friction dropped, the small irregularities in the force curve smoothed out, and the sound changed character. What did not change was the overall shape, which is the most useful thing to know.

What does lubrication actually remove?

It removes friction between moving parts: the stem against the housing, and the spring against whatever it touches. Friction is what produces grit, that slightly rough sensation during a press, and it is what produces most of the scratchy sound in an unlubricated switch.

It does not change the spring rate or the shape of the force curve. A tactile switch remains tactile after lubrication because the bump comes from the stem geometry, not from friction. Anyone expecting lubrication to remove a bump is treating the wrong mechanism.

Where does the change show up in a measurement?

In the fine structure of the curve. An unlubricated switch shows small irregularities superimposed on the underlying shape, and those irregularities are friction events. After lubrication the curve becomes noticeably cleaner while following the same overall path.

The total force also drops slightly, because friction was adding to the resistance the whole way. The drop is small, but it is measurable and it is one reason a lubricated switch can feel lighter without any change to the spring.

Why does too much lubricant make things worse?

Because excess adds drag. A thick or over-applied lubricant creates viscous resistance, which slows the return stroke and gives the press a dampened, sluggish character. That is what people describe as mushy, and it is the most common failure of a home treatment.

It also attracts dust. A heavily lubricated switch collects particles over time, and the resulting paste increases friction rather than reducing it. The treatment that felt good initially can degrade into something worse than the original.

Does the spring need lubricating separately?

Only if it is making noise. A spring can produce a ping or a crunch as it compresses, particularly if it is touching the housing or has a rough end. Lubricating it addresses that specific sound and nothing else.

Doing it adds risk, because lubricant on a spring can migrate and it changes the spring’s behaviour slightly. If your switches are not pinging, the spring is not the problem and treating it will not improve anything.

What about the stabilisers?

These matter more than the switches for wide keys. A stabiliser is a small mechanism with several contact points, and it is a common source of rattle and ticking. Lubricating it directly addresses a sound that is louder than anything a switch produces.

This is the highest value target on most keyboards. The wide keys are used constantly and their stabilisers are usually the least carefully assembled part of the board, which is why treating them produces the most noticeable improvement per unit of effort.

How consistent is the result across a board?

Less consistent than people hope, because hand application varies. We measured neighbouring switches in the same board that differed after treatment, and that inconsistency is itself noticeable during typing. A board that feels uneven is worse than one that is uniformly mediocre.

Doing a careful, repeatable job matters more than doing an elaborate one. A light, even application on every switch beats a heavy, uneven one, and the difference is obvious in both the traces and the hand.

Does lubrication change the sound?

Yes, and often more than it changes the feel. Removing friction removes the scratchy high-frequency component of the press, leaving the impact of the stem and the resonance of the case. The result is usually described as deeper and cleaner rather than simply quieter.

Over-application changes the sound again, adding a dampened quality with less attack. People who dislike the sound of lubricated switches are usually reacting to this, and it is worth knowing that it is a quantity problem rather than a principle.

How long does the effect last?

It depends on how much was applied and how the board is used. A light application on a lightly used board lasts a long time; a heavy one on a daily driver degrades sooner, partly through migration and partly through dust. Neither outcome is a failure exactly.

The measurable sign of ageing is the return of the small irregularities in the force curve. That happens gradually, which is why people rarely notice the change until they compare with a freshly treated switch.

Is it worth doing on an expensive board?

That depends on what is wrong with it. If the complaint is grit or scratchiness, lubrication is the correct treatment and it works. If the complaint is that the switch is too heavy, too light, or has the wrong bump shape, lubrication will not help because those are geometry and spring properties.

Diagnosing before treating is the part people skip. A force curve taken before any work tells you whether friction is actually the problem, and it takes less time than disassembling the board.

What are the alternatives?

Buying switches that come treated, using films to tighten the housing, or changing the keycaps. Each addresses a different part of the same system, and each has a different ratio of effort to result. Films address wobble, caps address sound, and lubrication addresses friction.

Combining them without diagnosing usually produces a board that is different but not better. Changing one thing at a time and measuring in between is slower and it is the only approach that reliably improves the result.

What should you conclude about the whole practice?

That lubrication is a targeted treatment for a specific problem, not a general upgrade. Applied to a switch with genuine friction it produces a clear, measurable improvement. Applied to a switch that was already clean it mostly adds effort and risk.

The trace is the honest judge. If the fine structure of the curve does not change, nothing you did made a difference, and no amount of reporting otherwise will make the board feel better.

What follows from understanding friction?

Does the type of lubricant matter?

Yes, mainly in viscosity and in how it ages. A thinner lubricant reduces friction without adding much drag, and a thicker one adds noticeable damping alongside the friction reduction. For a switch, thinner is almost always the right choice.

Ageing matters too. Some lubricants dry out or migrate, and a treatment that felt right initially can change character over months. Choosing one known to stay stable is worth the small extra cost.

Should you treat every switch identically?

Standard practice says yes, and there is a case for varying by position. Keys that are pressed constantly benefit most from a clean stroke, and stabilised keys benefit most from stabiliser treatment. Treating everything the same wastes effort on keys that do not matter.

The risk is inconsistency. Varying treatment across a board makes it feel uneven, and an even mediocre result is usually preferable to an uneven excellent one.

Does lubrication affect the switch’s lifespan?

Probably positively, because less friction means less wear. The effect is hard to measure over a reasonable period, and it is not a reason to lubricate, but it is a reasonable expectation.

The larger effect on lifespan comes from reduced contamination. A lubricated contact is somewhat protected from the environment, which slows the oxidation that eventually causes problems.

Should you lubricate a brand-new board?

Only if it needs it. A new board that feels clean gains little, and opening it introduces risk and usually voids something. Measuring or simply using it for a while first is the better order.

Where a board is gritty from the start, lubrication is the correct response and it works. Diagnosing before treating remains the rule.

Can lubrication be removed if it goes wrong?

Yes, with cleaning, and it is tedious rather than difficult. Disassembling and cleaning each switch takes far longer than lubricating them did, which is the best argument for applying sparingly in the first place.

An over-lubricated switch is recoverable, and the recovery costs an evening. Measuring the result before treating a whole board avoids discovering the problem forty times over.

Does amount or placement matter more?

Placement, because the wrong place adds drag without removing grit. Lubricating the contact surfaces that actually rub is what produces the improvement, and coating everything does not.

A light, correctly placed application beats a heavy, general one by a wide margin. That is the single most useful thing to know before starting.

The next variable is housing tolerance, which decides how much friction exists in the first place. Measuring wobble directly is the natural continuation, and it tells you whether lubrication is even the right tool for a given switch.