Low force microswitches

I was successful - it turned out that microswitches were of much better quality than any keyswitch I had previously encountered.

The best of the switches I tried was this one:

Saia-Burgess microswitch

G3M1T1PUL - Saia-Burgess G3 low force microswitch (dismantled)

G3M1T1PUL - Saia-Burgess G3 low force microswitch, 15g, 7A, 250V ac

G3M1T1PUL - Saia-Burgess G3 microswitch (upper contact bent down)

G3M1T1PUL - Saia-Burgess G3 low force microswitch (dismantled)

G3M1T1PUL - Saia-Burgess G3 low force microswitches

The specs of this switch may be found [here].

Video

Details

After preloading with a relegendable keycap, the required activation force was 8cN. This is very low - perhaps even lower than is desirable.

The activation travel is small - 1.2mm according to the specification.

The total travel is typically large compared to the activation travel, and the force increases in a pleasing non-linear manner as the total travel is approached.

Also, the tactile and auditory feedback with this switch seem to be of very high quality.

The switches are referred to by their manufacturer as being snap action switches - a reference to their tactile feedback.

The tactile feedback works by magnifying small changes in the activator displacement into larger displacements at the contact points. The catastrophic change in position of the spring results in an audible sound, and - significantly - a delicate click sensation being transmitted back to whatever is supplying the activation force.

The click sensation is extremely precise. Almost all other low force keyswitches I've tried (excepting the Omron microswitch below) feel mushy by comparison.

Most tactile feedback mechanisms operate at some risk of transmitting too much energy back into whatever activates it - with the potential of causing cumulative damage disorders. The action of this switch seems to be subtle enough to avoid causing problems - while being dramatic enough to provide the desirable activation signal at a level where it can be detected.

The overall effect is reminiscent of the old IBM "buckling spring" keyswitches - but with a much lower activation force.

The switch uses a "trident-shaped" beryllium copper spring.

The noise created by the switch is noticable. It is increased by the casing of the switch, which acts rather like the surface of a drum.

I have estimated the force diagram of the switch as I use it:

Saia-Burgess G3 low force switch - force curve

A physical catastrophe takes place at the activation and release points, which is what provides tactile feedback. The catastrophe leaves a signature in the form of a near-vertical precipice in the force curve. The switch shares this feature with the buckling spring keyswitches used by IBM.

The difference between the red and green lines represents the hysteresis of the switch, which prevents it from chattering around the activation point.

Tactile and audio feedback is also provided on the upstroke - though if you take your finger off the switch fast enough, the tactile element may not regiser.

Switch surgery is possible - and the switch is fairly hackable. The contact points can all be bent - changing the behaviour of the switch. The spring itself can be removed, inverted and reinserted - which makes it offer more resistance.

The activation force can be increased to at least 30 cN - or reduced to next to nothing. The activation distance can be increased or decreased. The strength of the tactile feedback can also be increased or decreased. However these changes can not all be made independently.

The factory shipped switch has a larger activation force, longer initial travel - and some initial prestress. Also a stop terminates the travel at 4mm, before the final small nonlinearity. I remove the stop if fitting a Cherry keycap.

The activation force of this switch is certainly extremely low - and the tactile and auditory feedback seem to be of excellent quality.

The resulting feel of these switches is excellent.

If I had to criticise I would point to the lack of initial resistance.

It seems to me that switches should have some initial resistance - and then 'give' a bit before activation - the idea being to give the user as much advance warning as possible about when the switch is about to make.

Perhaps my ideal switch would have a force curve something like this:

My ideal switch - force curve

In the far future, servo mechanisms may be used to provide tactile feedback - but until that happens, switches a lot like this one appear to be attractive as keyswitches.

There are a number of drawbacks to using these switches in a keyboard:

• Cost: these switches are not the cheapest available;
• Large size: these switches are huge;
• Too long: these switches are longer than a typical keycap;
• Awkward pins: these switches are not designed for PCB mounting;
• Lifespan may not be great - these switches are rated at 1-10 million operations;

Attaching a standard Cherry MX keycap can only realistically be done by leaving off part of the case - possibly creating some problems with dust and dirt.

I have used two of these switches to replace the left button of my mouse.

So far I have constructed three keypads and one full keyboard using these switches.

See here, here and here for more details of the keypads. The results are excellent.

I list suppliers of this switch in the links section - at the bottom of this page.

Omron microswitch

Interesting - but not as good as the Saia-Burgess switch - from my point of view.

The switch costs half-as-much again - and it may be more robust. However, the default spring force is greater, the audio feedback is not quite as good - and fitting keycaps looks as though it will be much more difficult.

Other microswitches

I quite like the idea of a lever - it alows a configurable tradeoff between activation distance and activation force to be made.

I bought these switches because the specifications claimed low activation forces. However, these switches were useless to me: the activation forces were far too high. The manufacturers must have been measuring the activation force at the far end of the lever.

I bought all the switches pictured here from [RS components].

Tactile feedback

I hypothesize that direct spinal circuits can be formed to deal with relaxing upon encountering tactile feedback - based on existing withdraw-on-finger-prick circuitry.

By contrast, auditory feedback necessarily has to travel via the brain - which is slower - and thus less effective.

Often good quality tactile feedback seems to require high activation forces.

The G3M1T1PUL microswitches show that very high quality tactile feedback is possible while retaining a very low activation force - provided the activation point is not constrained to be at exactly the same position as the deactivation point.

Saia Burgess G3 part numbers

Links

This site

Saia Burgess

Saia Burgess retailers

Low force switch design

Other