Use a rotary switch as rotary encoder

[Freezingmoon]

Hello,
For a project I need an incremental rotary encoder, I've tried dozens of various type but they all need too little force to "click". I really like the snap feedback of the traditional rotary switches, I have a few with continuos rotation I'd like to use. It is possible, do I need additional circuitery?

 

[D.A.(Tony)Stewart]

I think this suits your needs https://www.ebay.ca/sch/i.html?_trk..._nkw=Rotary+Encoder+detent&_sacat=0&_from=R40

But without a spec for peak torque you need, it is hard to compare.

Old fashioned rotary switches had lots of friction and contact bounce so a stiff spring loaded ball-bearing and socket detent overcame these hysteresis friction effects to accelerate the transition speed with stiff positive force feedback until position detent is reached with overshoot and then positive force feedback again , oscillating quickly to lock into position with minimal oscillations or well damped.

For continuous volume controls, you don't need this level of force feedback since the quadrature switches have low friction and you only want a light click tactile feel.

Discussion of solution options
To increase the force feedback, that you prefer, one needs a solution that is either passive mechanical or magnetic or electromagnetic force feedback.

Mechanical is obsolete and wear-prone with a spring loaded ball bearing detent if high force. You only need the spring to be twice peak force of the friction torque so modern detent torque is much less.

Magnetic force-position loaded uses permanent rare earth magnetics, which adds much cost to the rotary switch or uses capacitive quadrature position sensing with electromagnetic force feedback, which is much more expensive.


When you have done this analysis for force vs position between detents , you may decide the solution is a stepper motor with friction. https://www.google.ca/search?q=mini...7.4292j0j7&sourceid=chrome&es_sm=122&ie=UTF-8
They can have strong magnetic locking positions depending on rating of micro stepper motor. Adding more torque can be as simple as shorting with a shunt resistor and monitoring current across this 75mV shunt to determine detent position and direction from quadrature decoding of pulse position. More complex touchy feel detent detection can be done with active current feedback to accelerate and brake from detent to detent by sensing position of quadrature phases, velocity and then assist with electronic force feedback from negative force to positive force using magnets and electromagnet coils in a micro-stepper & driver to emulate any rotary switch force with any fancy force feedback you want by designing of these parameters for force vs rotary position, velocity and acceleration.

 

[crutschow]

A normal incremental rotary encoder has a two-phase pulsed output so the rotational direction can be determined. I don't see an easy way to generate those signals with a standard rotary switch. How many indents does your rotary switch have?

 

[alexan_e]

Maybe with resistors in a ladder style to give different output voltage in each step and read it from the adc,
like shown

 

[dick_freebird]

I think using a old-school mechanical rotary switch would
be OK for "front panel" type use, but is liable not to last
as part of a high-cycle mechanism. Maybe you want to
separate the mechanical detent function, from the encoding,
each to its strengths.

 

[Freezingmoon]

Hi all, thanks for the reply.
I really can't add a stepper motor or more complex circuitery. I really can't speak about the use of this switch, but it will be placed in a front panel for an application. No harsh conditions. The problem with rotary encoders is that sometimes since the low torque they need, the user accidentally made too many clicks, where the traditional rotary switch need a way higher torque per click.
This switch/encoder will be connected to an USB board with support for standard gray code encoders.
Maybe an encoder simulation solution with a micro is possible?

- - - Updated - - -

A normal incremental rotary encoder has a two-phase pulsed output so the rotational direction can be determined. I don't see an easy way to generate those signals with a standard rotary switch. How many indents does your rotary switch have?

12.
I was thinking with a solution like this is possible to know the rotation direction of the switch: "If I close contact 3 from 2 the rotation is clockwise, If I close the contact 3 from contact 4 is couter-clockwise etc" But I think I need a micro of some sort.

 

[FvM]

Incremental encoders are made with different torque and pulses per revolution, cheap mechanical and more expensive optical switches. In so far I guess, you just didn't find the right product out of the large market offer.

Many digital measurement instruments (e.g. oscilloscopes or generators) are using incremental encoders for their controls, I don't see a problem of too low meachanical feedback with these switches.

Some standard rotary switches have removable stops, so they can be continuously rotated. A 3x4 switch can be easily converted to a quadrature encoder, but it must be make-before-break type to avoid illegal stae transitions.

 

[Freezingmoon]

Incremental encoders are made with different torque and pulses per revolution, cheap mechanical and more expensive optical switches. In so far I guess, you just didn't find the right product out of the large market offer.

Many digital measurement instruments (e.g. oscilloscopes or generators) are using incremental encoders for their controls, I don't see a problem of too low meachanical feedback with these switches.

Some standard rotary switches have removable stops, so they can be continuously rotated. A 3x4 switch can be easily converted to a quadrature encoder, but it must be make-before-break type to avoid illegal stae transitions.

Hi, I know for an application like an oscilloscope a "standard" rotary encoder is good. But in my application sometimes the user don't really have the time to rotate the encoder for just one click. The rotary switch I'm sure will avoid this problem due to the higher torque its need.
Can you tell me more about this 3x4 switch? Thanks!

 

[D.A.(Tony)Stewart]

Hi all, thanks for the reply.
I really can't add a stepper motor or more complex

A micro stepper motor gives the cogging feel without any circuit due to magnetic poles. The two coils with a quadrature logic and generate the step pulse and direction outputs needed or if you prefer , step CW and CCW. This can be done with gates and RC edge detection or flip flops or any " quadrature detector"

Really simple rotation with strong magnetic stops between positions is a engineering matter of specs and design. This method is simpler than it sounds.

 

[Freezingmoon]

Found this document: https://www.737ng.co.uk/simple encoder.pdf the guy is using the same USB board as mine but I can't make it work...