Which is better, buttons or rotary encoder?

[metal]

I am planning a new device. This has a timer and dimmer function. While coding, I opted a direction that I no longer think is accepted, to be more precise, I am not sure about my decision, not any more. There is only on button, which is used to change back and forth between hours and minutes, while the rotary encoder inc/dec the timer hours/minutes. When timer is set, a long press on the button starts the timer and the dimmer function.

The rotary encoder now is used to inc/dec the dimmer, while a long press on the button terminates the timer and resets everything.

Which is better, to use rotary encoder or buttons to inc/dec hours/minutes for the timer? At the end I am a coder, I am unable to decide which is better. People who saw that it is possible to use the rotary to spin hours and minutes, were technically impressed, they are techs like me, unable to think the user way.

May be other techs have different opinions!?
All opinions are welcome :- )

[Vineyards]

If you are planning to use the device in a damp environment, you should prefer membrane switches which can be designed complete with LED's and buttons. If not you should find a way to keep your buttons or rotary encoder free from corrosion. If not, they will not last especially if they are used outdoors or in humid interiors. You could also consider a joystick that performs the function of several buttons in a tiny package. Corrosion will still be the major problem though.

[Magnox]

I'm another techy of course, although more on the hardware side really than coding, and I much prefer a rotary encoder over up/down buttons. So did my wife when she got to play (i.e. I showed it to her and said 'go on then, turn something') with the prototype of the function generator I'm building - and she's definitely not a techy!

So that's two votes for an encoder over buttons.

I use rotary encoders with a built-in button click (by pushing the shaft) and incorporate that for functions such as alternating between seconds, hours and minutes, or hertz/KHz/MHz for instance. Anything that makes sense in conjunction with whatever is being adjusted with the rotary action. I don't know if they all have that.

I also use the rotary encoder for menu selection and such like by dividing the count down so that it isn't too sensitive. In that case the included switch is usually the 'Enter' or 'Go' button.

[metal]

Thank you for your replies.

I am using the encoder attached, unfortunately it has no switch. This encoder is a direct replacement for ALPS STEC16B, datasheet is also attached, 3rd in order.

The thing that was irritating me is that while I was experimenting to see if the code is working fine, I had to use both hands, one to hold the encoder, the other to spin it, and then the same hand to push the button. It is unlikely that I will get the encoder changed at this phase, however, I feel that the encoder with a switch is much more comfortable. Ok foxyrick, I will go with your wife's opinion and leave the rotary encoder functions unchanged. But I am still thinking about the switch position relative to the encoder? Which is better, put it on the left side of the encoder, or the right side?

Vineyards, your idea of the membrane switches is awesome indeed, but I am practically unaware of readily available single membrane switches with LED support, I have not seen any yet, I better do a deeper search later. When I was looking at images from google search, another idea popped up. I started thinking about a switch that comes with bi-color LED, as in the 2nd attachment.

Concerning damp environments, it is an indoor device, do you think I will have to put some anti-moisture bags inside the enclosure to ensure that moisture can be kept away?

[Parmin]

Metal,
if you want to impress the masses, you may think of using capacitive touch rotary slider.
Very robust (and damned cheap cost you made it yourself) too..

[sfiga69]

Why not use an encoder with built-in switch?
http://www.bourns.com/pdfs/em14.pdf
or
http://www.switchchannel.com/switches/encoder%20switches/re11/mechanical%20rotary%20encoder%20switch.htm

example:
press switch more 3 sec -> enter into the menu
wheel encoder-> navigate menus
press switch -> enter submenu
press switch -> sets option.

[metal]

Parmin, do you have experience with capacitive touch switches and sliders? Can you write down your personal experience please, it can be useful to me, I know that you are a mass-products expert, I 've never tried to tackle this topic :- )

I tried an optical rotary encoders, those without a mechanical detent are very difficult to implement, you will end up counting the signals and dividing them on some number to obtain a decent response, furthermore, they are extremely fast. I have not tried those with a mechanical detent, may be they are more identifiable  in terms of pulse/detent, I am not sure about that. In addition, optical encoders are very expensive; the one I am using can be bought for less that a dollar for large quantities.

I also used RE11CT series, I did not like their mechanical detent :- ) What really impressed me was the one I showed earlier, its mechanical detent is very smooth compared to RE11CT types.

Actually, the capacitive thingy started spinning my head, I should give it a read today.

thanks to all

[solutions]

The touch slider is a copper rectangle with a small line of no copper running on the diagonal.  What you are doing is measuring the capacitance of the two triangular electrodes - more area touched by your finger is more capacitance. At the very left, you'll have max cap and almost no cap at the vertex of one of the two electrodes. The ratio of those two caps tells you where you are.  The downside is that it does not work with gloves or a thick overlay of glass or plastic.

A really good, easy to understand, app note on this at Microchip.

http://ww1.microchip.com/downloads/en/AppNotes/01334A.pdf

and

http://ww1.microchip.com/downloads/en/AppNotes/01298A.pdf

[pickit2]

funny I seen this in my bookmarks.
http://www.youtube.com/watch?v=5ZH1-eScY0M
also see
http://hackaday.com/2011/09/12/glass-pcbs/

[solutions]

I looked into the exact link on glass pcbs, pickit2, a couple of weeks ago.

I'm still scratching my head on how to do vias....and, no, not jumpers or crossovers, but vias

[Parmin]

The touch slider is a copper rectangle with a small line of no copper running on the diagonal.  What you are doing is measuring the capacitance of the two triangular electrodes - more area touched by your finger is more capacitance. At the very left, you'll have max cap and almost no cap at the vertex of one of the two electrodes. The ratio of those two caps tells you where you are.  The downside is that it does not work with gloves or a thick overlay of glass or plastic.

A really good, easy to understand, app note on this at Microchip.

http://ww1.microchip.com/downloads/en/AppNotes/01334A.pdf

and

http://ww1.microchip.com/downloads/en/AppNotes/01298A.pdf

As what he said..

[FTL]

What I hate with user interfaces is when there are a very limited number of controls that need an obscure process to control the device where different buttons have different functions at different times. I think that sort of thing is usually done to reduce the cost of a product since each switch costs money to add, while weird software functions are free once they've been written.

The thing with user interfaces is that the end-user has a model in their head of what this device is and how it should work. In a book about industrial design I've read about architecturally beautiful glass doors on an office building with the hinges hidden. People had a difficult time entering and exiting the building because the doors did not operate as they expected. That is, they expected to push on the side without a hinge, but could not easily determine which side that was so they struggled. Another example was study on fridge controls. In both cases there were two knobs. In one case one controlled the fridge temperature and the other the freezer temperature. In the other case one controlled the firide temperature, and the other the relative temperature of the freezer. The second case made sense to the engineers because it matched the internal model of the fridge (a thermostat and a door that directed cold air to the fridge or freezer). The first case made much more sense to the end users who wanted to be able to set the temperature of each portion of the fridge. That implementation more closely matched the model of teh device that they had in their heads.

In terms of your encoder, will it be possible to easily turn the encoder on the finished product with one hand? If it is small and light, two hands will be needed as you've found in your testing. If it is large and heavy, the encoder can easily be spun with two fingers and the button(s) pressed with another. I would avoid ever having to press two things at once, or havng to press and rotate at the same time.

The LCD provides a big help to the user interface in that it tells the user what to do and what state the device is in. It can provide instant feedback so the user knows whether what they are doing is the corect thing.

The bottom line is to make the user interface match the model the user has in their heads.

[metal]

solutions, the ADC idea is awesome, thanks!

PICKit, I also saw that last week :- )

FTL, this is leading us to the conclusion that I will have to implement some sort of capacitive switching. Your point was just correct, the device is not heavy, I have been thinking how I will rotate the encoder while experimenting and imagined the casing, it is just hell. In one f Dave's videos, he made many product reviews, and it was always important if the device is not heavy, to find some way to interface it to outside world without having to move the device backward. An example of this is new oscilloscopes, many reviews complained that both hands should be used to press a button. We should always think like an end user, our wives, children, parents are the best to ask to try something you made, they can report their experience on the fly, you were right about the fridge thingy, my dad always damned those engineers :- ))

I did a quick search yesterday on cap switches and sliders, I think they are pretty easy to implement. At least, one will no longer worry about changing mechanical parts. I will try to avoid cap sliders, just cap switches, I think I can implement long/short/repeat presses, then no matter how light the device is, user will not hassle to press a button. Once I am done with experimenting, I will post some photos.

thanks to all

[solutions]

Run a cap slider along the bottom or side(s) if you have space left after placing your switches. Nobody will know what it is, they'll think it's some kind of arty logo or whatever in copper, and you can play with it and "upgrade" the f/w later to a "linear" control, vs having to hold or tap switches.

[Ichan]

Hi Metal,

I play with AN1298 CVD some years back, I will post it as a new thread as I am gonna asking peoples to discuss about it.

-ichan

[metal]

I had long thought about the discussion we had, it seems that the project will under go major changes, it will take some time before I get something working.

P.S. thank you Ichan