Electrical engineers or tinkerers -- need rotary component for etch-a-sketch knobs

[Leros]

I'm going to build to build a digital etch-a-sketch for my final lab. We're designing our own controller board and then building whatever we want on top of that.

I'm trying to figure out what would be the best device for the two knobs.

It would also be neat if the knob could stop rotating once we hit the edge of the screen. A potentiometer wouldn't really allow this since we would reach a maximum ADC value before the pot stopped turning.

Any suggestions?

 

[Demon-Xanth]

If it's digitally controlled, rotary encoder.
If it's analog, put a series resistor to limit the ADC so that the pot would be full scale.

 

[tranceport]

ipod wheels... get with the times man..

 

[edro]

You may be able to limit it by getting a potentiometer that has a consistent amount of turns (10 is standard).
I would make the software have a limit on it though.

If your array is like 10,000 pixels wide, make it stop at 9,999 pixels.

 

[Leros]

Originally posted by: edro
You may be able to limit it by getting a potentiometer that has a consistent amount of turns (10 is standard).
I would make the software have a limit on it though.

If your array is like 10,000 pixels wide, make it stop at 9,999 pixels.

Ideally one knob would turn twice as far as the other knob. My screen is 128x64 pixels (very limited budget allowed).

 

[blahblah99]

Originally posted by: Leros
I'm going to build to build a digital etch-a-sketch for my final lab. We're designing our own controller board and then building whatever we want on top of that.

I'm trying to figure out what would be the best device for the two knobs.

It would also be neat if the knob could stop rotating once we hit the edge of the screen. A potentiometer wouldn't really allow this since we would reach a maximum ADC value before the pot stopped turning.

Any suggestions?

Why would you reach a max. adc before hitting a stop on the POT? That just means the amplifier is designed improperly.

Step down the supply voltage or resistor divider or gain values so you can maximize the dynamic range of the ADC with full turn on the pot.

 

[Leros]

Originally posted by: blahblah99

Originally posted by: Leros
I'm going to build to build a digital etch-a-sketch for my final lab. We're designing our own controller board and then building whatever we want on top of that.

I'm trying to figure out what would be the best device for the two knobs.

It would also be neat if the knob could stop rotating once we hit the edge of the screen. A potentiometer wouldn't really allow this since we would reach a maximum ADC value before the pot stopped turning.

Any suggestions?

Why would you reach a max. adc before hitting a stop on the POT? That just means the amplifier is designed improperly.

Step down the supply voltage or resistor divider or gain values so you can maximize the dynamic range of the ADC with full turn on the pot.

I've never actually messed around with maximizing the range before. Can I maximize completely with resistors without any amplification?

 

[So]

You want a Grayhill 62P. http://www.grayhill.com/web1/i...es/Opt_Encoder_62P.pdf It's painless to implement and easy to program for -- just poll the inputs and if they change, given the last position, and the current position you know which direction it's clicked. 16 detents per turn and a (axial) shaft push button.

Edit: I think I sourced them from Digikey or Jameco when I last bought a few. In my experience, using a pot with an ADC or a normal rotary encoder will require a LOT More work. These things are more expensive than other options, and in a mass produced device, it'd be worth developing for a cheaper solution, but for a one off lab project they're cheap enough.

 

[Leros]

Originally posted by: So
You want a Grayhill 62P. http://www.grayhill.com/web1/i...es/Opt_Encoder_62P.pdf It's painless to implement and easy to program for -- just poll the inputs and if they change, given the last position, and the current position you know which direction it's clicked. 16 detents per turn and a (axial) shaft push button.

Edit: I think I sourced them from Digikey or Jameco when I last bought a few.

Edit 2: In my experience, using a pot with an ADC or a normal rotary encoder will require a LOT More work.

So, I would need to makes sure I poll fast enough to get the next position before it moves more than 4 positions? Seems fairly easy to handle.

 

[So]

Originally posted by: Leros

Originally posted by: So
You want a Grayhill 62P. http://www.grayhill.com/web1/i...es/Opt_Encoder_62P.pdf It's painless to implement and easy to program for -- just poll the inputs and if they change, given the last position, and the current position you know which direction it's clicked. 16 detents per turn and a (axial) shaft push button.

Edit: I think I sourced them from Digikey or Jameco when I last bought a few.

Edit 2: In my experience, using a pot with an ADC or a normal rotary encoder will require a LOT More work.

So, I would need to makes sure I poll fast enough to get the next position before it moves more than 4 positions? Seems fairly easy to handle.

Yeah, you'd be surprised how slow a person is at turning a knob.

 

[Leros]

Originally posted by: So

Originally posted by: Leros

Originally posted by: So
You want a Grayhill 62P. http://www.grayhill.com/web1/i...es/Opt_Encoder_62P.pdf It's painless to implement and easy to program for -- just poll the inputs and if they change, given the last position, and the current position you know which direction it's clicked. 16 detents per turn and a (axial) shaft push button.

Edit: I think I sourced them from Digikey or Jameco when I last bought a few.

Edit 2: In my experience, using a pot with an ADC or a normal rotary encoder will require a LOT More work.

So, I would need to makes sure I poll fast enough to get the next position before it moves more than 4 positions? Seems fairly easy to handle.

Yeah, you'd be surprised how slow a person is at turning a knob.

So is there something special about the 62P or would any optical encoder do the trick?

Also, whats wrong with a normal rotary encoder?

 

[So]

Originally posted by: Leros

Originally posted by: So

Originally posted by: Leros

Originally posted by: So
You want a Grayhill 62P. http://www.grayhill.com/web1/i...es/Opt_Encoder_62P.pdf It's painless to implement and easy to program for -- just poll the inputs and if they change, given the last position, and the current position you know which direction it's clicked. 16 detents per turn and a (axial) shaft push button.

Edit: I think I sourced them from Digikey or Jameco when I last bought a few.

Edit 2: In my experience, using a pot with an ADC or a normal rotary encoder will require a LOT More work.

So, I would need to makes sure I poll fast enough to get the next position before it moves more than 4 positions? Seems fairly easy to handle.

Yeah, you'd be surprised how slow a person is at turning a knob.

So is there something special about the 62P or would any optical encoder do the trick?

Also, whats wrong with a normal rotary decoder?

I've only gotten the grayhill to work. IIRC, the grayhill is an absolute encoder and most others are relative ones. I couldn't get the one I was playing with to work right, but I didn't spend much time playing with it.

 

[FrustratedUser]

Use a counter triggered by an optical interupter on the wheel.

 

[chorb]

Odd, I made a wireless joystick that controlled an etchesketch for my senior design project. The joystick was a 2 axis pot (both could be used at the same time) which allowed for better diagonal drawing. We then had fine pitched stepper motors attached to the knobs.

Someone else in our class did something with an etchesketch too, wonder what the appeal is...

 

[Leros]

Originally posted by: chorb
Odd, I made a wireless joystick that controlled an etchesketch for my senior design project. The joystick was a 2 axis pot (both could be used at the same time) which allowed for better diagonal drawing. We then had fine pitched stepper motors attached to the knobs.

Someone else in our class did something with an etchesketch too, wonder what the appeal is...

This isn't senior design. This is just a microprocessor interfacing class. They wouldn't allow us to do something anywhere near this simple for senior design. This is just a week or two long project for fun at the end of the semester.

We had simple options of doing something like 1-d pong (with a row of LEDs), interfacing a stepper motor, interfacing a speaker to produce a little piano or we can come up with our own idea.


My thought for the etch-a-sketch was:
- simple
- fits within budget
- graphical display
- fun to pick up and play during demo
- expandable for future development

 

[FrustratedUser]

Make it run Windows.