Hi All, I'm working on a project where I need to use an alternative to conventional rotary potentiometers. The reason for this is that the position of each pot or should say control needs to be saved and recalled. This is further complicated because the feel of the system must be as it would be with a conventional pot with a rotation of 270 degrees, the client wants to sweep from one end of the range to the other in less than one turn.

My plan thus far has been to use a rotary encoder with either a ring of 21 leds around it to indicate the position or 3 x 7 segment displays above to provide a visual indication of its position – or possibly both. This side of things is relatively straightforward.

The problem has been that I have not been able to find an affordable encoder with a resolution higher than 24 pulse per rev, which means if I want to keep the feel of the control the same as a regular pot (270 Degrees) I would only have 18 pulses to play with which is too coarse. To get more means having to multi turn the control which feels wrong to the user.

So can anyone point me in the direction of a high resolution (around 340 pulses per rev) affordable rotary encoder? Or possibly affordable continuous potentiometer (I know this will require some coding to handle the point where it drops off one end on to the other)?

I have also considered a motor pot but again these were discounted because of cost, size and additional drive circuitry required.

Any other suggestions?

There will be 10 to 12 of these type of controls in each unit so it is very cost sensitive.

 

How about a 100p/rev quadrature encoder for $3.00?
Ebay 112030673020. It is quadrature so you would need to count x2 or x4 edges to increase the res. to 200p/rev or 400p/rev other than the native 100p/rev.
Or you can simply make your own with a slot opto.
Max.

 

I had considered making my own and having a disc laser cut with enough slots... may still do...

 

With the ebay one I posted you could just use one of the opto's if not needing x2/x4 count, with a incremental encoder, the home position usually has to be establish some how at power up.
Max.

 

I had considered making my own and having a disc laser cut with enough slots... may still do...

Chemically milled stainless steel is the way to go. Here are a few tips I can pass on after designing encoders into a few projects:

• Through-looking optos are easier to get right than retro-reflective ones.
• Precision counts. Eccentric runout will affect the 90degree relationship between the two phases. Shaft wobble will affect the signal quality of reflective optos as the spot moves in and out of focus on the wheel.
• Higher resolutions like 100PPrev require significant precision. Sketch a wheel out, do the math to see the slot dimensions then consider how little tolerance you have to work with.
  
• Signal conditioning is a must. Schmitt trigger inputs at the least but preferably comparators. Microcontroller inputs (PICs anyway) don't always flip at the same voltage, even the schmitts. Given the slow rise and fall times of the opto output, the flipping of the input at different voltages results in phase jitter.
• The optos don't have to be adjacent as long as the angular difference between their locations incorporates a 1/4 slot-angle displacement. Concentricity issues are mitigated when the optos approach 0 or 180deg apart and are aggravated on 90/270deg.
• Consider a dual opto package if you can get one for your resolution.
  
• Beware the plastics guys that tell you that they can replace that milled disk with a molded or printed part with equal precision or that a simple shaft to snap the wheel onto is as good as a bearing.
• If you do go the plastic wheel route, be sure to know the optical properties of the material in the infra-red spectrum. Materials that look opaque and reflective in visible light can be remarkably transparent or absorbent in the IR. Ask me how I know these things.
• Firmware using interrupt on change and an is-was lookup table is a good way to decode.
  
• Lower resolutions can be helped by firmware that incorporates 'ballistics' i.e. determine how fast the shaft is being rotated by noting the time between changes i.e. the faster the turn, the bigger the parameter change.

Hope this helps with your decision. Good luck!

 

So can anyone point me in the direction of a high resolution (around 340 pulses per rev) affordable rotary encoder?
.

Once you get into this resolution (340p) using slot count, it usually requires the specialized motion encoder that allows reading a photo etched glass with the added reading head (Moiré) grating in order to read extremely high resolutions.
These are typically quadrature style encoders.
Using the 100p/rev quadrature x4 configuration enables up to around 400p/rev from ordinary slot type opto's.
Max.

 

The problem has been that I have not been able to find an affordable encoder with a resolution higher than 24 pulse per rev,

What exactly is how you define affordable? You can have whatever you wish as to pulses per revolution. It will go back to your definition of affordable or inexpensive. US companies like Anaheim Automation market a line as well as Baldwin and as to low pulse counts like you mention there are off the boat and encoders like Max suggested.

Ron

 

It seems that you need a pot ganged with an encoder. A possibility is a 2 gang pot with a well regulated supply V. Record V of each pot with switchable V meter. I'm not sure how the 2 pots are attached, but might remove back pot & replace it with encoder ?
Could also display V with cascaded LM3914's