Re: Re:Re:encoders

"Arne Chr. Jorgensen" wrote:

> Ted wrote a long mail on encoders, and the use of exclusive or
> gates. This is not that simple. You will get errors. I would like a
> schematic drawing of it before I pass judgment, - but I "believe"
> that you may accumulate errors - caused by mechanical jitter. To
> put it very simple: Think of a simple switch. Without a debounce
> circuit, you will get many pulses, - not just a clean on/off. If you
> turn the encoder back 1/2 pulse length, then turn it forward again
> 1/2 pulse length, - any digital readout will read this as 1 step
> forward, - but the position has not really changed. With very high
> resolution, you may not notice this error at all, - but it can build
> up.

Nope. You have to treat it as a state table. If you allow it to go
from any of the 4 possible states to any other state, then, yes, that
WILL get errors. But, if you force it to only go from one state
to the next correct state, it will ignore these motions. In other words,
if, say, the A output is high, and the B output is flickering back and
forth between high and low, it is going back and forth between two
adjacent states. One transition is motion in the + direction, and
the other transition is motion in the - direction, so counts DON'T
accumulate. It is just flickering back and forth between two
immediately adjacent coordinate counts.

Jon