John De Villiers wrote: > My genius netmouse pro had only one pickup on each encoder wheel. I > took out the encoder when and moved the tip of a pen through the gap > and it still picked up motion in the correct direction. Seems like the > transistor they're picking the data up from is not all that standard. Tricked you! Is it standard? Yes, it's a "standard mouse transistor". You perhaps assume that because it has three pins, it's a common phototransistor. Again, it *is* common because mice use them, and mice are *really* common, but it is in fact an assembly/ package incorporating *two* phototransistors with the centre pin common to both emitters, and the outer pins, a collector each, no (external) connections to the base as such is rarely required to be made. (It is sometimes appropriate to have a resistor from base to emitter of a phototransistor to produce a minimum light threshold.) Peter Cousens wrote: > It does not require two complete sensors, one led and transistors or > one transistor and two (modulated) leds Indeed, one LED and two transistors, but the two transistors are in the one encapsulation. Multiplexing two LEDs is not particularly practical. Why? Well, it slows down the capture process, requires a faster phototransistor and worst of all, requires switching transistors in the LED lines. That isn't at all practical in mice which use one "handshake" line to power the (Holtek or Zilog, sometimes PIC) controller, and another to energise the LEDs in a series chain. Sure, PS/2 mice use a single 5V supply. Which is nice as you can energise exactly two LEDs in series, one for each sensor. Consider how much you would gain from multiplexing LEDs instead of phototransistors. Two sensor assemblies: Cointinuous LEDs, two P/Ts per assembly; four inputs. Two LEDs each on 2 sensors, one P/T each; two outputs for LED drive, two inputs = same number of I/O. Minimal savings even on a "wheelie" mouse. Six inputs for three encoders plus three buttons. What *really* paeves me on my "wheelie" mice though - the wheel on the "Genuine Micro$oft" one uses a crappy mechanical encoder while the "clone" one uses a proper optical encoder. What does that tell you? -- Cheers, Paul B.