A voltage quantizer takes a continuously variable input voltage and outputs a stepped (or quantized) voltage. Whereas the input voltage might ramp up from 0 to 1V continuously the output voltage of the quantizer will be discrete steps of .083V. In normal mode this quantizer always starts at digital 1 or .083V. As the input voltage starts to ramp from 0 up to 1 volt the output will step to .1666V, .2499, .3332, etc up to 1.0V. The circuit also features a whole step mode in which the in to out conversion is quantizd to steps of .1666V. It also features a fourth step mode in which the in to out conversion is quantizd to steps of .3332V. Useful with sample and hold circuits to bring voltages into musical scale steps (half, whole or fourths). Useful with sequencers to bring voltages into musical scale steps (half, whole or fourths).
In the previous mp3 samples the output of the quantizer is driving a VCO. These samples are dry. In the Ramp wave (half, whole, fourths) you hear the effect of quantizing a ramp wave. The Ramp wave's frequency is varied so you can hear the what happens when the frequency exceeds the ability of the quantizer to capture it. The input to the quantizer should be between 0 and +7V for the best quantization. Negative voltage will not damage the quantizer in any way but the output will be 0V as the quantizer does not handle negative voltage. Voltages above the recommended maximum (7V) will result in the maximum output from the quantizer (.08333 x 63)V.
Minor Performance Improvement (Integrated into documentation)
I have eliminated comparator speed-up cap C13 (10pF ceramic cap). I found that without the cap their is less chance for output jitter. This change makes the A to D convertor used in the front end of the quantizer more stable.
I have increased the value of output voltage glitch supressor C18 to .002uF. This results in less digital noise on the output of the quantizer.
I soldered a .01uF cap across the feedback resistor R37. This reduces the digital noise appearing on the output of U5-C and reduces jitter since the comparator is presented with a cleaner voltage.
R1 must be a 200K resistor. Make sure the clock frequency is between 40KHZ and 50KHZ. Clock rates above 50KHZ will increase jitter.
Lastly for greatest accuracy. I ordered 100 each of the 100K and 200K metal film resistors and then used an accurate bench meter to further sort them to find resistors that were within 0.1% of 100K and 200K respectively and then used those in the A to D and D to A R-2R circuit sections. By doing so I got excellent accuracy in the output voltage.
With these changes I am very pleased with the operation of this circuit. All four are operating quite nicely.
Specs & Downloads
Specs & Downloads