Sujet DAC 16+3bits du Korg Wavestation

Deciphering the mysteries of the famous #PublisonDHM89B2 (cont’)
• • • • • • •
Just finished a first redraw of the DHM’s so-called “Flying comma quasi-16-bit A/D Converter”.
Never seen such a complex stereo, discrete, successive approximation analog-to-digital converter from the late seventies era...
...and that’s probably why they called it a “quasi-16-bit” ADC: the rather unusually complex analog part of the converter (left and center sections of the first picture) contains no less than 10 op-amps, 2 comparators, and 14 transmission gates (and I’m not even couting in all the passive components!) which, together with the DAC, SAR and other logic gate chips (at the far right of the image) produce a digital representation for each sample of the audio signal on 12 bits only.
At that time memory ICs were much expensive, even for small capacity chips. So engineers had to find a way to fit high quality digital audio into the less possible number of bits. Not to mention that integrated 16-bit A/D converters were almost non-existent on the 1977 market, or came at too high a price tag for a very small company as Publison .
Then Mr. Philippe Petitdemange, the genius who designed the #DHM89B2 , came up with that complex way of combining two channels sampling into one circuit, as well as designing his own way of using the “gain ranging technique” to achieve a quasi-16-bit dynamic range with 12-bit coding.
It has to be mentioned that, in 1977-1979, VERY few pro digital audio manufacturers used 50 kHz sampling rate, quasi-16-bit converters... as can be found inside the DHM89.

FS Gain Error (DACs)
The full-scale gain error of an digital-to-analog converter (DAC) is the difference between the actual and the ideal output span. The actual span is determined by the output when all inputs are set to 1s, minus the output when all inputs are set to 0s. The full-scale gain error of any data converter can be affected by the choice of reference used to measure the gain error.

Gain Error
The gain error of an ADC or DAC indicates how well the slope of an actual transfer function matches the slope of the ideal transfer function. Gain error is usually expressed in LSB or as a percent of full-scale range (%FSR), and it can be calibrated out with hardware or in software. Gain error is the full-scale error minus the offset error.