Companding analog to digital converter

Analog to digital converter means for generating a variable bit length digital output comprising:

• means for receiving an analog signal;
• adjustable means for selectively attentuating the signal level of the analog signal in said receiving means;
• signal converter means for converting the attentuated analog signal into an N-bit binary word representing said attentuated signal level;
• first storing means for storing the M most significant bits of said binary word;
• decoder means responsive to selected bits of said binary word for adjusting said attentuating means to alter the attentuation of the analog signal in said receiving means as a function of the value of the selected bits stored in said first storing means;
• means for causing said first converter means to convert the analog signal level now presented by said adjustable means into an N-bit binary word;
• second storing means for receiving and storing said N-bit binary word from said signal converter means, whereby the contents of said first and second storing means are utilized to form a binary word representing the analog signal in said receiving means;
• an output line;
• means for shifting the contents of said first storing means to said output line in serial fashion;
• means for transferring the contents of said second storing means to said first storing means in serial fashion.

A method for converting an analog signal, whose magnitude lies within a predetermined range, into a variable length digital signal comprising the steps of:

• temporarily storing said analog signal;
• attentuating said analog signal a predetermined amount so that its magnitude is reduced to a value near the low end of said range;
• converting the attentuated signal into a multi-bit binary word representing the magnitude of said attentuated signal;
• storing a portion of said multi-bit binary word;
• examining said stored portion and altering the amount of attentuation of said stored analog signal in accordance with the value of the stored portion of the multi-bit binary word;
• converting the analog signal whose magnitude has been so altered into a second multi-bit binary word;
• combining one or more of said bits initially stored with said second multi-bit binary word to form a multi-bit binary word whose resolution (i.e. word length) is a function of the magnitude of the stored analog signal wherein the smaller the magnitude of the stored analog signal, the greater the resolution, (i.e. length) of its multi-bit binary word.