Abstract: A method and apparatus for converting a plurality of digital signals to an analog signal and m-bit digital-to-analog converter having an LSB current switch. A most significant m bits of an m bit digital signal are applied to m bits, respectively, of the m bit digital-to-analog converter, and successive values of the n bit digital signal are substituted at a first frequency. An interpolation signal is produced at the first frequency. A least significant n-m bits of the n bit digital signal are applied to the interpolation circuit to cause it to interpolate between successive values of the n bit digital signal. An LSB current switch distinct from the m bit digital-to-analog converter produces an output signal in response to the interpolation signal. An analog output signal produced by the m bit digital-to-analog converter is combined with the output signal produced by the LSB current switch to produce an analog output signal with n bit resolution.

Abstract: In view of the limited speed times accuracy product of A-to-D and D-to-A converters, a required digital accuracy or sufficiently smooth analogue waveform cannot always be obtained. The present invention is useful in enhancing available converters. In the A-to-D version a stepped interpolation waveform obtained from a generator 13 is added to samples of an analogue waveform obtained from a sample-and-hold circuit 11 at a frequency 2f.sub.B (where f.sub.B is the bandwidth of the input signal). The resultant signal is applied to an n-bit A-to-D converter 15 operating at 2.sup.m times 2f.sub.B and the converter output is averaged using an accumulator 16 reset at 2f.sub.B. The accumulator output is a higher accuracy signal since it has m+n bits. An analogous technique is described for D-to-A conversion.

Abstract: In view of the limited speed times accuracy product of A-to-D and D-to-A converters, a required digital accuracy or sufficiently smooth analogue waveform cannot always be obtained. The present invention is useful in enhancing available converters. In the A-to-D version a stepped interpolation waveform obtained from a generator 13 is added to samples of an analogue waveform obtained from a sample-and-hold circuit 11 at a frequency 2f.sub.B (where f.sub.B is the bandwidth of the input signal). The resultant signal is applied to an n-bit A-to-D converter 15 operating at 2.sup.m times 2f.sub.B and the converter output is averaged using an accumulator 16 reset at 2f.sub.B. The accumulator output is a higher accuracy signal since it has m+n bits. An analogous technique is described for D-to-A conversion.

Abstract: A method and apparatus for converting a plurality of digital signals to an analog signal and m-bit digital-to-analog converter having an LSB current switch. A most significant m bits of an .[.m.]. .Iadd.n .Iaddend.bit digital signal are applied to m bits, respectively, of the m bit digital-to-analog converter, and successive values of the n bit digital signal are substituted at a first frequency. An interpolation signal is produced at the first frequency. A least significant n-m bits of the n bit digital signal are applied to the interpolation circuit to cause it to interpolate between successive values of the n bit digital signal. An LSB current switch distinct from the m bit digital-to-analog converter produces an output signal in response to the interpolation signal. An analog output signal produced by the m bit digital-to-analog converter is combined with the output signal produced by the LSB current switch to produce an analog output signal with n bit resolution.