We present a programmable continuous-time floating-gate Fourier processor t
hat decomposes the incoming signal into frequency bands by analog bandpass
filters, multiplies each channel by a nonvolitile weight, and then recombin
es the Frequency channels. A digital signal processor would take a similar
approach of computing a fast Fourier transform (FFT), multiplying the frequ
ency components by a weight and then computing an inverse FFT. We decompose
the frequency bands of the incoming signal using the transistor-only versi
on of the autozeroing floating-gate amplifier (AFGA), also termed the capac
itively coupled current conveyer (C-4). Each band decomposition is then fed
through a floating-gate multiplier to perform the band weighting, Finally,
the multiplier outputs are summed using Kirchoff current law to give a ban
d-weighted output of the original signal. We examine many options to reduce
second-order harmonic problems inherent in the single-sided C4. We present
a method for programming arrays of floating-gate devices that are used in
the weighting of the bands. All of these pieces fit together to form an ele
gant and systematic Fourier processor.