We are using an AD9826 with an FFT-CCD detector. All timing seems to be OK, but the data is unstable. Attached is waveforms of the 9826 timing and the "right" side of the 0.1uF capacitor. Any suggestions, anyone? I have tried all sorts of timing combinations and can't seem to get anything stable.
Yellow line is RG
Burgundy is ADCCLK
Green is CDSCLK2
Magenta is CDSCLK1
Cyan is the capacitor "right" side
Notice the spike when ADCCLK transitions from high to low.
Here are some other "right" side waveforms for different timing arrangements.
Last image is for CDSCLK1 falling at same time as ADCCLK falling