User`s guide

Elan Digital Systems Ltd. 6 AD125 USER’S GUIDE
2. ABOUT THE AD125
2.1 QUICK THEORY OF SUCCESSIVE APPROXIMATION
CONVERTERS
The type of converter used in the AD125 approximates the analogue
level being applied to its input using a D to A converter and a
comparator. The converter starts in “track” mode where it is
following the input voltage and applying it to a track and hold
amplifier. Once the converter is told to perform a conversion, it
holds the current input voltage level on a capacitor while it
approximates its value.
The converter uses a clock to break the approximation process down
into 12 steps, one per bit. Each step attempts to approximate the
held input voltage to one more “bit” of resolution. The logic in the
A to D makes the most significant bit decision first as this is the
most “coarse” level, i.e. is the signal positive or negative.
Subsequent decisions are then made on the difference between the
output of the internal D to A converter and the held input value: if
the comparison is “greater” then the bit is set, if “less” the bit is
cleared. After 12 clocks the complete word is ready to be read out of
the converter.
2.2 NOISE
Noise in an A to D converter system will degrade the “effective
resolution” of the conversion. The noise can be power supply noise,
thermal noise, pick-up noise etc. All contribute to the degradation.
The Effective Number Of Bits for a converter expresses a measure
of the noise level relative to the input signal level.
The ENOB of a converter is expressed as:
ENOB = (SNR(dB) - 1.76) / 6.02
So, the better the signal to noise ratio the higher the effective
resolution. Be warned however, that this computation is based on
RMS noise. Taking individual samples from the AD125 will reveal