TLV1562

2.7 V TO 5.5 V, HIGH-SPEED LOW-POWER RECONFIGURABLE ANALOG-TO-DIGITAL

CONVERTER WITH 4-INPUT, DUAL S/H, PARALLEL INTERFACE, AND POWER DOWN

SLAS162 – SEPTEMBER 1998

21

POST OFFICE BOX 655303

Full-Scale Error

(Before Calibrayion)

Ideal Transfer Function

Transfer Function Before

Mid-Scale Calibration

Mid-Scale Error

(Before Calibration)

Mid-Scale Error

(After Calibration)

Analog Input

REFP

V-Mid (REFCM)

Zero-Scale Error

(Before Calibration)

Zero-Scale Error

(After Calibration)

Full-Scale Error

(After Calibration)

Transfer Function After

Mid-Scale Calibration

REFM

Zero Scale

Mid Scale

Full Scale

Code Output

Figure 16. Mid-Scale Error Calibration

resume normal conversion from mid-scale error calibration modes

A follow on write operation sets CR0.(7,6) to 00 which resumes the normal conversion mode. Typically

mid-scale error calibration needs to be performed only once after power up. If however the operation mode is

changed from single ended to differential, then preferably mid-scale error calibration should be performed

again.

The user writes a bit to enable mid-scale error calibration. Inputs to the ADC are internally shorted therefore

the offset value can be converted to a digital word. The result (a digital word representing the offset) is stored

in a latch. This offset value is then subtracted from the digital output of all conversions except when in mid-scale

error calibration mode.

system design consideration regarding to mid-scale error calibration

Mid-scale error calibration may limit the dynamic range of the ADC. If the offset is negative and has a magnitude

x, then the range of the converter codes is x to 1023. If the offset is positive and has a magnitude of x, then the

range of converter codes is 1 to 1023 –x. Thus the ADCs dynamic range is reduced by x (say x = 20 codes) on

either side of the range effectively with mid-scale error calibration. However this should not be a limitation for

most users.