I have a question on the way the AD7656 is specified in the Specification Table.

The DC Accuracy is specified as Positive Full Scale Error (±0.75% FSR max, ±0.22% FSR Typ) and Negative Full Scale Error (the same values).

When I use the device in a Bipolar mode at ±10V it looks like the ADC could have an error of ±150mV (0.75% x 20) at the limits of positive and negative 10 volts. That would end up as an error of ±1.5% at 10V. Am I looking at the specification correctly?

Thanks,

John

Hi John,

Firstly, from the datasheet terminology section, the Positive Full Scale Error is described as the deviation of the last code transition from the ideal ( +4Vref - 1 LSB for the +-10V range) after adjusting for the bipolar zero code error.

In terms of your calculation for the +-10V range this would be correct - however the Positive Full Scale Error you measure would be different for the +-5V range. This is due to the fact the reference is scaled differently.

So the +-5V range would have a maximum Positive Full Scale Error of +-75mV (0.75% x 10)

Regards

Niall