12.1 Layout Guidelines

TI recommends employing best design practices when laying out a printed-circuit board (PCB) for both analog and digital components. This recommendation generally means that the layout separates analog components [such as ADCs, amplifiers, references, digital-to-analog converters (DACs), and analog MUXs] from digital components [such as microcontrollers, complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), radio frequency (RF) transceivers, universal serial bus (USB) transceivers, and switching regulators]. An example of good component placement is shown in Figure 123. Although Figure 123 provides a good example of component placement, the best placement for each application is unique to the geometries, components, and PCB fabrication capabilities employed. That is, there is no single layout that is perfect for every design and careful consideration must always be used when designing with any analog component.

The following outlines some basic recommendations for the layout of the ADS1248 to get the best possible performance of the ADC. A good design can be ruined with a bad circuit layout.

• Separate analog and digital signals. To start, partition the board into analog and digital sections where the layout permits. Route digital lines away from analog lines. This prevents digital noise from coupling back into analog signals.
• The ground plane can be split into an analog plane (AGND) and digital plane (DGND), but this is not necessary. Place digital signals over the digital plane, and analog signals over the analog plane. As a final step in the layout, the split between the analog and digital grounds must be connected to together at the ADC.
• Fill void areas on signal layers with ground fill.
• Provide good ground return paths. Signal return currents will flow on the path of least impedance. If the ground plane is cut or has other traces that block the current from flowing right next to the signal trace, it will have to find another path to return to the source and complete the circuit. If it is forced into a larger path, it increases the chance that the signal will radiate. Sensitive signals will be more susceptible to EMI interference.
• Use bypass capacitors on supplies to reduce high frequency noise. Do not place vias between bypass capacitors and the active device. Placing the bypass capacitors on the same layer as close to the active device yields the best results.
• Consider the resistance and inductance of the routing. Often, traces for the inputs have resistances that react with the input bias current and cause an added error voltage. Reducing the loop area enclosed by the source signal and the return current will reduce the inductance in the path. Reducing the inductance will reduce the EMI pickup and reduce the high frequency impedance seen by the device.
• Watch for parasitic thermocouples in the layout. Dissimilar metals going from each analog input to the sensor may create a parasitic themocouple which can add an offset to the measurement. Differential inputs should be matched for both the inputs going to the measurement source.
• Analog inputs with differential connections must have a capacitor placed differentially across the inputs. Best input combinations for differential measurements use adjacent analog input lines such as AIN0, AIN1 and AIN2, AIN3. The differential capacitors must be of high quality. The best ceramic chip capacitors are C0G (NPO), which have stable properties and low noise characteristics.

12.2 Layout Example