4 Input Common-Mode Distortion

The following plots demonstrate the CMRR and THD+N performance with different input levels in each of the tolerance modes. These curves are measured on an ADC5120EVM-PDK board configured for AC coupled inputs using an AP2700 and a precision function generator. A 1-kHz, -60-dBFS differential input tone is applied from the Audio Precision and a 750-Hz common-mode tone is applied to the reference node of the AP2700. The -60-dBFS differential tone is not large enough to produce its own distortion products. The 750-Hz common-mode is used to easily differentiate any harmonics resulting from the common-mode variation.

When referring to the figures it is helpful to keep in mind the relationship between output dBFS and the input voltage. The following is an example calculation:

2-Vrms = 2.8-Vpp

100-mVpp = 0.05-Vp

100-mVpp in dBFS = 20*log(0.05/2.8) = -35dBFS

Figure 4-1 shows the output spectrum in mode 0 with a 100-mVpp (-35-dBFS) common-mode tone applied to the input. From this plot we can observe that no significant distortion occurs and the 750-Hz tone is reduced by around 55dB (-35dB - (-90dB) = 55dB). The measured THD+N level for this test was -93-dBFS.

Figure 4-2 shows the output spectrum in mode 0 with the input common-mode amplitude increased to 1Vpp (-15-dBFS). Note that this is well outside the maximum common-mode tolerance of mode 0 and significant distortion can be seen as expected. The primary 750-Hz tone is at approximately -70dB, so the CMRR is still around 55dB (-15 - (-70) = 55dB), but the THD+N level has degraded to approximately -73-dBFS.

Figure 4-3 shows the same 1-Vpp common-mode signal applied to the input in mode 1. Mode 1 extends the supported common-mode range, slightly improves the common-mode rejection of the fundamental 750Hz tone, and significantly reduces the distortion introduced from the large common-mode signal.

Figure 4-4 again shows the same 1-Vpp common-mode signal applied to the input in mode 2. Mode 2 further extends the common-mode range, improves the common-mode rejection, and eliminates the distortion produced from the large input common-mode tone.