As discussed previously (e2e.ti.com/.../913234), significant noise is observed across the PGA capacitor for the ADS1292 when the chip is reset by a START command (via START pin or SPI command). This makes sampling at rates that do not correspond to rates achievable by continuous mode of operation and which require sampling synchronization between more than one ADS1292 chip extremely difficult when detection of pacemaker pulses is required as the START reset spikes are large and "in-band" in terms of frequency content.
We had a solution to this, which basically amounted to increasing the pacing pulse amplitude relative to the START noise. However, recent builds have revealed that some ADS1292 chips are much "noisier" than others. We have recently observed chips which are nominally in identical circuit boards, running identical firmware, that exhibit START spike noise across the PGA capacitor with 5x the amplitude and our previous solution of reducing system gain for pacing detection is no longer effective on these "noisy" chips.
As TI has developed several reference designs for pacing pulse detection with the ADS1292 chip, I'm assuming this has to have been observed. Even in the absence of additional pacing circuitry, measurement of the signal across the PGA capacitor during START operation will show the spikes. How can we address this?