Hello,
I have a design in the field that we're experiencing a number of issues with. As I've been investigating potential issues with the design, I've noticed that when the power supplies for the TMS320F28035 are switched off, the supply rails, digital and analog 3.3V appear to be hovering around 0.4-0.6 V. I suspect that the supply rails are being held to 0.4-0.6V by ADC inputs that are likely conducting through the clamping diodes of the TMS320F28035. By design, series resistance limits the maximum DC current into any of these ADC pins to be less than 110 uA. I noticed that here's ~ 5.7 nF (nominal) of MLCC capacitance on these pins, and if the ADC pins voltage were say 1.7V while the VDDA supply source was being powered down, then as VDDA potential dropped below the V_F of the clamping diode, current would conduct from the analog input pin, through the diode, and onto the analog supply rail, with very little series resistance to limit the current. However the VDDA and VDDIO supplies both have large capacitors (severs uF) by comparison to the 5.7 nF filter caps on the ADC input pins, so I'd expect both the VDDA rail and the analog input pins to decrease in unison, with the analog input pin offset by the forward voltage drop of the clamping diode, until a steady state is reached where the small (<100 uA) current per analog input pin is flowing through the clamping diode onto the VDDA/VDDIO rails and through any impedance on those rails into ground. If my understanding is correct, then I don't believe the design is violating the abs max of +/- 20 mA of clamping current, and note 3 in the abs max seciton of the datasheet states: "Continuous clamp current per pin is ±2 mA. Do not operate in this condition continuously as VDDIO/VDDA voltage may internally rise and impact other electrical specifications.". Again, I do not think the design is violating the +/- 2mA per pin limit, however, it is continuously flowing current through the clamp diodes, and the VDDIO/VDDA supply voltages are rising to 0.4-0.6V. Is the scenario I'm describing something that we should be concerned could damage the TMS320 part? I've included a simplified schematic diagram for illustration, the scenario is when SW1 is open, the A3V3 and D3V3 rails remain around 0.4-0.6V, I believe due to current flowing from ADC pins with structures similar to that shown below.
Thanks for your support!
