Hi,
I am using a C5515 DSP SAR to measure the voltage of a battery. I have tested the A/D sampled data (with SAR at 2MHz clock rate) to ensure it corresponds to the actual input voltage, and I get a fairly linear correspondence, except for a DC offset of about 380mV in the A/D sampled data. This offset persists even when I short the SAR input to ground momentarily. I do not understand why there should be this offset included in the measurement as I use the VDD_ANA as the SAR reference (1.3Volts). On another PCB with a C5515 this offset is larger, so it’s not entirely consistent from PCB to PCB but the offset persists. I am not expecting such a large difference between the A/D digitized values and the actual SAR voltage input as measure with a voltmeter - a few mVs OK, but NOT 100s of mVs.
I am not using ANA_LDOO to power the VDD_ANA, however the VDD_ANA is powered directly by an external 1.3V regulator which also powers the CPU core and the PLL etc. VSS_ANA and the Band Gap grounds are coupled to digital ground through a ferrite bead, and the 1.3V supply supplies the VDD_ANA through a ferrite bead to minimize noise etc.
I would appreciate some suggestions for what this offset may be due and if it is a SAR configuration problem or a supply regulator problem, or some kind of leakage offset on the SAR input etc ? Many thanks in advance !
Regards, M