TMS320F28069: ADC output leakage current

Sam,

I  noticed the spec for input leakage is listed as +/-2uA, which to me is indicating that it could either output or input even though the parameter is defined as input.

Can you provide information on which pins show the output leakage vs input?  Is it always the same pins?  I'm wondering if it is the A2/A4/A6 B2/B4/B6 pins that have IO functions muxed and we can disable that. 

Best,

Matthew

Matt,

I also notice the AIO pins, but the output current is also on ADC only pins. Actually this should not be a problem when the source signal resistance is small, but right now we have an external 30K ohm resistance in the ADC pin, so the current will make big difference. If the leakage could also be output, then what's the range of the source resistance?

Thanks.

BR,

Sam  

Sam,

Thanks for the clarification, it looks like this is for all ADC pins then and doesn't have a dependence on the AIO function.  

The max we specify in the DS is the 2uA, so at 30kOhm we could see as much as 60uV or 82 LSBs of ADC code caused by this.  It looks like you are seeing 1uA so this would be ~40 codes.

I'm wondering, however, if when the ADC is sampled for a longer duration if we will see this leakage dissapate.  Can you comment on the value you are using for the ACQPS/aquisition window when sampling that channel?  Can you try to increase this value and see if you read the voltage from your system more accurately?

Another alternative, we could use the ADCOFFSET register/bit field to artifically correct for this skew, either globally or when you sample this channel.

Best,

Matthew

Matt,

Thanks for your help, we tried the ACQPS/aquisition window but not much help, +/- 1uA is in the datasheet range, so we will try to reduce the source resistor to get better result.

Thanks.