[FAQ] PGA900: Questions ADC Related to the PGA900, PGA300, PGA302 and PGA305 Devices

Bob Benjamin said:

Can I float the ADC inputs?

https://e2e.ti.com/support/sensors-group/sensors/f/sensors-forum/1226647/pga305-inp-inp--is-not-used-also-floating-ok-and-adc-sampling-rate?

Bob Benjamin said:

What is the ADC sampling rate?

https://e2e.ti.com/support/sensors-group/sensors/f/sensors-forum/1226647/pga305-inp-inp--is-not-used-also-floating-ok-and-adc-sampling-rate?

Bob Benjamin said:

What temperature accuracy can I expect and what about drift?

https://e2e.ti.com/support/sensors-group/sensors/f/sensors-forum/842849/pga305-unable-to-temperature-compensate/3129231?

Bob Benjamin said:

What is the relationship to TGAIN and TADC_GAIN, or PGAIN and PADC_GAIN?

TADC_GAIN and PADC_GAIN refer to gain applied at the PGA prior to the ADC.  TGAIN and PGAIN are digital gain that are applied after conversion result is posted.  It is always best to limit the digital gain to 1 (or nearly 1) so that all gain takes place prior to ADC.  This becomes more important when linearizing and temperature compensating the result.  It is possible to utilize the OFFEN bit for high offset sensors so that the compensation and computation include the digital gain times the result of the ADC conversion result and the offset.  Otherwise the digital gain is applied to just the ADC result with the offset added later.  The result becomes the same if the digital gain is 1 regardless if you use OFFEN or not.

Bob Benjamin said:

For the PGA305, should I use 24-bit or 16-bit and why?

Internally, the PGA305 is set up to always run in 24-bit mode, so the PADC and TADC raw  values will always be 24 bits. When the compensation engine is running, the 24 bit values are passed to the compensation algorithm and all of the compensation is done at 24 bits. There are 2 output methods: one is analog where the output is passed to a 14 bit DAC, and one is digital where the output can be either 16 bit or 24 bit through I2C. If 16 bit mode is selected, then only the 16 MSBs of the compensated output are passed to the I2C output. The analog output is always truncated to 14 bits because of the DAC’s resolution limitation.

 If the customer places the PGA305 in digital interface mode, they will only be able to read 24 bit data from the PADC and TADC data registers regardless of the mode that it is placed in. If the device is placed into compensation mode and 16 bit mode is selected, then the output on I2C (see PGA305 I2C Runtime Commands in the datasheet) will be 16 bits. Note that the source values will always be 24 bits, but the compensated values will be 16 bits if the device is placed in 16 bit mode.

Bob Benjamin said:

Why is my internal temperature sensor value so much different from the ambient temperature?

The internal temperature is actually the IC die temperature and not the outside the package temperature.  So, if the supply voltage powering the PGA is large, then the voltage drop from the supply source to the output of the AVDD and DVDD LDO will have a thermal affect which will cause the die temperature to increase.  So, supply voltages closer to the LDO output voltage (5V for example) will have a much more accurate internal temperature with respect to the ambient than a voltage closer to the maximum (24-30V).