PGA2500 CMRR data

Hi, tg,

Let's see what we can find for you... 

-d2

Hi,

The PGA2500 inputs are meant to be driven using a floating source.  A standard CMRR test driving the inputs with a grounded source will not work well.

Both VIN+ and VIN- are biased approximatlely 0.65V below the common-mode voltage VcomIN and the use of DC coupled capacitors is highly recommended for the analog inputs of the PGA2500.

Best Regards,

Luis

Luis

   Thank you for taking the time to reply.  However, your answer did not address my original question.  Let me break down the question further just to be clear on what I am asking:

1.   Is there any CMRR data that you can share, for any mode of operation?

2. How much CMRR degradation (if any) can be expected as a result of grounding VcomIN?

3. Figure 8 in the datasheet clearly shows an example with VcomIN grounded.  Do we have any CMRR data for that example?

In advance, thank you.

tg

Hello TG,

The typical mode of operation for the PGA2500 device is with supplies +VA=5V, -VA=-5V and VD=-5V; and with VcomIN=GND.  

Setting VcomIN=GND allows for the highest input and output voltage range; where the maximum output swing is 0.9V from the rails.  Therefore, connecting the VcomIN  to ground will not cause any degradation in the performance of the device, and will allow for the full output voltage range.  Most of the performance specifications given in the electrical table and plots provided in the datasheet show VCOM=0V.  The device can support up to 3.5Vrms output signal when VcomIN=GND without an increase in distortion.  The typical plots for THD+N vs Frequency with VcomIN=0V and signals Vout=3.5 Vrms and 4.0Vrms are shown at the bottom of page 5.

In applications where the PGA2500 is driving audio ADC inputs, and the output common mode signal needs to be shifted , the VcomIN could be set to 2.5V, in order to set the optimal output common-mode (at the middle of the supplies of the ADC).  However, in this case, the output range of the device is limited to approximately 1.0Vrms; above this voltage the distortion increases.  The plot at the bottom of the datasheet on page 5 shows the THD+N distortion with VcomIN=2.5V, VOUT=1Vrms and VOUT=2Vrms for different gains.    

In this device, the inputs are intended to be self-biased by the internal circuitry; therefore, a standard DC CMRR test would not work well.  A possible test could be performed by driving the VcomIN pin; or by driving the inputs with an AC common-mode signal.  The device also incorporates a common-mode servo function that can be enabled and disabled.  The common-mode servo provides negative feedback that results in very low common-mode impedance.  This function improves the low and medium frequency common mode rejection in the typical application where the input source is floating. 

According to notes in the documentation, the measured typical performance for CMRR is around 20uV/V; and the the common-mode servo loop can improve CMRR significantly when the source is floating.  I will have to dig a little more to find more data or perform a bench measurement to provide a typical result .

Best ,

Luis 

Luis

   Thank you for taking the time to perform the test.  Your help is greatly appreciated!

tg