Hi
I am trying to use PGA870 with ±2.5V supply to drive an input of ADC.
PGA870 input is DC-coupled to output of op-amp with low output impedance.
It would be of great help if you could provide more information on how to correctly connect a DC-coupled input to PGA870 for Single-Ended to Differential Operation with ±2.5V supply?
In the datasheet there is a note that in this case the non-driven input pin of the PGA870 should be ac-coupled to ground through a capacitor, but is this correct?
Should this capcaitor and bypassing capacitors for Vmid1, Vmid2 then be connected to ground or to -2,5V?
I have also tried to simulate this design in TINA-TI, but there was no output from PGA870.
Thank you,
Dejan P.
Dejan P.,
For +/- 2.5V split supply operation, all bias voltages need to be shifted by -2.5V as stated in the apps section. For DC coupling, tie the undriven input to GND or 0V, this is now mid-supply. If you use cap to GND, the gain will not be correct at DC.
Vmid2 should be bypassed with capacitor to GND as with single supply. This may seem odd since is should be GND, but Vmid2 is an active ouput; so use cap. Vmid1 is the midpoint of internal resistor divider between the power rails and also should be GND. You can tie it to GND, or bypass with capacitor to GND as with single supply. Maybe just as easy to bypass both with capacitors to GND as with single supply.
I have tested this in the lab and know it works. But you are right the Tina simulation does not work right. There must be voltage levels inside the macro that do not shift in relation to the external supply pins.
Regards,
Jim Karki
TI High Speed Amplifiers
Hi Jim Karki,is it possible to shift supply between +3.5V and -1.5V to perform a common-mode level near 1V for DC coupling without using a resistor network? In this case the mid-supply is now 1V, can I tie the undriven input to this value or to GND?should Vmid1 and Vmid2 be bypassed with capacitor to GND?Thank youLuca Toscano