OPA192 output swing over temp / stability

Hello Ue,

1. I will address your capacitive drive question first:

I ran some simulations and found that the OPA192 is unstable at the given load. This instability can be compensated by adding a 10 Ohm resistor and a 1uF Cap in parallel with the load (as shown in the schematic). I have attached the TINA circuits for the OPA192 in both stable 0-5V and 0-10V configurations.

OPA192_5Vout.TSCOPA192_10Vout.TSC

2. Power Supply:

For the 0-5V output configuration: Max output current = 5V/2kOhm = 2.5mA

For the 0-10V output configuration: Max output current = 10V/2kOhm = 5mA

If you look at figure 20 in the datasheet (Output voltage swing vs Output current), you can calculate the output voltage swing from both rails.

For the positive rail, the worst case output voltage swing over temperature at 2.5mA load current is (V+) - 0.3V. For the 5mA load, the worst case output voltage swing is (V+) - 0.5V.

For the negative rail, I don't think you would see any significant swing because of the lower load currents. But just to be sure, I would recommend a negative rail of -0.232V. This can be generated easily by using a LM7705 (but be careful about the Vin_max to the LM7705 !).

To summarize:

0-5V configuration: You would need -0.232V / +5.5V supplies.

0-10V configuration: You would need -0.232V / +11V supplies.

I hope this helps! Please let me know if you have any questions. You can contact the power teams for recommendations for a power supply solution.

Best Regards,

Mihir Gupta

Field Applications Engineer

Texas Instruments

Hi Ue,

The opamp was not stable with a 22nF load. The easiest method of compensating for this is using the R_iso method (shown below), where a resistance is placed in series with the opamp output in order to create a zero in the frequency response of the system. This is the method used in the TI Design you linked to and it is mentioned in the TI Precision Labs video series as well.

While the Riso method will stabilize the system, it has a drawback of creating a voltage divider at the output with the R_iso and the RLoad. Hence the output voltage will be reduced by a factor of (RLoad/RLoad+R_iso). 

To overcome this problem, we can instead use the Output Pin compensation method (shown below):

In this method, a much larger capacitive load Cco is added in parallel to the CLoad. This has the effect of "hiding" the actual load from the OPA. Since the OPA now only effectively sees Cco at the output, we only have to stability compensate for the larger load by placing a series resistor Rco. There is no voltage drop with this stability compensation method.

Of course, the R_iso with dual feedback method can also be used in this case. This method is just easier to design :)

If you'd like to learn more about the other stability compensation methods for OpAmps, please go through the PPTs on the page below:

Regards,

Mihir