OPA549: OPA549 Transient Response

Hello Akash,

The OPA549 power op amp's transient characteristics are seen two places in the datasheet. The Electrical Characteristics table under Frequency Response lists; Settling Time: ±0.1%, 20 µs typical for G = –10 V/V, and 50 V step. And under the Typical Characteristics, the LARGE-SIGNAL STEP RESPONSE, G = 3 V/V, CL = 1000 pF, graph. In each case the conditions are for a voltage voltage change at the output, but certainly the current is changing as well. The horizontal axis on the graph is 5 us/div so time wise the transient events are over in microseconds.

Just about every circuit, load and stimuli are different so the transient behavior can certainly be different for each of them. The specific characteristics of the load applied will affect just how the circuit settles. You are using an e-load and it too will have specific settling characteristics when it slews from one current level to another. And it likely has some dynamic impedance when it is slewing that is different than when it is providing a constant current. If that impedance is complex (R+/-jX), then that will affect how the op amps settles.

You mention, "I cannot add capacitance at the output as that makes the output unstable." Viewing the datasheet SMALL-SIGNAL OVERSHOOT vs LOAD CAPACITANCE graph, one can infer that the OPA549 should be able to drive a reasonably high capacitance (~>20 nF, G = +2 V/V), without breaking into oscillation. Certainly as the capacitance on the output is increased there will be reduced phase margin and increased overshoot, but I would expect that it remains stable. The fact that you are having stability issues when you add a capacitance to the OPA549 output is unexpected unless the e-load appears as a high capacitance.

The OPA549 should be able to drive an LDO, or buck regulator, providing the load load impedance that it represents to the OPA549 output doesn't cause the phase margin to become badly degraded. Should that occur the transient response and settling time suffers, and instability can even occur.

If your application can tolerate it, you may find that adding a small resistance between the OPA549 output pin and the e-load (a few tenths of an ohm) can really affect the settling behavior. That is if the load appears predominately capacitive to the op amp output.

Regards, Thomas

Precision Amplifiers Applications Engineering

Thomas,

Thank You for the detailed explanation. 

I am able to get a good response for load current slew-rate of 0.3A/us which is enough for the application. 

Regrads

Akash