OPA564: Driving large Capacitive load

Hi David,

Much about the high C-load stability question depends on how you intend to configure the OPA564. The circuit configuration, inverting or non-inverting, and the gain certainly are factors. Also, the actual load impedance presented to the OPA564 output needs to be fully understood. You mention a 200 mA load current and a capacitance of 1 to 5 uF. Therefore, there is some resistive load in parallel with the large load capacitance. If you can provide a drawing of your circuit idea, then we can assess it from a stability standpoint. Make sure you indicate a single, or dual supply for the OPA564.

In general, any op amp can be externally compensated for a specific capacitive load. There are some common compensation techniques that can be applied. The more difficult challenge is when the C-load can vary; you mention 1 to 5 uF. Is the capacitance going to be a fixed value, or will it vary?

Sometimes, a high C-load circuit might actually result in stable operation when we might expect otherwise. The reason that occurs is because the load capacitor, in conjunction with its equivalent series resistance (ESR) modify the gain-phase response of the amplifier circuit resulting in enough phase margin to maintain stability. Thus, the electrical characteristics of the load capacitor need to be included in an analysis. The one outcome of this sort of unexpected compensation is it much reduces the circuit's otherwise wide bandwidth.

When driving a high C-load the transient current can be much higher than anticipated; especially, if the edge rate of the applied waveform is fast. Remember i = C dv/dt. The duration of these current pulses associated with the waveform edges is usually short and the power dissipation doesn't increase by much. The main concern is planning for the higher current and making sure the power op amp can supply it.

Regards, Thomas

Precision Amplifiers Applications Engineering

Hi Thomas and thank you for your answer.

I will elaborate a bit more.

The DUT requires quite many different power rails, each with different nominal voltage and maximum current load.

For example:

• 0.6V at max 50mA
• 3.8V at max 5mA
• 0.6 at max 900mA
• 1V at max 1mA
• etc.

Obviously the type and total decoupling capacitance will be configured for each rail separately.

This step is finding a compromise between maximum rail voltage ripple and ease of driving noise.

I don't intend driving with wave-forms other than sinusoidal ones - just will vary amplitude and frequency (whilst always monitoring actual voltage wave-form at DUT package).

Would prefer to use the amp in non-inverting configuration but not absolutely necessary - which would give better response with C load?

Gain probably low - between 1 and 5 (or minus).

Another caveat that I forgot to mention is that I will need to place a current sense shunt resistor somewhere after the amp. Depending on the particular rail, it's value would be between say 0.1 and 10 ohm.

So, given the new information, how could I best compensate?

Would you recommend non-invert or invert config?

Do you think the OPA564 is the best choice?

Thanks again

David

Hi David:

Here are my answers to your questions:

Would prefer to use the amp in non-inverting configuration but not absolutely necessary - which would give better response with C load?

Use which ever configuration works best for your application. The compensation can be optimized for either one, inverting or non-inverting.

Another caveat that I forgot to mention is that I will need to place a current sense shunt resistor somewhere after the amp. Depending on the particular rail, it's value would be between say 0.1 and 10 ohm.

Actually adding a resistance in series with the power op amp output, after the loop return, helps isolate its output from the C-load. A common op amp C-load compensation technique is the inclusion of an "Riso" in series with op amp output. It is discussed in the op amp stability information I will recommend.

So, given the new information, how could I best compensate?

I suggest having a look at the following materials to gain a better understanding about the reasons behind op amp instability, and how to apply compensation:

Section 10. TI Precision Labs - Op Amps: Stability

Would you recommend non-invert or invert config?

As mentioned previously, us whatever configuration serves your application best. Once you have your circuit designed we can assist you with the compensation effort should you need help.

Do you think the OPA564 is the best choice?

You mention an output level such as 0.6 V at max 900mA. Since the voltage level is low relative to the Vsupply you might use the internal power dissipation of the op amp can be high. It all comes down to the thermal considerations described in the OPA564 datasheet. If the dwell isn't long, then the op amp may not heat up much. Otherwise, you may want to consider a higher power op amp such as the OPA548. You would have to check its spec's to make sure they meet all of the application requirements.

You can find the OPA548 datasheet here:

Regards, Thomas

Precision Amplifiers Applications Engineering