Editor’s note: Art Kay is this month’s guest blogger on Texas Instruments’ Signal Chain Basics blog
A typical precision operational amplifier (op amp) may have a 1 MHz gain bandwidth product. Theoretically, you might expect RF signals that are in gigahertz to be attenuated to very small levels because they are far outside the amplifiers bandwidth. However, in practical cases this isn’t what happens. In reality, RF signals are “rectified” at the input of the amplifier by electrostatic discharge (ESD) diodes, input structures, and other non-linear elements contained within the amplifier. In a practical sense, the RF signal is converted to a DC offset voltage that adds with the amplifiers input offset voltage.
You might ask, “How do I determine the amplitude of the DC offset voltage generated by a given RF signal?” In fact, the amplifier’s sensitivity to RF interference depends on the amplifier’s design and technology. For example, many modern amplifiers have built in RF filters to minimize the problem. This filter is most effective for low-gain-bandwidth op amps as the filter cutoff frequency can be set to a lower frequency, which provides higher attenuation of the RF signal. Beyond this, some technologies have better inherent immunity to RF signals. For example, most CMOS devices have better immunity to RF than bipolar devices. Other factors, such as the input stage design, influence RF immunity.
Considering all these factors, how does the board and system level designer choose an amplifier? The answer is to look at the electromagnetic rejection ratio (EMIRR). This specification is similar to power supply rejection and common-mode rejection in that it translates the effects of RF interference to a DC offset at the amplifier’s input. As an example, Figure 1 shows the EMIRR curve for the OPA333. Notice from the curve that this op amp has an EMIRR of 120 dB at 1000 MHz. This is a very high rejection level, making it possible to directly compare this curve to the curve of other devices.
The EMIRR curve is a measurement of an op amp’s conducted immunity to an RF signal applied to the non-inverting input. The term “conducted” means that the RF signal was directly applied to the op amp input using an impedance-matched PCB. The reflections at the amplifiers input are characterized and accounted for.
Finally, the DC offset introduced by the RF signal is measured by a digital multimeter. Note that a low-pass filter is used between the amplifier and the multimeter to prevent potential errors caused by residual RF signals that pass through the amplifier. Figure 2 illustrates the test circuit used to characterize EMIRR.
The mathematical definition of EMIRR is given in equations (1) and (2). The two equations are re-arranged versions of each other. Equation (1) illustrates the relationship between the applied RF signal and the offset shift. Notice that the offset changes to the square of the applied RF signal. This means that a relatively small increase in the incident RF signal can cause a significant increase in the offset. Also notice that the EMIRR term acts to attenuate the effect of the RF signal; in other words, large EMIRR(dB) causes a large reduction in the offset shift. Equation (2) is the form of the equation used to calculate EMIRR(dB) during characterization.
Lastly, note that many other factors such as PCB layout and shielding impact the RF immunity of your system. However, once these factors are optimized in your design, the best performance can be achieved using an amplifier with good EMIRR. Furthermore, you don’t need to do any complex calculations. Simply compare the EMIRR curves of different amplifiers to select the best one for your application. I hope that you are able to make use of the EMIRR specification to optimize your system’s immunity to RF signals.
Please join us next time when we will discuss rethinking system-level management through subsystem over-current detection and monitoring.
- Chris Hall and Thomas Kuehl, “EMI rejection ratio of operational amplifiers,” Applications Report (SBOA128), Texas Instruments, August 2011
- Gerrit de Wagt and Arie van Staveren, “AN-1698 A specification for EMI hardened operational amplifiers,” Applications Report (SNOA497A), Texas Instruments, January 2010
- Chris Hall, “OPA333 EMI immunity performance,” Technical Brief (SBOZ004A), Texas Instruments, August 2012
- OPA333 datasheet