Analogue circuits tend to be more susceptible to continuous interference. RF interference within the signal frequency range is processed along with normal signals and is inseparable from them. In the case of frequency selective circuits such as PLLs or clocked ADCs there may be a narrowband response as the interfering frequency approaches the operating frequency, at which the circuit operation is completely disrupted. At other frequencies there is a worsening in signal-noise ratio, which may extend beyond the signal passband if the circuit bandwidth exceeds it. A circuit will be less sensitive to RF outside its passband but non-linearities will create “audio rectification” and intermodulation products which may be within the signal passband, or which may cause overload and saturation of the circuit with consequent complete failure.

Transient interference has in general less effect on analogue circuits than on digital. But transients may overload circuit functions (especially integrators) so that they become non-linear or fail to operate, and may then take a considerable time to recover, so that the duration of a transient’s effect is much greater than its actual pulsewidth. Such effects are minimized by ensuring a high overload margin or by including overload suppression using appropriate clamping devices or other limiting circuits.

As with digital circuits, the first defence is layout to keep ground interference currents away from the non-linear parts of the circuit. Filter the I/O leads or isolate them, to define a preferential safe current path for interference. Radiated RF fields that could create differential-mode voltages within the circuit need the minimum circuit loop area, for which a continuous ground plane is always the best solution.

Audio rectification

This is a term used rather loosely to describe the detection of RF signals by low-frequency circuits. It is responsible for most of the ill effects of RF susceptibility of both analogue and digital products.

When a circuit is fed an RF signal that is well outside its normal bandwidth, the circuit can respond either linearly or non-linearly, depending on the level of the interfering signal. If the signal level is low enough for it to stay linear, it will pass from input to output without affecting the wanted signals or the circuit’s operation. No, or at least only slight, susceptibility effect will be seen.

If the level drives the circuit into non-linearity, then the envelope of the signal (perhaps severely distorted) will appear on the circuit’s output. At this point it will be inseparable from the wanted signal and indeed the wanted signal will itself be affected by the circuit’s forced non-linearity. DC signals such as transducer measurements will be shifted in level, maybe saturated; audio signals will have the interfering modulation superimposed on them. This is the reason for using a standard 1kHz tone for RF immunity tests.

Therefore, using circuit components and configurations which have a large linear operating region will make for inherently better immunity. This usually means going for the maximum supply voltage which is available for the application. Or, making sure that the interfering voltage induced at the non-linear points is as low as possible.

Op-amp susceptibilities

The response of the circuit depends on its linear dynamic range and on the level of the interfering signal. Some research has been published on the susceptibility of several types of op amp; this can be found at the Elmac website, Annex F.

The test circuit was a unity-gain op-amp in which the interfering RF signal is applied in series with the non-inverting input. The DC conditions of the op-amp depended on the intended supply rails of the type under test. The unwanted signal is an unmodulated RF carrier across the frequency range 0.5MHz to 200MHz developed from a signal generator feeding an RF amplifier, and applied to the circuit via a broadband transformer.

In all cases the devices were subjected to an increasing RF level at spot frequencies to determine their RF-to-output transfer function. The resulting data, given here for one example (the National LM358N), shows the value of the deviation from the undisturbed level, not taking into account the wanted DC output offset. Conclusions are:

different op-amp technologies have different characteristic functions;

for a given type, different responses can be observed at different frequencies;

the DC operating conditions often have a marked effect on the response, and this is particularly noticeable when single-supply devices are operated close to their lower common mode voltage limit , i.e. with an input DC voltage of 0V on a single supply;

only one of the devices, the LM301A which has the oldest design, showed a response that was approaching linearity over the range tested. All others showed greater or lesser non-linearity. In many cases, a sudden change to a saturated state was observed at a particular RF level.

To maximise immunity to RF and transients, you should configure the active circuits to prevent interfering out-of-band signals from appearing at significant levels at the most non-linear points in the circuit: usually the inputs of amplifiers or comparators. The following diagram gives a selection of possibilities for this.

Bandwidth should be restricted to the minimum acceptable by input RC or LC filtering (1), feedback RC filtering (2), and low value (10 - 33pF) capacitors directly across device terminals, pin-to-pin and/or pin-to-0V (3). Small capacitors directly across non-linear points, e.g. transistor base-emitter junctions, are especially helpful. Care should be taken with RC filters and feedback components to ensure that they do not affect stability or worsen HF common-mode rejection, and it would be wise to check the stability of a particular circuit using Spice or similar before finalising it. The threat is that marginal stability in a circuit can lead to increased susceptibility at frequencies where the circuit is only just stable.  For balanced circuits, where the high frequency balance could be affected by the tolerance of the filter capacitors, using X2Y capacitors can be beneficial, as also is paralleling a capacitor differentially across the inputs. An even simpler method is to use low value stopper resistors or ferrites (4) at the inputs of all amplifying circuits. These work as low-pass filters in conjunction with the device's input capacitance in the range of hundreds of MHz, when parasitic coupling is at its most significant.

Signal level should be maintained as high as possible throughout, consistent with other circuit constraints, in order to maximise the signal-to-induced-interference ratio. The supply voltage should be as high as possible, and the DC conditions should be set so that the active device is not near the extreme of its common-mode range. (Both of these requirements are in opposition to the desire for low-voltage single supply amplifiers!) Supply decoupling with a parallel capacitor and series ferrite (5) is just as essential in analogue circuits as it is in digital ones; here the purpose is to prevent any incoming RF or transient interference from affecting the circuit via its power rail.

Balanced circuit configurations take maximum advantage of the inherent common-mode rejection (CMR) of op-amp circuits. But note that CMR is poorer at high frequencies and is affected by capacitive and layout imbalances. While balancing a circuit markedly improves its in-band interference immunity, it is less effective at improving RF and transient immunity.