Data Sheet

AD8541/AD8542/AD8544

APPLICATIONS INFORMATION

analog.com

Rev. H | 14 of 18

NOTCH FILTER

The AD854x have very high open-loop gain (especially with a

supply voltage below 4 V), which makes it useful for active filters

of all types. For example, Figure 36 illustrates the AD8542 in the

classic twin-T notch filter design. The twin-T notch is desired for

simplicity, low output impedance, and minimal use of op amps.

In fact, this notch filter can be designed with only one op amp if

Q adjustment is not required. Simply remove U2 as illustrated in

Figure 37. However, a major drawback to this circuit topology is

ensuring that all the Rs and Cs closely match. The components

must closely match or notch frequency offset and drift causes

the circuit to no longer attenuate at the ideal notch frequency. To

achieve desired performance, 1% or better component tolerances

or special component screens are usually required. One method

to desensitize the circuit-to-component mismatch is to increase

R2 with respect to R1, which lowers Q. A lower Q increases

attenuation over a wider frequency range but reduces attenuation at

the peak notch frequency.

Figure 36. 60 Hz Twin-T Notch Filter, Q = 10

Figure 37. 60 Hz Twin-T Notch Filter, Q = ∞ (Ideal)

Figure 38 is an example of the AD8544 in a notch filter circuit.

The frequency dependent negative resistance (FDNR) notch filter

has fewer critical matching requirements than the twin-T notch,

where as the Q of the FDNR is directly proportional to a single

resistor R1. Although matching component values is still important,

it is also much easier and/or less expensive to accomplish in the

FDNR circuit. For example, the twin-T notch uses three capacitors

with two unique values, whereas the FDNR circuit uses only two

capacitors, which may be of the same value. U3 is simply a buffer

that is added to lower the output impedance of the circuit.

Figure 38. FDNR 60 Hz Notch Filter with Output Buffer

COMPARATOR FUNCTION

A comparator function is a common application for a spare op

amp in a quad package. Figure 39 illustrates ¼ of the AD8544 as

a comparator in a standard overload detection application. Unlike

many op amps, the AD854x family can double as comparators

because this op amp family has a rail-to-rail differential input range,

rail-to-rail output, and a great speed vs. power ratio. R2 is used

to introduce hysteresis. The AD854x, when used as comparators,

have 5 µs propagation delay at 5 V and 5 µs overload recovery

time.

Figure 39. AD854x Comparator Application—Overload Detector

PHOTODIODE APPLICATION

The AD854x family has very high impedance with an input bias

current typically around 4 pA. This characteristic allows the AD854x

op amps to be used in photodiode applications and other applica-

tions that require high input impedance. Note that the AD854x

has significant voltage offset that can be removed by capacitive

coupling or software calibration.

Figure 40 illustrates a photodiode or current measurement applica-

tion. The feedback resistor is limited to 10 MΩ to avoid excessive

output offset. In addition, a resistor is not needed on the noninvert-

ing input to cancel bias current offset because the bias current-re-

lated output offset is not significant when compared to the voltage

offset contribution. For best performance, follow the standard high

impedance layout techniques, which include the following:

►

Shielding the circuit.

►

Cleaning the circuit board.