AD637

Data Sheet

Referring to Figure 8 for optional external gain and offset trim

schematic. The following sections describe trimming for greater

accuracy in detail.

Offset Trim

Ground the input signal (VIN) and adjust R1 until the output

voltage at Pin 9 measures 0 V. Alternatively, apply the least

expected value of VIN.at the input VIN and adjust R1 until the dc

output voltage at Pin 9 measures the same value as the rms input.

Scale Factor Trim

Insert Resistor R4 in series with the input to decrease the range

of the scale factor. Connect a precision source to Pin 13 and

adjust the output for the desired full-scale input to VIN, using

either a calibrated dc or 1 kHz ac voltage, and adjust Resistor R3

to give the correct output at Pin 9 (that is, 1 V rms at the input

results in a dc output voltage of 1.000 V dc). A 2 V p-p sine

wave input yields 0.707 V dc at the output. Remaining errors

are due to the nonlinearity.

25kΩ

–VS

–VS

–VS

1

25kΩ

DEN

INPUT

CAV

BUFF IN

BUFF

OUT

CS

dB

OUTPUT

3 COMMON

ABSOLUTE

VALUE

BIAS

2 NIC

4

OUTPUT

OFFSET

4.7kΩ 5

6

7

8

9

10

11

12

NIC

13

14

NC

VIN

VIN

R4

147Ω

CAV

+

R3

1kΩ

R2

1MΩ

R1

50kΩ

+VS

+VS

+VS

+VS

OUTPUT

OFFSET

TRIM

SCALE FACTOR TRIM

AD637

RMSOUT

SQUARER/

DIVIDER

–

+

–

+

RMS

OUT

–

+

NIC = NO INTERNAL CONNECTION

Figure 8. Optional External Gain and Offset Trims

CHOOSING THE AVERAGING TIME CONSTANT

The AD637 computes the true rms value of both dc and ac

input signals. For dc inputs, the output tracks the absolute

value of the input exactly. However, when the voltage is ac,

the converted dc output voltage asymptotically approaches the

theoretical rms value of the input. The deviation from the ideal

rms value is due to the implicit denominator inherent to averag-

ing over an infinite time span. Because the error diminishes as

the averaging period increases, it quickly becomes negligible.

The remaining error components are the ac ripple and dc offset

voltage, if any. The ac and averaging error components are both

functions of the input-frequency (f) and the averaging time

constant τ (τ: 25 ms/µF of averaging capacitance). Figure 9 shows

the output errors, which are enlarged for clarity. The frequency

of the ac component (ripple) is twice the frequency of the input,

the dc error is the RSS sum of the average rectified error and

any fixed value dc offset.

The value of CAV and the 25 kΩ feedback resistor establish the

averaging time constant, and solely determines the magnitude

of the rms-to-dc conversion error. Furthermore, any post-

conversion filtering does not improve the dc component

composite result.

Equation 1 defines the approximate peak value of the ac ripple

component of the composite output.

)

/

1

(

where

reading

of

%

in

3

.

6

50

f

f

>

τ

τ

(1)

DOUBLE-FREQUENCY

RIPPLE

EO

TIME

AVERAGE ERROR

IDEAL

EO

DC ERROR = AVERAGE OF OUTPUT – IDEAL

Figure 9. Enlarged Composite Conversion Result for a Sinusoidal Input

Increasing the value of the averaging capacitor or adding a post-

rms filter network reduces the ripple error.

The dc error appears as a frequency dependent offset at the

output of the AD637 and follows the relationship

reading

of

%

in

4

.

6

16

.

0

1

2

2 f

τ

+

SINE WAVE INPUT FREQUENCY (Hz)

100

0.1

1.0

10

10k

1k

100

10

DC ERROR

PEAK RIPPLE

Figure 10. Comparison of Percent DC Error to the Percent Peak Ripple over

Frequency Using the AD637 in the Standard RMS Connection with a 1 × µF CAV

The ac ripple component of averaging error is greatly reduced

by increasing the value of the averaging capacitor. However, the

value of the averaging capacitor increases exponentially while

the settling time increases directly proportion to the value of

the averaging capacitor (TS = 115 ms/µF of averaging capacitance).

Rev. L | Page 14 of 25