MX536A/MX636
_______________Detailed Description
The MX536A/MX636 uses an implicit method of RMS
computation that overcomes the dynamic range as well
as other limitations inherent in a straightforward computation of the RMS. The actual computation performed
by the MX536A/MX636 follows the equation:
V
RMS
= Avg. [V
IN
2
/V
RMS
]
The input voltage, V
IN
, applied to the MX536A/MX636 is
processed by an absolute-value/voltage to current converter that produces a unipolar current I1(Figure 1).
This current drives one input of a squarer/divider that
produces a current I4that has a transfer function:
I
4
=
I
1
2
I
3
The current I4drives the internal current mirror through
a lowpass filter formed by R1 and an external capacitor, CAV. As long as the time constant of this filter is
greater than the longest period of the input signal, I4is
averaged. The current mirror returns a current, I3, to the
square/divider to complete the circuit. The current I4is
then a function of the average of (I
1
2
/I4), which is equal
to I
1
RMS
.
The current mirror also produces a 2 · I
4
output current,
I
OUT
, that can be used directly or converted to a voltage using resistor R2 and the internal buffer to provide
a low-impedance voltage output. The transfer function
for the MX536A/MX636 is:
V
OUT
= 2 · R2 · I
RMS
= V
IN
The dB output is obtained by the voltage at the emitter
of Q3, which is proportional to the -log VIN. The emitter
follower Q5 buffers and level shifts this voltage so that
the dB output is zero when the externally set emitter
current for Q5 approximates I3.
Standard Connection
(Figure 2)
The standard RMS connection requires only one external component, CAV. In this configuration the
MX536A/MX636 measures the RMS of the AC and DC
levels present at the input, but shows an error for lowfrequency inputs as a function of the CAVfilter capacitor. Figure 3 gives practical values of CAVfor various
values of averaging error over frequency for the standard RMS connections (no post filtering). If a 3µF
capacitor is chosen, the additional error at 100Hz will
be 1%. If the DC error can be rejected, a capacitor
should be connected in series with the input, as would
typically be the case in single-supply operation.
The input and output signal ranges are a function of the
supply voltages. Refer to the electrical characteristics for
guaranteed performance. The buffer amplifier can be
used either for lowering the output impedance of the circuit, or for other applications such as buffering highimpedance input signals. The MX536A/MX636 can be
used in current output mode by disconnecting the internal load resistor, R
L
, from ground. The current output is
available at pin 8 (pin 10 on the “H” package) with a
nominal scale of 40µA/V
RMS
input for the MX536A and
100µA/V
RMS
input for the MX636. The output is positive.
True RMS-to-DC Converters
6 _______________________________________________________________________________________
ELECTRICAL CHARACTERISTICS—MX636 (continued)
(TA= +25°C, +VS= +3V, -VS= -5V, unless otherwise noted.)
Rated Performance +3/-5 V
Dual Supplies +2/-2.5 ±16.5 V
Single Supply +5 +24 V
UNITS
MIN TYP MAX
CONDITIONSPARAMETER
Quiescent Current (Note 10) 0.8 1 mA
Note 1: Accuracy is specified for 0 to 7V
RMS
, DC or 1kHz sine-wave input with the MX536A connected as in Figure 2.
Note 2: Error vs. crest factor is specified as an additional error for 1V
RMS
rectangular pulse stream, pulse width = 200µs.
Note 3: Input voltages are expressed in volts RMS, and error as % of reading.
Note 4: With 2kΩ external pull-down resistor.
Note 5: Accuracy is specified for 0 to 200mV, DC or 1kHz sine-wave input. Accuracy is degraded at higher RMS signal levels.
Note 6: Measured at pin 8 of DIP and SO (I
OUT
), with pin 9 tied to COMMON.
Note 7: Error vs. crest factor is specified as an additional error for 200mV
RMS
rectangular pulse input, pulse width = 200µs.
Note 8: Input voltages are expressed in volts RMS.
Note 9: With 10kΩ external pull-down resistor from pin 6 (BUF OUT) to -V
S
.
Note 10: With BUF input tied to COMMON.
POWER SUPPLY