ETC 429A

Accurate, Wideband Multiplier,
Divider, Square Rooter
Model 429
1.0%10.5% Accuracy Without Trimming (429A/B)
Low Drift to 1.0mV/°C max
Wideband - 10MHz
0.2% Nonlinearity max (429B)
External Amplifiers Not Required
MTBF: 169, 268 Hours
Fast Divider
Modulation and Demodulation
Phase Detection
Instrumentation Calculations
Analog Computer Functions
Adaptive Process Control
Trigonometric Computations
The Model 429, an extremely fast multiplier/divider, should be
considered if bandwidth, temperature coefficient, or accuracy are
critical parameters. Based on the transconductance principle to achieve
high speed, the model 429 offers a unique combination of features,
those being ½% max error (429B) and 10MHz small signal bandwidth.
the offset is zeroed out? Can one get less by interchanging inputs?
1. Nominal output is XY/10 = (5)(1)/10 = 50mV
2. Expected error is (5) (0.5%) + (1) (0.3%)
28mV, 5.6% of output (0.28% of F.S.)
Both models 429A and 429B are internally trimmed achieving
max errors of 1.0% and 0.5 % respectively. By fine trimming
the offset and feedthrough with external trim potentiometers
typical performance may be improved to 0.5% for the 429A and
0.2% for the 429B.In addition to high accuracy and high
bandwidth, the model 429 offers exceptionally good stability for
changes in ambient temperature. Model 429B is 100%
temperature tested in order to guarantee an overall accuracy
temperature coefficient of only 0.04%/"C max. Additionally,
offset drift is held to only I mV/°C max. To satisfy OEM
requirements of low cost, the 429 uses transconductance
principles with the latest design techniques and components to
achieve guaranteed performance at competitive prices.
Multiplication accuracy is generally specified as a percentage of full
scale output. This implies that error is independent of signal level.
However, for signal levels less than 2/3 of full scale, error tends to
decrease roughly in proportion to the input signal. A good
approximation of error behavior is:
3. Interchanging inputs (1) (0.5%) + (5) (0.3%) =
20mV, 4.0% of output (0.20% of F.S.)
Compare this with the overly conservative error predicted by the
overall 1% of full scale specification: 100mV, or 20% of output.
Accuracy, and its components, feedthrough, linearity, gain, (and phase
shift) are frequency dependent. Feedthrough is constant up to 100kHz
for the Y input, and up to 400kHz for the X input. Beyond these
frequencies it rises at approximately a 6dB/octave rate due to
distributed capacitive coupling. A plot of typical feedthrough vs.
frequency is shown in Figure 1. For this measurement one input is
driven with a 20V p-p sine wave while the other input is grounded and
the feedthrough is measured at the output. This error will decrease
roughly in proportion to the input signal, and will also vary with
temperature (about 0.01%/°C of the nonzero input). Low frequency
feedthrough error can be further reduced from the internally trimmed
limits by the use of optional external potentiometers.
f (X, Y) ≈ |X| ∈X, + |Y| ∈Y, where ∈X and ∈Y are the fractional
onlinearities specified for the X and Y inputs
EXAMPLE.- For Model 429A ∈X = 0.5 %, ∈Y = 0. 3%. What
maximum error can one expect for x = 5V, y = 1V, providing
Information furnished by Intronics is believed to be accurate and reliable. However, no
responsibility is assumed by Intronics for its use; nor for any infringements of patents or
other rights of third parties which may result from its use. No license is granted by
implication or otherwise under any patent or patent rights of Intronics.
1400 Providence Highway, Building #2
Norwood, MA 02062
Phone (781) 551-5500
FAX (781) 551-5555