Ecler DPA2500T, Service Manual

Page: 4 / 69

52-0031-0100 EP04-05-05 DPA2500T DPA4000T Anglès.doc 1 of 2
POWER MODULE.
Due to the high power level required on the output load, the amplifier final stage's structure differs from the design
used until now. This is due to the breakdown voltage limit on P-channel MosFET's, with is 200V. This final stage is
formed by several shunted MosFET's, where those of the positive branch are common-drain configured, and the
negative branch are mounted in common-source configuration.
The system's controlling device is a NE5534-type operational amplifier, with is internally compensated in order to
obtain gain levels equal or higher than three. The amplifier's feedback is given by a resistor and a capacitor
associated to the operational amplifier's non inverting input.
Transistors BF471 and BF472 are common-base configured, forming a current supply structure. This specific
transistor type is used because of the higher Vce voltage level required by this design. They perform
simultaneously two functions: they polarize the MosFET's gate-source junctions, keeping them on the conducting
edge, and they transfer the OpAmp's output voltage variations referred to signal ground.
The signal variations normally reflected by Q107 and referred to the positive power supply, are now needed to be
floating variations, and referred to the outputs. This function is done by Q109-110 (BF472), with are mounted
forming a Wilson-type mirror current supply. This mirror current supply transfers all of the current variations
detected while descending through Q109's collector, to similar variations on Q110's collector also downward
current. Resistors R167 and R174 are used to balance the current mirror, in order to avoid the use of transistors
with forcibly the same beta value. C138 and C141 suppress their resistance when high frequency signal is
processed. Diodes D126 and D127 avoid the transistors to get saturated, and R171 eliminates the loads on
BF472's bases (Baker Circuit).
The system requires about 12Vdc additional voltage upon the usual Vcc level, this allows a correct saturation and a
symmetric clipping at the higher MosFET's.
The correct polarization current value is adjusted by a 4K7 potentiometer connected to the BF transistor's emitter.
This adds an additional current to the current source output on the BF transistor's loading resistors.
In order to maintain the appropriate stand-by current level against varying temperature conditions, BD437-type
transistors are used. As they have a particular temperature-depending base-emitter voltage curve, this voltage is
used to keep a correct voltage reference for the current supply. As the temperature rises, the reference voltage
level decreases, the gate-source voltage also decreases and, finally, the bias current also decreases.
Transistors Q111 and Q112, and their corresponding twins at the lower branch, form a current-buffering circuit with
allows a fast charge and discharge of the power MosFET's gates.
The Zobel circuit, a resistance-capacitance-inductance formed network associated to the amplifier's output, tries to
keep the amplifier's output load impedance constant no matter with load value is connected to the output, or with
frequency is processed, in order to avoid phase shifts on the feedback signal.
To avoid the presence of DC voltage on the output, a diac-triac based system is used, with shorts the output to
signal ground in case the DC level reaches the diac's triggering value. To avoid this to happen when processing
correct signal (sine waveform, music...), the diac obtains its reference level from a filtering network formed by a
100K resistor and a 1µF capacitor.
The protections circuitry overhauls the MosFET's power consumption. Basically, this circuitry consists of two
important sections: MosFET's Id current monitoring, and MosFET's Vds value detection.
author:
J. Colomines
date:
051006
project:
EP04/05-05
product:
DPA2500T/4000T
ECLER
approved:
num:
52..0031
version:
01.00
title:
FUNCTIONING DESCRIPTION