CH679807A5 - - Google Patents

Description

1

CH 679 807 A5

2nd

description

The invention relates to a partial discharge sensor according to the preamble of claims 1 and 8 and relates to the field of partial discharge measurement on electrical equipment such as transformers, generators, motors, switchgear including solid-insulated and SF6 switchgear, cable networks and others.

In addition to the cyclical or permanent status diagnosis, there are in principle also possibilities for early anomaly detection and thus the triggering of shutdowns of endangered objects, while preventing a destructive breakdown.

In addition to the classic and standardized partial discharge measurement technology (“Partial discharge measurement. IEC publication 270/1981”), a small probe and a more favorable useful interference signal ratio are aimed for on-site diagnosis. For this purpose, it is known to compensate for the interference signals in bridge circuits (Zaengl, W .; Klauss, A .: On-site surveillance of potential transformers by means of DP measurement. CIGRE Symposium, Vienna 1987, Section 700-01) or the temporal to hide limited interference components with electronic discriminator circuits (Lemke, E .: diagnosis of electrical equipment based on very broadband partial discharge measurements. ELEKTRIE, Berlin 38/1984/10, pp. 387-398).

A method based on 50 Hz test alternating voltage for detecting internal TE pulses is presented in DE-OS 2 721 353. Disturbances, e.g. External discharges that occur in the area of the maximum of the test voltage are suppressed by special electronic short-circuiters and are therefore not used for evaluation.

WP 253 333 discloses an arrangement for suppressing the RF interference voltages superimposed on the partial discharge signals, which essentially consists of a push-pull transistor circuit with diodes and a controllable resistor between the two bases for generating a quiescent current. The output of a control amplifier amplifying the useful interference signal mixture is connected to the control electrode of the controllable resistor via a isolating capacitor and a rectifier circuit with optimal timing and basic level adjustment, and the control pulse is connected via a decoupling diode so that only the partial electrical discharge pulses are transmitted via the push-pull transistors when the control signal is active.

DE-OS 3 630 026 discloses an ultrasonic sensor for the indication of TE from a partially flat, light guide loop which is fixed in several curvatures and attached to the test specimen and which is coupled at one end to a monochrome light source and at the other end to a light sensor. The ultrasonic waves caused by TE cause elastic changes in the radii of curvature of the light guide, thus coupling and loss or modulation of the light, which is used for indication and evaluation.

DE-PS 3 408 256 discloses a measuring probe for detecting the electromagnetic radiation field which arises when TE is ignited, with frequency components far above 100 MHz, using a broadband antenna which consists of a three-electrode system connected to one another via a compensation network. On the side facing the test specimen, a frame-shaped compensation electrode is arranged between a measuring electrode and a comprehensive reference electrode designed with an opening and housing, and differential amplifiers are integrated within the reference electrode, the rise time of which is very small compared to the duration of the TE pulses and im Nanosecond range. The arrangement of the electrodes results in a good straightening effect and their interconnection results in increased interference signal reduction from far fields.

The condition monitoring described at the outset requires the TE probe to be constantly ready for operation. Since the problematic suppression of interference signals can only be fully effective in potential-free operation (without a mains connection), a cyclical replacement of the internal batteries or their recharging is essential for permanent condition monitoring, so that interruptions in operation can not be ruled out in addition to time-consuming maintenance. As a result, a second probe should be available for each measuring point to ensure redundancy if a critical error occurs during maintenance.

The circuit complexity for the previously known TE probe (DD-WP 214 461) is also not insignificant, although its possible uses are based on the above. Area of application of the accident protection goes far.

Finally, for permanent condition monitoring, the TE probe would have to be permanently installed on the object to be protected. Problems arise with ensuring accessibility for changing the battery. In addition, the size and weight of the known TE probe do not allow easy attachment to the object. In the case of open-air objects, there is also the problem of climate protection. Finally, with a view to selective fault detection, it would be desirable to take measurements at several vulnerable points within electrical equipment (e.g. in the transformer, i.e. under oil). The previously known TE probe is not suitable for this, since it cannot be hermetically sealed.

The broadband amplifiers customary in the prior art for amplifying the high-frequency TE signals (> 100 MHz) deliberately need to keep working resistances (<100 ohms), with switching times in the nanosecond range, small and therefore large operating currents (approx 10 mA per transistor), which is an obstacle to battery operation over a sufficient operating time.

For classic rectifier circuits (including avalanche diodes), the An5 is

10th

15

20th

25th

30th

35

40

45

50

55

60

65

3rd

CH 679 807 A5

4th

Talk threshold about 2 to 10% of the final level, so that only dynamic values from 1:50 to 1:10 can be achieved. The need for a fixed operating point setting in the classic rectifier by means of a voltage divider also increases the current requirement.

The aim of the invention is to implement a PD sensor for PD diagnosis including early diagnosis of TE anomalies as well as the accident protection of high-quality electrotechnical high-voltage equipment. The aim is:

- simplest circuit construction,

- small number of components,

Feasibility as an integrated circuit,

- low energy consumption,

- light weight,

- small dimensions,

- possibility of hermetic encapsulation,

- high immunity to interference,

- high dynamic of the modulation without changing the measuring range,

- Simple adaptation options on the test object (implementation if necessary)

- responsiveness only to short-term impulses in the nanosecond time range,

- TE signal detection on metal surfaces with a distance of the measuring points of only 10 cm.

The objectives mentioned above are in part directly linked and complement each other.

Based on the above objective, the task is to record TE signals from the test object by means of a transducer while reducing the interference signal level and to shape them in a processing unit so that the output signals at an increased level and thus reduced risk of further interference with subsequent signal processing in the control center ( control room).

According to the invention, this object is achieved by the features listed in the characterizing part of patent claim 1.

An alternative solution is given in claim 8.

If the test object has a metallic vessel, it is advantageous to tap partial discharge signals at exposed locations on the wall, it being provided according to the invention that the adapter consists of two contact tips with a distance which determines the active measuring zone of a metallic test vessel.

Another possibility of TE signal detection is that the adapter consists of a single-pole contact to the test object and the conductive housing of the sensor forms an antenna.

According to the invention, an expedient embodiment provides that the single-pole contacting in ferromagnetic test specimen vessels takes place by means of a holding magnet.

In some applications, however, it is expedient for the electrically conductive housing of the sensor to be electrically and mechanically connected to the housing of the test object and for an antenna, preferably a disk-shaped one, to be arranged as the opposite pole.

According to the invention, in particular for the latter arrangement, in order to increase the useful interference signal ratio, a short-circuited waveguide with a wave propagation time of less than 100 ns is connected from the signal line to the sensor housing.

It is possible that the waveguide consists of a coaxial cable, the inner conductor of which is connected on the input side to the signal line and on the output side to the metallic housing of the sensor, or that a parallel strip line is connected accordingly. A pulse transformer can also be used to separate interference pulses of more than 100 ns from the short TE signals, thus improving the signal-to-noise ratio.

A variant of a pulse transmitter provides that the adapter has two inputs, which are connected to the beginning and end of the inner or outer conductor of a coaxial line, preferably wound on a Manifer shell core, the outer or inner conductor of which is connected to the signal line at one end the two charge storage capacitors are connected at the other end to the ground of the circuit arrangement.

Instead of a battery power supply or for buffering the same, it is also possible to feed a supply current via the output signal line from the control room, whereby according to the invention one contact of the output socket is connected via a resistor to the positive and the other contact via a resistor to the negative operating voltage line and at the end the output signal line in the control room is choked and a power supply is connected.

To avoid oscillation tendencies and asymmetrical modulations, it is provided that a low-inductance smoothing capacitor is connected to the operating voltage potentials via the connection points of the transistor emitters.

In order to increase the detection limit for the TE sensor, it is possible according to the invention that a second complementary transistor arrangement with the same structure and with charge storage capacitors as a preliminary stage is inserted between the signal line and the bases of the two complementary bipolar transistors.

A possible configuration of the invention, which is advantageous in some applications, provides that the transducer consists of an active broadband antenna, preferably of the three-electrode system with fast differential amplifiers known per se.

The evaluation in the control center is expediently carried out using digital signals, for this purpose it is expedient that, instead of the emitter followers, switched OP amplifiers are arranged as voltage followers, the outputs of which are connected to an arrangement generating a unipolar pulse sequence.

The TE signals of a TE sensor at high voltage potential are preferably transmitted to the control center in that the output socket is connected to the modulation input of a downstream light transmitter or laser arrangement, which is connected via fiber optic cables or

5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

5

CH 679 807 A5

6

is coupled via a focused beam to a light receiver upstream of the signal processing stage with a demodulator.

A simple analog signal evaluation and evaluation of critical insulation faults provides that a measurement value is formed in the signal processing stage which corresponds to the mean value of the square of the TE pulse charges.

Appropriate capacitive coupling of the signal processing stage requires a zero level clamp in which an OP amplifier with its non-inverting input at reference potential and with its inverting input via a high-pass capacitance at the output of the processing unit, via a high-pass resistance at reference potential, via a diode in parallel a resistor to the output of the OP amplifier and to the input of an integrator.

The integration forms a signal level that corresponds to the mean square value of the TE-lm pulse charge. An economic variant provides that the integrator consists of a resistor-capacitor-resistor series circuit, the output signal being tapped via the capacitor, or that an OP amplifier switched as an active integrator instead of or with the resistor capacitor -Resistor series connection are arranged.

The combination of the resistor-capacitor-resistor series connection also prevents the TE sensor from being influenced by external high-frequency interference signals, which can be picked up by the signal line from the environment due to the antenna effect, due to the low-pass behavior seen from the output socket.

In the following exemplary embodiment, the invention is explained in more detail with reference to seven figures.

Show it:

Fig. 1 The basic principle of the TE sensor technology Fig. 2 The adaptation by means of a holding magnet and use of the probe housing as an antenna

Fig. 3 The adaptation with mechanical attachment and simultaneous electrical contacting of the probe housing on the metal housing of the test object and connection of a disk-shaped radiator on the signal line

Fig. 4 An overall circuit diagram of the TE sensor. Fig. 5 A circuit example for analog pulse processing

Fig. 6 Two variants of a pulse transmitter. Fig. 7 A circuit example for averaging the square of the TE pulse charges

1 shows the basic principle of partial discharge sensor technology. The partial discharges TE emitted by the different measurement objects MO can be in various ways at the most can also be removed from their metallically conductive vessels MG using an adapter AD, reinforced and shaped in the processing unit VE and evaluated after transmission to the central unit SV.

In the simplest case, the adapter AD is partly realized by the metallic vessel MG of the test object PO, the area between the metal-contacted pick-up points MPi; MP2 serves. The distance MP1-MP2 determines the response threshold. It can be reduced to less than 10 cm.

Another possibility of TE signal detection is single-pole contacting with the test object, in which case the opposite pole is realized by the housing of the sensor itself, which acts as a radiator (antenna) for the rapid electromagnetic compensation processes for TE signals. For single-pole contacting, the use of holding magnets HM is also expedient for iron vessels MG (FIG. 2).

On the other hand, it is also possible to use an inverse arrangement, i.e. the metallic housing SG of the sensor is electrically and mechanically connected to the test object housing MG, and the opposite pole is connected by an additional radiator SR (antenna), e.g. is disc-shaped, realized (Fig. 3).

A coaxial cable KL is preferably used for the latter arrangement in connection with the adapter for emulating a fictitious short circuit for signals with a duration of more than a few 10 ns. Its inner conductor is connected on the input side to the signal line SL and on the output side to the metallic housing SG of the sensor (FIG. 4). The outer conductor of the KL cable is connected on both sides to the sensor housing SG. The optimal length of the KL cable is between 10 and 50 cm. With regard to universal use, it is expedient to fully integrate this HF cable KL as part of the adapter in the circuit (FIG. 4). In principle, modified waveguides can be used instead of the coaxial cable KL, e.g. Pulse transformer or parallel stripline.

Fig. 6 shows schematically the possibility of winding a coaxial line KL on a Manifer shell core or through a ring core MK and interconnecting it as a pulse transformer.

In variant A, the two inputs are MP1; MP2 connected to the beginning and end of the inner conductor IL1 of the coaxial line KL, while the beginning or end of the outer conductor AL1 via the signal line SL to the charge storage capacitors Csi; Cs2) and the end or beginning of the outer conductor AL1 is connected to the ground of the circuit arrangement.

In variant B, however, the two inputs are MP1; MP2 is connected to the outer conductor AL2 of the beginning and end of the coaxial line KL, while the inner conductor IL2 is connected to the signal line SL and to ground.

This pulse transformer with a Manifer shell core MK and a few turns of the coaxial line KL represents a space-saving arrangement which can be tuned to sharply separate disruptive pulses.

5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

7

CH679 807 A5

8th

For signal amplification in the nanosecond time range, concepts are used in classic circuit technology whose current consumption is in the range of a few 10 mA due to the required rapid recharging of parasitic stray capacitances (operating voltage approx. 10 V). In addition, the rectification of nanosecond pulses of low amplitude (response threshold e.g. 10 mV) with high dynamics (modulation e.g. up to 5 V) in the classical circuit technology requires considerable effort, which also increases the power consumption significantly.

An unexpected effect in the field of nanosecond pulse technology with regard to the lowest current consumption (<1 mA) and high dynamics is achieved by the arrangement according to the invention in which a bipolar transistor T2 of the npn type with a pnp transistor Ti is switched in the same way that the collectors of both transistors directly and the base of each transistor Ti; T2 are connected via a high-resistance resistor Rb in the MOhm range (FIG. 4). The emitter of the p-n-p transistor Ti is connected to the positive operating voltage pole (+) and the emitter of the n-p-n transistor T2 to the negative operating voltage pole (-). The input signal line SL is via the charge storage capacitors Csi and Cs2, whose capacitance is in the range of a few 10 pF, each with the base of the transistors Ti; T2 connected. The connection point of the collectors of both transistors Ti, T2 is connected on the one hand via a coupling capacitor Cki in the range of less than 10 nF with a clamping resistor R «1 of the order of 50 kOhm, which with its other connection is connected to the negative operating voltage (-) is present. The connection point of coupling capacitor Cki and resistor Rki is connected to the emitter of a p-n-p transistor T3, the base of which is connected to the positive operating voltage (+) via a charging capacitor Cli of capacitance on the order of 100 pF. At the connection point of the charging capacitor Cli and the base of T3, the base of a p-n-p transistor T4 is connected in a collector circuit with the load resistor Rai.

In an analogous manner, a second measuring branch is realized, consisting of the corresponding components C «2> Rk2> 0l2, Ra2 and the n-p-n transistors T5 and T6. The values of the resistors Rai and Ra2 are in the order of 10 kOhm. The transition frequency of all transistors is above 500 MHz.

At the connection point of transistor T4 and resistor Rai as well as of transistor T6 and resistor Ra2, a series capacitor Cm and Cr2 is connected, the counter connections of which are connected to one another. At the connection point of the components Cri, Rr2, the coaxial output socket K3 is connected, on the one hand via which the signal is forwarded to the subsequent signal processing stage SV and, if necessary, a supply of the operating voltage of approximately 8 to 10 V is ensured by then connecting it with a contact via a resistor Rr, the size of about 1 kOhm, to the positive (+), and with the other contact via the isolating resistor Rt to the negative (-) operating voltage line.

At the end of the output signal line connected to the output signal socket K3 in the control room, the signal processing stage SV is connected via a capacitor and the power supply via a choke. The outer conductor of the output signal line is clamped to the reference potential (ground).

The smoothing capacitor Cg is designed as a ceramic disk capacitor and has a capacitance of approximately 100 nF. It is connected to avoid oscillation tendencies and asymmetrical modulation in the event of short-term pulses directly to the connection point of the emitter of transistor Ti with the negative operating voltage (-) and to the connection point of the emitter of transistor T2 with the positive operating voltage (+), with the supply line length on both sides of the Capacitor Cg is to be limited to less than 2 cm.

The principle of operation of the circuit according to FIG. 4 is as follows:

The signal adapter AD in conjunction with the coaxial cable KL limits the maximum duration of the TE signals detectable on the test vessel MG in such a way that the highest measuring sensitivity is achieved with pulse rise times in the nanosecond time range. The short-circuit line KL in particular fulfills this purpose, unless a fictitious short-circuit for long-term signals already exists due to the contact to the metallic vessel MG of the test object. As a result, the useful-to-interference signal distance can be considerably improved, since usual interference signals include are characterized by rise times in the range of a few 100 to 1000 ns and are therefore significantly longer than the above. Value of a few ns for maximum measurement sensitivity.

By counter-switching the transistors Ti and T2 while simultaneously connecting the bases of these transistors via a high-ohm resistor (> 1 MOhm), this transistor arrangement is to a certain extent in the "slumber" state, i.e. the current consumption is far below 1 mA. If a short-term pulse of a duration of a few ns occurs, the voltage amplitude of which exceeds approximately 1 mV, a charge portion is taken over by the storage capacitor Csi when the pulse is positive and injected into the base of the transistor Ti. After transient charge storage in transistor Ti, a collector current surge is triggered in transistor Ti after the input pulse has almost subsided. The transistor T2 remains in the “slumber” state at the positive polarity under consideration, i.e. no additional current flow is excited by transistor T2. Accordingly, a voltage change appears at the connection point of the collectors of the transistors Ti and T2, which essentially results from the discharge of the storage capacitor Cli, given by the

5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

9

CH 679 807 A5

10th

Current path from the collector of the transistor Ti via the coupling capacitor Cki and a limited avalanche breakdown of the emitter-base path of the transistor T3 to the charging capacitor Cli. In contrast, the reverse-biased transistor Ts remains in the blocked state at the polarity under consideration. Corresponding to the current amplification and thus also the charge amplification of transistor Ti, a signal amplitude is reached at the charging capacitor Cli which exceeds the input signal, the duration of the signal at the charging capacitor Cli also being strong as a result of the termination of the avalanche breakdown and the re-starting blocking phase of transistor T3 is stretched. This creates a rapidly rising and almost linearly falling signal edge, i.e. a triangular pulse is obtained, which contains the square value of the pulse charge in its area. This parameter is advantageously suitable for alerting a critical TE intensity in the sense of anomaly lie early detection.

Analog conditions exist with opposite polarity of the input signal. In this case, transistor Tz becomes active and transistor Ts suffers the limited avalanche breakdown, while transistor T3 remains in the blocking state.

When using transistors with a crossover frequency of more than 500 MHz, this memory effect proportional to the input signal can be demonstrated down to the sub-nanosecond time range.

It should be noted that this circuit arrangement according to the invention differs significantly from the “classic broadband amplifier”, in which, in contrast to the circuit in question, the working resistances are deliberately kept small (order of magnitude <100 ohms for switching times in the nanosecond range) and also the operating current of each transistor in the range of approx. 10 mA.

The peak value storage according to the invention with transistors T3 and Ts also differs from conventional circuit arrangements. A rectifier effect already takes place with a voltage swing of only 10 mV, which is retained with the transistors used with a crossover frequency of over 500 MHz even with switching times of a few ns, i.e. the blocking delay time is extremely short. In the case of classic rectifier circuits, however (including avalanche diodes), the response threshold increases 10 to 50 times, so that instead of a dynamic range of 1: 500, only values from 1:50 to 1:10 at the specified operating voltage of Ub s 10 V can be reached.

The physical effect is based on the avalanche breakdown of the emitter-base path of the transistors T3, which is biased in the reverse direction; Ts. It is particularly advantageous that no fixed preload has to be set by means of a voltage divider. In contrast, in the case of classic rectifier circuits, an operating point setting is only possible on the one hand by means of voltage dividers, which increases the current requirement of the circuit. In addition, this measure is only effective when a diode current is excited, so that the blocking characteristic deteriorates.

Due to the simplicity of the circuit arrangement, in particular due to the small number of components, the sensor can also be implemented using integrated circuit technology. This also provides hermetic encapsulation so that the sensor can be installed not only outside but also inside electrical equipment (e.g. under oil). This results in a further improvement in the signal-to-noise ratio, since on the one hand the PD signals no longer have to penetrate the metallic encapsulation with appropriate damping and on the other hand this encapsulation as a Faraday cage considerably dampens the interference signals inside. By installing several sensors in high-quality equipment (e.g. transformers), a multi-point measurement can also be advantageous for fault location.

The small size of the sensor is also advantageous for the precise localization of TE-damaged bar windings of large motors and generators, as has been demonstrated by experimental studies.

It should also be noted that the sensor can of course also be operated with an internal battery, whereby an operating time of at least one month can certainly be achieved. This period of time appears to be sufficient to keep damaged equipment in operation for a certain period of time until an exchange unit is procured. With permanent condition monitoring, total failure with all consequential damage can be avoided.

If the detection limit for TE detection is to be increased, it is advisable to expand the circuit with a pre-stage, which is analogous to the input circuit in accordance with. Fig. 4 is realized with the components Csi, Cs2, Rb, Ti and Tz.

The circuit arrangement according to the invention can also be combined with an active broadband antenna as a pickup. The well-known three-electrode system with differential amplifier arrangement is preferably used as the active broadband antenna.

Digital technology lends itself to further signal processing at the headquarters. For this purpose, it is recommended to start with an analog pulse processing when using circuit components of the TE sensor. An exemplary embodiment is shown in FIG. 5. This circuit arrangement, in which the emitter T4; Operational amplifiers connected as voltage followers are used to generate unipolar triangular pulses from bipolar needle pulses, which, using a threshold value trigger, allow amplitude-time conversion and thus digitalization in a simple manner.

If the entire partial discharge sensor is battery-fed at high voltage potential, the transmission of the amplified and shaped TE signals to the evaluation devices is preferably by means of focused light or laser beam or

5

1Q

15

20th

25th

30th

35

40

45

50

55

60

65

6

11

CH 679 807 A5

12

to be carried out using high-voltage-resistant fiber optic cables. The partial discharge sensor according to FIG. 5 is connected with its signal output K1 to a modulation input of a light transmitter, not shown, which is also at high voltage potential. Partial discharge sensor and light transmitter can then preferably also be arranged in the interior of encapsulated, oil or inert gas-filled electrical equipment at high voltage potential, directly at the possible point of origin of partial discharges.

FIG. 7 shows a simple analog circuit arrangement for obtaining the mean square value from the unipolar triangular pulses, as are supplied by the circuit arrangement according to FIG. 5 to the output socket K1. The evaluation of this mean square value is a good and cheap alternative to the digital evaluation and, in deviation from the applicable standards (IEC 270), allows a better assessment of a critical insulation fault, since large TE pulses are particularly emphasized by the square value and also a dense one by averaging Sequence of TE impulses also lead to a high output signal.

The circuit arrangement preferably consists of two OP amplifier stages, which have a high pass Ci; Ri are connected to the output socket K1 of FIG. 5. The decoupling via the high-pass capacitor Ci prevents lower-frequency noise or shifting of the operating point of the upstream stage from shifting the output signal level and thus falsifying the measured values. The time constant of the high pass Ci * Ri is approx. 1 ms.

With the diode D1 and the output resistor R2 in the feedback branch of the OP amplifier, the first stage effects a compounding of the zero level, which would otherwise shift depending on the pulse train. The second OP stage represents an active integrator whose integration capacitor Ci in the feedback branch forms a time constant Ci * R3 approx. 10 ms with the series resistor R3 and a time constant Cj * R4 approx. 50 ms with its adjustable parallel resistor R4.

The signal level thus formed corresponds to the mean square value of the TE pulse charges.

The resistor-capacitor-resistor series circuit on the output side, in which the output socket K2 is connected via the capacitor, represents a further passive integration element which, with suitable dimensions, can also take over the function of the active integrator, so that the second OP stage in one economic variant can be omitted.

In addition, due to the low-pass behavior, this passive integration element prevents the TE sensor from being influenced by external high-frequency interference signals which, due to the antenna effect, could be coupled into the output socket K2 from the environment via the output signal line.

Claims (19)

Claims 1. Partial discharge sensor, in particular for on-site diagnosis, early detection of anomalies and protection against accidents on electrical equipment, without their activation, with a sensor for partial discharge signals and a processing unit, characterized in that - That the transducer represents an adapter (AD) adapted to the respective test object, which is connected via a signal line (SL) and two charge storage capacitors (Csi, Cs2) each to the bases of two complementary bipolar transistors (T1; T2) that are interconnected by means of high-resistance (Rb) is - That their emitters are each connected to a pole of an operating voltage (+; -) and their collectors are connected to each other and via a coupling capacitor (Cki; Ck2), each has a complementary emitter-base path (T3; T5) and a charging capacitor (Cli; Cl2) are connected to one pole of the operating voltage, - In addition, each connection of the two charging capacitors (Cli; Cl2) is connected to the base of emitter followers (T4; T6) which are complementary to one another, - That in each case the connection point of the coupling capacitor (Cki; C «2) and the emitter-base path (T3; T5) is connected via terminal resistors (R« i; Rk2) to the pole of the operating voltage at which the collector of the associated emitter follower (T4; T6) and - That of the two load resistors (Rai; Ra2) in the emitter branches, series capacitors (Cri; Cr2) couple the amplified and stretched TE signals to an output socket (K3) to which a signal processing stage (SV) is connected via an output signal line . 2. Partial discharge sensor according to claim 1, characterized in that the adapter (AD) consists of two contact tips (MP1; MP2) with a distance determining the active measuring zone of a metallic test specimen vessel (MG). 3. Partial discharge sensor according to claim 1, characterized in that the adapter is a waveguide consisting of a coaxial line, which is wound on a highly permeable core, for pulse transmission, a first input (MP1) with the beginning of the inner conductor (IL1) and a second input (MP2) is connected to the end of the inner conductor (IL1) and the signal line (SL) is connected to the beginning of the outer conductor (AL1) and the end of the outer conductor (ALi) to the housing of the sensor. 4. Partial discharge sensor according to claim 1, characterized in that a contact of the output socket (K3) of the processing unit (VE) via a resistor (Rr) with the positive (+) and the other contact of the output socket (K3) via a separating resistor ( Rt) is connected to the negative (-) operating voltage line, whereby 5 10th 15 20th 25th 30th 35 40 45 50 55 60 65 13 CH679 807A5 14 at the end of the output signal line in the control room next to the signal processing stage (SV) a power supply is choked. 5. Partial discharge sensor according to claim 1, characterized in that a low-inductance smoothing capacitor (Cg) is connected to the operating voltage potentials (+; -) via the connection points of the transistor emitters (T1; T2). 6. Partial discharge sensor according to claim 1, characterized in that between the signal line (SL) and the bases of the two complementary bipolar transistors (T1; T2) a similarly constructed second complementary transistor arrangement with charge storage capacitors is inserted as a preliminary stage. 7. Partial discharge sensor according to claim 1, characterized in that the sensor consists of an active broadband antenna, preferably a three-electrode system with differential amplifiers. 8. Partial discharge sensor, in particular for on-site diagnosis, anomaly early detection and accident protection on electrical equipment, without their activation, with a sensor for partial discharge signals and a processing unit, characterized in that - That the transducer represents an adapter (AD) adapted to the respective test object, which is connected via a signal line (SL) and two charge storage capacitors (Csi; Cs2) each to the bases of two complementary bipolar transistors (T1; T2) that are interconnected by means of high-resistance resistors (Rb) is - That their emitters are each at a pole of an operating voltage (+; -) and their collectors are connected to each other and via a coupling capacitor (Cki; Ck2), each have a complementary reverse emitter-base path (T3; T5) and a charging capacitor (Cli; Cl2) is connected to one pole of the operating voltage, that the connection point of the coupling capacitor (Cki; Ck2) and the emitter-base path (T3; T5) are each connected to one pole of the operating voltage via clamping resistors (Rki; Rk2), - That the charging capacitors (Cli; Ci_2) are each connected to a voltage follower, consisting of OP amplifiers, the outputs of which are connected to the inputs of a further OP amplifier, which generates unipolar voltage pulses in the case of bipolar input pulses and whose output to one Output socket (K1) is connected to which a signal processing stage (SV) is connected via a signal connection. 9. partial discharge sensor according to claim 8, characterized in that the output socket (K1) is connected to the modulation input of a downstream light transmitter or laser arrangement, the light receiver with demodulator upstream of the signal processing stage (SV) via a light guide cable or a fo-focused beam is coupled. 10. Partial discharge sensor according to claim 8, characterized in that a measurement value is formed in the signal processing stage (SV) which corresponds to the mean value of the square of the TE pulse charges. 11. Partial discharge sensor according to claim 10, characterized in that an OP amplifier with its non-inverting input (+) at reference potential (Ubo) and with its inverting input (-) via a high-pass capacitance (Ci) at the output (K1) Processing unit (VE), are connected via a high-pass transverse resistor at reference potential (Ub-), via a diode (D1) in parallel with a resistor (R2) to the output of the OP amplifier and to the input of an integrator. 12. Partial discharge sensor according to claim 11, characterized in that the integrator consists of a resistor-capacitor-resistor series circuit, the output signal (K2) being tapped via the capacitor. 13. Partial discharge sensor according to claim 12, characterized in that an OP amplifier connected as an active integrator is arranged with the resistance-capacitor-resistance series circuit. 14. Partial discharge sensor according to claim 8, characterized in that the adapter (AD) consists of a single-pole contact to the test object (PO) and the conductive housing (SG) of the sensor forms an antenna. 15. Partial discharge sensor according to claim 8, characterized in that the single-pole contact in ferromagnetic test specimen vessels (MG) takes place by means of a holding magnet (HM). 16. Partial discharge sensor according to claim 8, characterized in that the electrically conductive housing (SG) of the sensor is electrically and mechanically connected to the housing (MG) of the test object (PO) and a preferably disc-shaped antenna (SR) is arranged as the opposite pole is. 17. Partial discharge sensor according to claim 8, characterized in that a short-circuited waveguide with a wave propagation time of less than 100 ns is connected from the signal line (SL) to the sensor housing (SG). 18. Partial discharge sensor according to claim 17, characterized in that the waveguide consists of a coaxial cable (KL), the inner conductor of which is connected on the input side to the signal line (SL) and on the output side to the metallic housing (SG) of the sensor. 19. Partial discharge sensor according to claim 17, characterized in that the waveguide consists of a parallel strip line. 5 10th 15 20th 25th 30th 35 40 45 50 55 60 65 8th