DE102010029496A1 - Method for energy supply of a load in a deposition process, comprises providing first and second electrical variables by a first and second controllable electric ignition means, where the electrical variables differ itself in its amounts - Google Patents

Abstract

The method for energy supply of a load (5) in a deposition process, comprises providing first and second electrical variables by a first and second controllable electric ignition means (4a, 4b), where the electrical variables differ itself in its amounts and are switchable alternative to the first or the second electrical variables to the load. The energy supply is carried out in cycle with a given time duration, where each cycle comprises a determined number of entire oscillations of the base frequency of the supply voltage. The method for energy supply of a load (5) in a deposition process, comprises providing first and second electrical variables by a first and second controllable electric ignition means (4a, 4b), where the electrical variables differ itself in its amounts and are switchable alternative to the first or the second electrical variables to the load. The energy supply is carried out in cycle with a given time duration, where each cycle comprises a determined number of entire oscillations of the base frequency of the supply voltage that is interconnected to a first part number of the first electrical variable and a second part number of the second electrical variable for the energy supply of the load. The first and second electrical variables are current or voltage. The sum of the first electrical variable to begin a cycle is less than the sum of the second electrical variable at the end of the cycle. The sum of the first and second part number corresponds to the determined amount. Informations are considered during the energy supply of the load over cycle, switch-on times and the ignition means of other load.

Description

The invention relates to a method for supplying energy to a load in a deposition process in which at least a first and a second electrical variable are provided by means of a first and a second controllable electrical switching means, wherein the electrical quantities differ in their amounts and alternatively the first or the second electrical variable second electrical variable to the load is switchable.

In deposition systems, converter plants or reactors preferably in the field of silicon production by means of a CVD process (chemical vapor deposition), the energy supply is usually carried out by means of a so-called phase angle control method (phase angle control).

In this case, a desired value of an electrical variable to be controlled for a load is specified for a phase-angle control. The control is effected by connecting or disconnecting a plurality of controllable electrical switching means connected in parallel with different output voltages or currents, which are arranged in a circuit between an AC voltage generator and the load, for example the silicon rods to be coated.

To be able to cover a large adjustment range by means of such a phase control, several taps of a transformer are used. A disadvantage of this solution are the unwanted disturbances in the supply network, which are caused by the phase control itself. One way to operate this process Netzrückwirkungsarm is to always switch two windings within a network half-wave to the load. Such voltage sequential control systems are also referred to as "voltage sequence control" or "multi-step control".

In a deposition system for silicon deposition, in which a CVD process is to take place, for example, according to the Siemens method, usually several pairs of rods made of silicon are juxtaposed. Rod pairs are understood to mean upright silicon thin rods, which are electrically connected to one another at the upper free end, forming a pair. The lower ends of the rod pairs are connected via plug connections to a voltage source. These pairs of rods, which represent the load to be supplied, must be brought to a temperature at which the deposition process from the gas phase becomes possible. In this process, polysilicon is deposited on the pairs of rods, whereby a progressive increase in thickness of the rod pairs takes place.

The starting material, so-called silicon thin rods, as used in the process, are very high-impedance. The heating of the silicon thin rods takes place electrically in the sense of resistance heating. In order to work with technically sensible voltages, first of all a high voltage is used in a parallel connection of the silicon thin-rod pairs. Each rod pair receives the same full voltage. Since the currents are only a few amperes in this case, this is technically easy to control.

Various methods for powering a deposition process are known in the art.

The so-called ramp method works with a constantly changing phase angle to the measurement technique, which after the Standard EN 61000-4-7 is realized pretend a smaller reading.

Another possibility is to operate by means of a synchronized method in which the stages and switching times of a normal oscillation packet control are synchronized in order not to simultaneously switch all systems on or off.

A third possibility is the use of the KAPAS method (combined phase-angle control), in which a combination of phase control and full-cycle clock is used. Here, a cycle consists of mains periods of equal amplitude of a voltage U or a current I with truncated or fully controlled periods.

The disadvantages of these methods according to the prior art are a high outlay in the field of circuit technology, especially for reducing the disturbances of the supply network generated by the circuit itself, the so-called harmonic harmonics.

Harmonics of the grid frequency can have adverse effects on other consumers and the utility grid, for example, through resonances. Therefore, low-backlash circuits are important for such high power power supply assemblies. A construction of compensation systems which is necessary for the reduction of the disturbances is complex, both during the installation as well as during the ongoing maintenance and control.

Other disadvantages are the high cost of the necessary calculation and measurement methods to reduce the disturbances and a necessary approval process for operating the power supply arrangement at the responsible energy suppliers.

The invention has for its object to provide a method for powering a deposition process, with which a reduction of the disturbances of the supply network is achieved.

According to the invention, in a method of powering a load in a deposition process, the object is achieved by powering cycles in a predetermined time T _o , each cycle comprising a fixed number n of whole oscillations of the fundamental frequency f of the supply voltage for a first part number n ₁ the first electrical quantity and for a second part number n ₂ the second electrical quantity for supplying energy to the load is switched through.

According to the invention, the power supply is made in cycles. A cycle here represents a period of time which repeats periodically. The cycle length is not arbitrary, but depends on the fundamental frequency f of the supply or mains voltage. The cycle lengths are always an integer multiple n of the period T after T = 1 / f of the fundamental frequency f, where n ≥ 2.

The essential difference from the described prior art is that in each case only whole oscillations of the network or fundamental frequency f are used to supply energy to the load, wherein the amplitudes of the whole oscillations may be different. In the case of using two oscillations of different amplitude in one cycle, the switching from the first electrical quantity to the second electrical quantity occurs at the zero crossing of the oscillation. Thus, no disturbing network perturbations occur as known from the prior art.

According to the method, at least a first electrical quantity and a second electrical quantity, which can be, for example, a voltage U or a current I, are provided. For better energy supply, in practice, for example, more than two voltages are provided, since in this way the control range and the resolution, ie the accuracy with which a desired value can be set, increases the energy supply. In practice, usually transformers are used, which have on their secondary side the desired number of taps. For example, the voltages of these taps are at least indirectly the first, second to nth electrical size.

A setpoint value predetermined by the superordinate control, which corresponds, for example, to a current I flowing through the load, is generated by means of this control by selecting one or two required electrical variables (currents I ₁ and I ₂ ) for the energy supply in the current cycle.

If the desired value of I corresponds to one of the two currents I ₁ or I ₂ , only the corresponding current, for example I _{2, is switched} through to the load.

In most cases, the setpoint value of I lies between the two currents I ₁ and I ₂ . In this case, the controller makes the required distribution between the currents I ₁ and I ₂ . determined. In this case, the distribution is understood as meaning a first part number n ₁ and a second part number n ₂ of whole oscillations, which together correspond to the specified number n of whole oscillations of the fundamental frequency f.

With an appropriate distribution of n ₁ and n ₂ , the rms current I _eff corresponds to the setpoint of I. The calculation of the effective current I _eff is shown below.

The distribution of the specified number n of whole oscillations into the number of parts n ₁ and n ₂ can, for example, take place at n = 5 according to the following distributions:
n ₁ = 0 and n ₂ = 5; n ₁ = 1 and n ₂ = 4; n ₁ = 2 and n ₂ = 3; n ₁ = 3 and n ₂ = 2; n ₁ = 4 and n ₂ = 1 or n ₁ = 5 and n ₂ = 0.

In one embodiment of the invention, it is provided that the first and the second electrical variable is a current I or a voltage U.

The first and second electrical quantities are, as required, a current I or a voltage U. This is determined, for example, by the electrical variable to be controlled at the load.

In an advantageous embodiment of the invention it is provided that no power supply of the load for a period of time T _P occurs in the event that the amount of the first electrical variable at the beginning of a subsequent cycle is less than the amount of the second electrical variable at the end of the previous cycle ,

This brief interruption, for example, of the power supply to the load prevents a short circuit in the transition from the higher to a lower voltage at the zero crossing of the oscillation. In practice, a break time of 5 to 15 ° el is usually sufficient.

In a further embodiment of the invention, it is provided that the sum of the first part number n ₁ and the second part number of n ₂ corresponds to the specified number n.

The number of whole oscillations n corresponding to a whole cycle of duration T _o is subdivided into a first part number n ₁ and a second part number n ₂ .

In one embodiment of the invention it is provided that in the power supply of the load information about cycles, turn-on and related electrical switching means of other loads are taken into account.

In practice, a plurality of loads are arranged in a reactor in the silicon production which are each controlled by a working according to the inventive control unit. To avoid a simultaneous switching on, for example, the respective highest voltage or the highest current through the control units, the control units exchange information with each other about their operating conditions such as cycle lengths, switch-on, switched electrical switching means and others and can thus a time-delayed switching and thus a further reduction of the flicker effect guarantee.

The invention will be explained in more detail with reference to an embodiment. In the accompanying drawings shows

1 a schematic diagram of a phase control with multiple voltage-time diagrams for powering a deposition process and

2 a diagram of the voltage-time course when using the method for power supply according to the invention.

In 1a a principle circuit arrangement for supplying voltage to a deposition process is shown. One from a primary winding 2 and a multiple taps having secondary winding 3 existing transformer 1 is the input side with a supply network, not shown, which operates, for example, with a voltage of 230 V ~ or more connected.

At the secondary side 3 of the transformer 1 are two controllable electrical switching devices 4a and 4b arranged, which with a first connection of a load to be supplied 5 are connected. As a controllable electrical switching means usually antiparallel connected thyristors or triacs are used, which are controlled by a supply circuit, not shown. Weight 5 is with a second connection to the secondary winding 3 of the transformer 1 connected.

Phase gating and partially controlled setting circuits are in the book of W. Keuter: "The switching and switching of alternating current" and are also known on the market as voltage sequence control (SFS) or voltage sequence controll (VSC).

In this control, first close to the zero crossing, a first voltage 6 switched to the load, which is too small for the task, with respect to the parameters current, voltage, power. Within each half-cycle, the next higher, too high second voltage is reached 7 switched over, as in the 1b shown.

The 1c and 1d set the respective switching range of the two controllable electrical switching means 4a and 4b represents.

The resulting rms value corresponds to the actually required value at the load 5 ,

For such a power supply are a plurality of transformer taps and a plurality of controllable electrical switching means 4 necessary.

Due to this type of power supply, the unfavorable characteristics of a pure phase-angle control (PA) can be leased, i. H. Less reactive power is generated which reduces harmonic content while increasing resolution.

When starting the deposition process with the thin rods is, if you are working with the mains frequency, because of the dynamics no other operating mode than the phase control method possible.

If the deposition process has progressed further, the thermal inertia of the system is large enough to operate according to the invention with another network-friendly operating mode according to the full-cycle clock principle.

The control device, not shown, which converts the method according to the invention switches over automatically or after a manual input into this operating mode, if this is enabled by the user.

The Vollschwingungstaktverfahren works without a phase control. Thus, harmonics of the line frequency are not generated, which is particularly important in the course of a deposition process in which large currents must be controlled, since harmonics of the grid frequency have deleterious effects on other consumers and the utility grid.

When using the method for a CVD process, it is advantageous to use a phase-angle method in the area of ignition with thin rods and to switch to the operating mode full-scale cycle method after reaching a predefined thickness of the silicon rods.

For example, if only the two illustrated electrical switching means 4a and 4b operate, the calculated by the controller, not shown control angle of the switching means is converted into a duty cycle. Such a power supply for a deposition process usually has more than two stages, but this has no effect on the implementation of the method according to the invention.

When controlling, it must be ensured that when switching from a higher voltage level to a lower voltage level, a pause time of approx. 5-15 ° el (load angle, 0-90 ° el limit view) must be observed to prevent a short circuit.

When changing from a low voltage level to a higher voltage level, switching can be done easily at zero voltage.

Due to the low voltage difference, the resulting flicker effect is insignificant.

The total time duration T _o , which is also to be referred to as the playing time, comprises a preselected variable number n of oscillations of the used AC frequency f. The choice of the number n depends on the thermal inertia of the system and should be chosen such that no noteworthy temperature change can be detected on the silicon rods glowing by the current flow. Time T _o can be set to be in the range between 100 ms and more than 10 seconds.

mit T_S₂ = T_o – T_S₁.The rms value of the current I _{eff is} calculated by a full oscillation cycle

with T _S₂ = T _o - T _S₁ .

The performance is then: P = P ₁ + P ₂ P ₁ = P _o₁ · T _S₁ / T _o P ₂ = P _o₂ · T _S₂ / T _o

I _o , U _o and P _o are the uninfluenced values of the stages 4a and 4b ,

T _o is the total duration of a cycle comprising the turn-on time T _S₁ and T _S₂ the respective stage 4a or 4b within a switching cycle T _o .

In the example, the Amplitute is the first stage 1 So for example 1 A (or 1 V) as in the 2 shown. The voltage of the second stage should be 1.5V. In the example n = 5, n ₁ = 4 and n ₂ = 1.

Thus, the RMS of the currents I _{eff ₁} , I _{eff ₂} and I _eff calculated after

The performances P ₁ , P ₂ and P are calculated according to

The selection of the required voltages or currents is carried out as described above by a higher-level control, so that each intermediate value between two voltages or currents can be achieved by a corresponding distribution of n ₁ and n ₂ .

LIST OF REFERENCE NUMBERS

1

transformer

2

primary

3

secondary winding

4

controllable electrical switching means

5

load

6

first tension

7

second tension

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited non-patent literature

• Standard EN 61000-4-7 [0008]
• W. Keuter: "The switching and switching of alternating current" [0037]

Claims (5)

A method of powering a load in a deposition process, wherein at least first and second electrical quantities are provided by first and second controllable electrical switching means, wherein the electrical quantities differ in magnitude and, alternatively, the first or second electrical quantities Load is switchable, characterized in that the power supply in cycles with a predetermined time T _o is carried out, each cycle a fixed number n of whole oscillations of the fundamental frequency f of the supply voltage comprises that for a first part number n _1, the first electrical variable and for a second part number n _2, the second electrical quantity for supplying power to the load is switched through. A method according to claim 1, characterized in that that the first and the second electrical variable is a current I or a voltage U. A method according to claim 1 or 2, characterized in that for a period of time T _P no power supply of the load takes place in the event that the amount of the first electrical variable at the beginning of a subsequent cycle is smaller than the amount of the second electrical variable at the end of the preceding Cycle. Method according to one of claims 1 to 3, characterized in that the sum of the first part number n ₁ and the second part number of n _{2 of} the specified number n corresponds. Method according to one of claims 1 to 4, characterized in that in the power supply of the load information about cycles, turn-on and related electrical switching means of other loads are taken into account.

Images