Matching of pulse forming network and pulse transformer parameters in pulse modulator circuits for a klystron power supply
Збережено в:
| Дата: | 1999 |
|---|---|
| Автори: | , |
| Формат: | Стаття |
| Мова: | English |
| Опубліковано: |
Національний науковий центр «Харківський фізико-технічний інститут» НАН України
1999
|
| Назва видання: | Вопросы атомной науки и техники |
| Онлайн доступ: | http://dspace.nbuv.gov.ua/handle/123456789/81527 |
| Теги: |
Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|
| Назва журналу: | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| Цитувати: | Matching of pulse forming network and pulse transformer parameters in pulse modulator circuits for a klystron power supply / I.V. Kazarezov, A.A. Korepanov // Вопросы атомной науки и техники. — 1999. — № 4. — С. 44-46. — Бібліогр.: 3 назв. — англ. |
Репозитарії
Digital Library of Periodicals of National Academy of Sciences of Ukraine| id |
irk-123456789-81527 |
|---|---|
| record_format |
dspace |
| spelling |
irk-123456789-815272015-05-18T03:02:03Z Matching of pulse forming network and pulse transformer parameters in pulse modulator circuits for a klystron power supply Kazarezov, I.V. Korepanov, A.A. 1999 Article Matching of pulse forming network and pulse transformer parameters in pulse modulator circuits for a klystron power supply / I.V. Kazarezov, A.A. Korepanov // Вопросы атомной науки и техники. — 1999. — № 4. — С. 44-46. — Бібліогр.: 3 назв. — англ. 1562-6016 http://dspace.nbuv.gov.ua/handle/123456789/81527 en Вопросы атомной науки и техники Національний науковий центр «Харківський фізико-технічний інститут» НАН України |
| institution |
Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| collection |
DSpace DC |
| language |
English |
| format |
Article |
| author |
Kazarezov, I.V. Korepanov, A.A. |
| spellingShingle |
Kazarezov, I.V. Korepanov, A.A. Matching of pulse forming network and pulse transformer parameters in pulse modulator circuits for a klystron power supply Вопросы атомной науки и техники |
| author_facet |
Kazarezov, I.V. Korepanov, A.A. |
| author_sort |
Kazarezov, I.V. |
| title |
Matching of pulse forming network and pulse transformer parameters in pulse modulator circuits for a klystron power supply |
| title_short |
Matching of pulse forming network and pulse transformer parameters in pulse modulator circuits for a klystron power supply |
| title_full |
Matching of pulse forming network and pulse transformer parameters in pulse modulator circuits for a klystron power supply |
| title_fullStr |
Matching of pulse forming network and pulse transformer parameters in pulse modulator circuits for a klystron power supply |
| title_full_unstemmed |
Matching of pulse forming network and pulse transformer parameters in pulse modulator circuits for a klystron power supply |
| title_sort |
matching of pulse forming network and pulse transformer parameters in pulse modulator circuits for a klystron power supply |
| publisher |
Національний науковий центр «Харківський фізико-технічний інститут» НАН України |
| publishDate |
1999 |
| url |
http://dspace.nbuv.gov.ua/handle/123456789/81527 |
| citation_txt |
Matching of pulse forming network and pulse transformer parameters in pulse modulator circuits for a klystron power supply / I.V. Kazarezov, A.A. Korepanov // Вопросы атомной науки и техники. — 1999. — № 4. — С. 44-46. — Бібліогр.: 3 назв. — англ. |
| series |
Вопросы атомной науки и техники |
| work_keys_str_mv |
AT kazarezoviv matchingofpulseformingnetworkandpulsetransformerparametersinpulsemodulatorcircuitsforaklystronpowersupply AT korepanovaa matchingofpulseformingnetworkandpulsetransformerparametersinpulsemodulatorcircuitsforaklystronpowersupply |
| first_indexed |
2025-07-06T06:32:20Z |
| last_indexed |
2025-07-06T06:32:20Z |
| _version_ |
1836878183365869568 |
| fulltext |
MATCHING OF PULSE FORMING NETWORK AND PULSE
TRANSFORMER PARAMETERS IN PULSE MODULATOR CIRCUITS
FOR A KLYSTRON POWER SUPPLY
I.V. Kazarezov, A.A. Korepanov
Budker Institute of Nuclear Physics, SB RAS, Novosibirsk, Russia
The most widespread circuit of high-power
microsecond pulse generation for a power supply of
various pulse microwave devices is the circuit of line
type modulator on the basis of an artificial pulse
forming network (PFN), switch and step-up pulse
transformer (PT). The efficiency of the modulator is
defined both by active losses in elements of the
generator and losses of energy at pulse rise and fall time
generation, the influence of energy losses associated
with pulse rise and fall generation being the most
essential. Pulse forming efficiency, taking into account
the given losses, is determined as follows:
∫
= tbot
top
dttitu
W
0
)()(
η
(1)
Here topW is an energy, dissipated in a load at a pulse
top generation (the pulse top is determined with the
given accuracy); bott is the pulse duration at the
bottom. It is clear from (1), that increase in energy
efficiency requires reducing the pulse rise and fall times
at the given pulse top duration. The basic parameters
having influence on the pulse rise and fall times are time
constant of a transformer circuit, PFN type, and number
of its cells. The paper deals with a choice of the
specified parameters.
TIME CONSTANT OF A TRANSFORMER
CIRCUIT AND CHOICE OF PFN TYPE
It is shown in [1] that with a sufficient accuracy
the processes of pulse rise and fall forming by PT may
be analyzed using the simplified circuit, shown on
Fig. 1. All parameters of the equivalent circuit are
reduced to the secondary winding. Here
1LsLstLs += is a sum of PT leakage inductance
Lst and inductance of a primary circuit 1Ls ,
2CCC T += is a sum of PT dynamic capacitance
TC and capacitance of a load 2C , 1R and 2R are
resistance of the generator and load, accordingly. The
time constant of a transformer circuit is determined in
view of parasitic parameters of the primary and
secondary circuits:
( ) ( )21 CCLsLst T ++=τ . (2)
Fig. 1: Equivalent schematic of a transformer circuit.
One of the basic issues of the pulse modulator
designing is a choice of PT transformation ratio or, for
the specified parameters of the switch, of PFN type. The
most widespread type of power microsecond PT's
windings is one layer winding with additive polarity. It
is known (see, for example, [1]), that dynamic
capacitance of such windings is proportional to
( ) 221~ nnCT − (3)
and the inductance does not depend on n at all. Thus, if
the influence of the primary circuit inductance 1Ls is
neglected, the dependence between the time constant τ
and PT transformation ratio n will be
2
2
1
1~ C
n
nc +
−τ . (4)
Here the constant 1c does not depend on n . This
relation shows, that the time constant of a transformer
circuit grows weakly at n increase. For example, at PT
transformation ratio variation from 14 down to 7 the
time constant τ can increase maximum by 1.08 times
(at zero load capacitance 2C ). However, in general it is
necessary to take into account the influence of 1Ls on
increase of pulse rise time. Inductance of a primary
circuit 2' 11 nLsLs = (the values reduced to the PT
primary side are marked with primes) mainly consists of
thyratron inductance and inductance of connections
between PFN, thyratron, and PT. The correct analysis of
dependence between value of '1Ls and n is hindered,
as this inductance does not linearly depend on a voltage,
therefore further, for simplicity, it is supposed, that
'1Ls does not depend on a primary voltage, so does not
depend on PT ratio. This assumption will give an upper-
bound estimate for τ dependence on n . Really, at
given '1Ls value a relation for time constant (2) can be
copied as follows:
( )
+
−⋅⋅+= 2
2
1
2' 11 C
n
ncnLsLstτ (5)
ВОПРОСЫ АТОМНОЙ НАУКИ И ТЕХНИКИ. 1999. № 4.
Серия: Ядерно-физические исследования (35), с. 44-46.
44
This equation shows that at large n the influence of
1Ls on time constant can be significant. Therefore
double PFN (DPFN, also called Blumlein PFN) with an
output voltage equal to charged one should be used for
PT turns ratio decrease and primary voltage increase. It
should be noted, that DPFN has the disadvantages:
The voltage polarity reversal on capacitors of the first
part of DPFN results in reduction of their lifetime;
Increased losses in the DPFN’s first part, connected
with increased duration and amplitude of a current,
passing through it;
High-voltage performance of a PT primary winding
and connections with DPFN becomes complicated;
More energy, than in the circuit with single PFN, is
disappeared at klystron during its breakdown; the
protection circuits become complicated;
At least double quantity of network elements is
required, that complicates and gets up DPFN
performance.
As it follows from the above, at given parameters
of the switch it is better to use a single PFN instead of
DPFN, if pulse distortions (the tightening of rise and
fall) are almost identical.
In terms of the generator, which supplies a
klystron for the project of the linear collider JLC [3], the
comparison of the pulse shape and energy efficiency for
the pulse formers based on single and double PFNs with
two PT turns ratios 14 and 7 accordingly was carried
out. Parameters of the pulse on a load are:
Voltage amplitude − 560 kV
Current amplitude − 503 A
Pulse top duration − 1.5 µs
PT leakage inductance and total capacitance of a
secondary circuit are taken from [2]: 120=Lst µH,
250=C pF. For simplicity during simulation the
magnetization current may be ignored. The number of
cells in both PFN circuits was chosen as 20 (in DPFN
2x20). The increase of capacitance C at PT ratio by the
factor of 2 can be neglected, as PT capacitance makes
only ~1/3 of the total capacitance C .
Table 1: Pulse shape efficiency in the circuits of single
and double PFN.
'1Ls , µH 0.4 0.5 0.6 0.8 1
glesinη (n=14) 0.782 0.78 0.77 0.757 0.749
doubleη (n=7) 0.777 0.776 0.775 0.764 0.752
Therefore during numerical simulation of two
PFN circuits they differ only in additional inductance of
a primary circuit 2'11 nLsLs ⋅= . The analysis results
of these circuits are given in Table 1. From the given
table it follows, that pulse shape efficiency in the single
PFN circuit glesinη is not worse than in the double PFN
doubleη down to primary circuit inductance values
11 ≈LsLs (at 14=n ) or 25.01 ≈LsLs (at
7=n ). Thus, in the case of identical efficiencies a
single PFN is favoured.
CHOICE OF PFN CELL NUMBER
The features of PFN designing for powerful
generators with PT are the fixed parameters of a
transformer circuit: its total inductance and capacitance.
To obtain high energy efficiency, PFN parameters (cells
inductance and capacitance) should be such that the rise
time is determined mainly by parameters of a
transformer circuit, but not by PFN parameters. The
influence of finite time of switch triggering on the pulse
rise time may be ignored, since in most cases thyratrons
have turn-on time several times smaller than pulse rise
time. The technique of PFN cell number choice can be
illustrated by the example of a single PFN. The circuit
of single PFN with the equivalent schematic of a
transformer circuit and resistive load is shown on Fig.2.
Fig. 2: Simulated circuit of a single PFN.
The energy efficiency η dependence of a single
PFN cells number N at various time constants of a
transformer circuit (the characteristic impedance of this
circuit was set approximately equal to the load
resistance) was obtained (Table 2) as a result of
numerical simulations of transient processes in the
given circuit.
ВОПРОСЫ АТОМНОЙ НАУКИ И ТЕХНИКИ. 1999. № 4.
Серия: Ядерно-физические исследования (35), с. 44-46.
44
Table 2: Energy efficiency dependence on SPFN cells
number at various values of transformer circuit time
constant.
ptτ η N
0.82 10
0.05 0.87 20
0.884 30
0.814 15
0.075 0.83 20
0.843 25
0.775 10
0.1 0.79 15
0.79 20
The top non-uniformity in this case was about
%1± . The data obtained allow us to choose the
number of cells necessary to reach the maximum
efficiency when the transformer circuit time constant is
fixed. As follows from Table 2, for any transformer
circuit time constant there is a value of cell number N ,
over which the energy efficiency does not practically
change. Therefore the choice of too large value of N is
not justified from the point of view of rise time and
efficiency. Moreover current load over the capacitors of
a network is increased when number of cells is
increased. It is connected with reduction of duration of
current pulses through capacitors at increase of cell
number.
Besides, increase of network cell number can
result in gain of influence of PFN capacitor parasitic
inductance Lc . The pulse shape efficiency falls
because of tightening of a pulse fall time. The 0LLc
value dependence of η was investigated by the
example of 10-cell single PFN, loaded on matched
resistive load. The non-uniformity of a pulse top was
maintained at a level of %1± . The results of these
simulations are shown on Fig.3. The inductance Lc
grows up N~ , and the cell inductance 0L falls
N1~ when cells number N increases. Therefore at
increase of N the drop of efficiency due to capacitor
inductance Lc will increase.
Fig. 3: The energy efficiency versus parasitic
inductance of 10-cells SPFN capacitors.
Thus, a conclusion may be drawn: the time
constant of a transformer circuit determines
unambiguously a number of PFN cells, at which a pulse
has the best characteristics.
REFERENCES
[1] Vdovin S.S. Pulse transformers design. “Energy”,
Leningrad, 1971, p. 148.
[2] Pulse transformer R&D for NLC klystron pulse
modulator. / M. Akemoto, S. Gold, A. Krasnyh and R.
Koontz, Stanford Linear Accelerator Center, SLAC-
PUB-7583, July 1997.
[3] M. Akemoto, S. Tokumoto, S. Anami, M. Mizuno.
Blumlein type of modulator at KEK. - Third Annual
Klystron-Modulator Workshop, June 23-25, 1998,
Workshop paper.
ВОПРОСЫ АТОМНОЙ НАУКИ И ТЕХНИКИ. 1999. № 4.
Серия: Ядерно-физические исследования (35), с. 44-46.
46
|