Charging systems
The choice of the charging system depends of the repetition
rate and the modulator system.
The following charge systems will be described:
Resistor charging
Chopper charging
Resonant free running charging
Resonant charging by command
Each system has its own type of leveling and stabilisation
control.
Power transport is levelled, regulated and
adjusted by a converter and step-up transformer. This is a fair new technique on this
power level. Today's chopper components are in the range of 30kV@30A, made up of a large
number of MOSFETs or IGBTs, laying parallel and in series, which are combined in a compact
low-inductance bank [1].
Due to the high frequency chopper components, the transformer can be
build to a small size with high efficiency. The pulse width of the chopper is influenced
by the maximum current in the chopper components and by the asked voltage in the
pulse-forming part.
Chopper charging to high voltage PFN
On this stage the pulse-forming takes place too. The pulse forming part
acts like a capacitor in a first approach. A transformer with a center tap on the primary
is used to use the full magnetising capabilities of the core. To handle the magnetising
energy there are two ways.
- During charging of the capacitor the energy is load to the capacitor by the rectifying
circuit. It is a full rectifying with center tap. By the use of a center tap the transformer
is always hard connected to a ground even by lost components. After switching off the chopper
pulse, by maximum chopper current or maximum pulse width, the opposite coil delivers most of
the magnetising energy into the pulse forming capacitor too.
- The chopper is stopped and the magnetising energy is dissipated if the capacitor has
reached its voltage level.
The chopper has to be switched on just before the expected klystron pulse,
so the losses of the stored energy are minimised. In fact this is a so called
'command charging'. The maximum pulse width, due to the transformer, and the minimum pulse
width, due to the maximum chopper current, dictates the pulse switching. The flowchart
of fig.5 shows the control flow of the chopper charger. The leakage inductance of the
transformer provids mainly the current risetime. If necessary the leakage inductance can
be increased with a dedicated reactor in series connection witch the chopper switches.
Simulation of the chopper
Flowchart of chopper charger control
[1]
Fast high voltage transistor switches - model series
HTS, Behlke Electronic GmBH, Germany
The charge takes place after each discharge
of the PFN. This is a no command charge (diode instead of a switching device) or free
running charge. The repetition rate is dictated by the charge choke, the PFN-capacitors
and the delay to the discharge command. During that time there is a certain voltage drop
by the leakage of the PFN, so the repetition rate is adjustable only within a small range.
To stabilize the PFN unit a de-Q-ing system has added[1] on a
secondary of the charging choke. Whenever the correct pfn-voltage has attained, the
de-Q-ing thyristor has triggered and the stored energy of the charge choke has stored
in the capacitor and dissipated in the resistor. Hereby it provokes a voltage drop on
the charging device which blocks the pfn-voltage. To reach a stable voltage on the PFN,
the de-Q-ing system can be triggered on the bases of;
Power-supply level.
To change the PFN-level we had to change the power supply level.
The de-Q-ing go with it and dissipate only the ripple of the power supply.
Minimum and maximum delay after the start of charging.
A too short delay means high dissipation in the de-Q-system, so the power supply
level has decreased to keep the PFN level by a reasonable dissipation.
A too long delay means no de-Q-ing, so no stable PFN voltage, so the power supply
level has increased.
In the example the charge time is about 1ms. The manufacturing of chokes
is laborious. Chokes are dissipating elements too but are stable, robust and reliable.
A linetype modulator with resonant charging.
Simulation of resonant charged modulator with
de-Q-ing.
[1]
G.N.Glasoe,
L.V.Lebacqz, "Pulse Generators", McGraw-Hill 1948
To explain a resonant charging by command
there is chosen for a complete solid state module as used in the MEA-modulator [1].
A solid state linetype modulator module
In this modulator the module is used with a cycle pulse transformer.
In this design is used a command charge system and a non-dissipating level system
(SLS = Stabilizing and Levelling System).
Just before a pulse the charging of the PFN takes place. The advantage
is that there is always the same leakage during a short as possible time. This means a
more stable output then a free running system independent of the repetition rate.
If the PFN is charged to the right level the SLS-thyristor is triggered.
After triggering of the SLS-thyristor the remaining energy of the charging coil together
with the nominal charge is stored in the capacitor of the SLS-system. The charge switch
is not conducting (anode lower voltage then cathode) and afterwards the energy is stored
back in the main power supply by the diode/thyristor. In such a case a not regulated power
supply can be used because there are only transfer losses.
simulation of a resonant charge system with
a SLS-sytem
Solid state linetype modulator module
45kg, 125x60x20cm, 1000V-500A-50us pulse, rep.rate <500Hz
[1] P.J.T.Bruinsma,
E.Heine et al., "An all solid state linetype modulator", IEEE Trans. on Nucl.Sci.
NS-20 1973
previous
--
next