For this reason, better stability will be obtained if controlled volt-seconds reset is used
rather than constant-current reset (particularly at light loads). Voltage reset will increment
the flux density B rather than the magnetization force H. This is more controllable in the
negative-slope region shown in Fig. 2.21.10.
For voltage reset, a fast-slew-rate, low-output-impedance voltage-controlled amplifier
is preferred. The decoupling capacitor C2 shown in Fig. 2.21.9 converts the current control
of Q1 to virtual voltage control at high frequencies, but degrades the transient response and
is a compromise solution only.
21.13 SATURABLE REACTOR DESIGN
The most difficult design decision is the selection of core material and core size. As previ-
ously discussed, this depends on the application, frequency, and required performance.
However, once the core selection has been made, the rest of the design procedure is
21.13.1 Core Material
The choice of core material is normally a compromise between cost and performance.
At low frequencies, there are many suitable materials, and the controlling factors will be
squareness ratio, saturation flux density, cost, and core losses. Also, at low frequencies,
core losses are less important, giving a wider selection. At medium frequencies, up to, say,
35 kHz, the core loss starts to be the predominant factor, and Permalloys, square ferrites,
or amorphous materials will be chosen. At high frequencies, above 50 kHz, the core loss
tends to become excessive, and the parameters of the cores rapidly degrade. Ferrite materi-
als are probably the best choice. (The author’s experience is limited to frequencies below
50 kHz at this time.)
For very high frequency operation, more than 75 kHz, better results may be found using
sine-wave converters and true magnetic amplifier techniques. Sinewave operation extends
the useful frequency range of a material. (Many of the/core losses are proportional to rates
of change of induction rather than to frequency per se.)
21.13.2 Core Size
The practical requirements of the power circuit usually determine the core size. In low-
voltage, high-current applications, the reactor winding may be three or four turns of large-
gauge wire, and the practical difficulties of winding this wire on the core will determine
the core size. In many cases the winding will be a continuation of the transformer second-
ary, and the same wire will be used. To minimize the turns on the saturable reactor, a large
flux density excursion, typically 300 to 500 mT, will normally be used; hence the core loss
will be relatively large compared with the copper loss for high-frequency operation.
A typical example of the single-ended forward saturable reactor regulator will be used
to demonstrate one design procedure.
21.14 DESIGN EXAMPLE
Consider a requirement for a 5-V, 20-A saturable reactor to operate at 35 kHz in the single-
ended forward converter shown in Fig. 2.21.4.
21. HIGH-FREQUENCY SATURABLE REACTORPOWER REGULATOR
21.14.1 Step 1, Select Core Material
From Table 2.21.1, suitable materials would be Permalloy, square ferrite, or Vitrovac 6025.
In this example, assume that cost is less important than performance, so that the best overall
material, Vitrovac 6025 amorphous material, will be used.
21.14.2 Step 2, Calculate the Minimum Secondary Voltage Required
from the Converter Transformer
The maximum “on” time is 50% of the total period, or 14.3 Ms at 35 kHz. When the
SR is fitted, there will be an unavoidable minimum delay on the leading edge of the
“on” pulse, as a result of the time required to take the core from B
, even when
the reset current is zero. Previous experience with the 6025 material using fast diodes
indicates that this delay will typically be 1.3 Ms. (The actual value can be calculated
when the turns, core size, and secondary voltage have been established.) Therefore,
the usable “on” period will be approximately 13 Ms. The minimum secondary voltage
required from the converter transformer, to develop the required output voltage can now
513 15 6
21.14.3 Step 3, Select Core Size and Turns
In this example, it will be assumed that the transformer secondary has been brought out as
a flying lead, and that this wire is to be wound on to the saturable reactor core to form the
winding. The core size is to be such that the winding will just fill the center hole. Further, it
will be assumed that the wire size on the transformer secondary was selected for a current
density of 310 A/cm2 and that 10 wires of 19 AWG were used. Assuming a packing factor
of 80%, the area required for each turn will be 19.5 mm2.
The next step is an iterative process to find the optimum core size. The larger the core
size, the smaller the number of turns required, but the larger the center hole size.
Consider a standard Series 2 toroid, size 25–15–10. From the manufacturer’s data, this
toroid has a core area of 50 mm2 and a center hole area of 176.6 mm2. The turns required on
this core (if the flux density change is to be 500 mT and the core is to control to full pulse
width) may be calculated as follows:
11 14 3
time that forward voltage is applied, Ms
$B flux density change, T
core area, mm2
The area required for six turns of 10 r 19 AWG at 80% packing density is 117 mm2, and
this will just fit the core center hole size.
At higher input voltages, the flux density excursion will be larger, but as the core can
support a total change of 1.8 T (
there is an adequate flux density margin.
C# Word - Insert Blank Word Page in C#.NET
how to rotate Word document page, how to delete Word page Add and Insert a blank Page to Word File in C#. following C# demo code to insert multiple pages of a delete page from pdf acrobat; delete pages from pdf in preview
21.14.4 Step 4. Calculate Temperature Rise
The temperature rise depends on the core and winding losses and the effective surface area
of the wound core. The core loss of Vitrovac 6025 at 35 kHz and 500 mT is approximately
150 W/kg. The weight of the 25 r 15 r 10 core is 17 g, so the core loss is 2.5 W. The copper
loss is more difficult to predict, as an allowance must be made for the increase in effective
resistance of the wire as a result of skin effect. With a multiple-filament winding of this
type, the F
ratio (ratio of DC resistance to effective ac resistance) is approximately 1.2,
giving a winding resistance of 0.0012 7 and a copper loss (I2R loss) of 0.48 W. Hence,
the total loss is approximately 3 W. The surface area of the wound core is approximately
40 cm2, and from Fig. 3.1.9, the temperature rise will be 55°C. Since much of the heat
will be conducted away by the thick connection leads, the actual rise will normally be
less than this.
1. Explain the basic principle of the saturable reactor regulator.
2. What are the desirable core properties for saturable reactors?
3. How does the saturable reactor delay the transmission of the leading edge of a secondary
4. How is the saturation delay period adjusted?
5. Why is the saturable reactor particularly suitable for controlling high-current outputs?
6. Why is constant-voltage reset preferred to constant-current reset in high-frequency satu-
rable reactor regulators?
7. Why are fast secondary rectifier diodes recommended for saturable reactor regulators?
VB.NET PDF: Get Started with PDF Library
Fill-in Field Data. Field: Insert, Delete, Update Field. a new PDF Document object with 2 blank pages Dim doc Create(2) ' Save the new created PDF document into delete pages from a pdf in preview; delete page from pdf
Most engineers will be very familiar with the general performance parameters of constant-
voltage power supplies. They will recognize that these power supplies have a limited
power capability, normally with fixed output voltages and some form of current- or
power-limited protection. For example, a 10-V 10-A power supply would be expected to
deliver from zero to 10 A at a constant output voltage of 10 V. Should the load current try
to exceed 10 A, the supply would be expected to limit the current, with either a constant
or a foldback characteristic. The well-known output characteristics of one such supply
is shown in Fig. 2.22.1.
FIG. 2.22.1 Output characteristics of a constant-voltage power sup-
ply, showing constant-current and reentrant-current protection locus.
22.2 CONSTANT-VOLTAGE SUPPLIES
From Fig. 2.22.1, the output characteristics of the constant-voltage supply will be recog-
nized. The normal working range for the constant-voltage supply will be for load resis-
tances from infinity (open circuit) to 1 7. In this range, the load current is 10 A or less. The
voltage is maintained constant at 10 V in this “working range.”
How to C#: Cleanup Images
returned. Delete Blank Pages. Set property BlankPageDelete to true , blank pages in the document will be deleted. Remove Edges or Borders. delete pages pdf; delete page in pdf
At load resistances of less than 1 7, the current-limited area of operation will be entered.
In a constant-voltage supply, this is recognized as an overload condition. The output voltage
will be decreasing toward zero as the load resistance moves toward zero (a short circuit). The
output current is limited to some safe maximum value, but since this is normally considered
a nonworking area, the characteristics of the current limit are not very closely specified.
22.3 CONSTANT-CURRENT SUPPLIES
The constant-current supply is not so well known, and therefore the concept can be a little
more difficult to grasp. In the constant-current supply, the previous constant-voltage
characteristics are reversed. Figure 2.22.2 shows the output characteristics of a typical
It should be noted that the controlled parameter (vertical scale) is now the output cur-
rent, and the dependent variable is the compliance voltage. The normal “working range”
is now from zero ohms (short circuit) to 1 7, and in this load range the output current is
At load resistances in excess of 1 7, a compliance-voltage-limited protection area is
entered. For the constant-current supply, this voltage-limited area would be considered an
overvoltage condition. Since this is normally a nonworking protection area, the output
voltage may not be well specified in this area.
22.4 COMPLIANCE VOLTAGE
The terms used to describe the operation of a constant-current supply are somewhat less
familiar than those used for the constant-voltage supply. For a variable constant-current
unit, the output current may be adjusted, normally from near zero to some maximum value
(simply described as the constant-current range).
FIG. 2.22.2 Output characteristics of a constant-current power supply, showing constant-
voltage compliance limits.
22. CONSTANT-CURRENT POWER SUPPLIES
To maintain the load current constant, the output terminal voltage must change in
response to load resistance changes. The terminal voltage range over which the output
current will be maintained constant is called the “compliance voltage.” This compliance
voltage usually has a defined maximum value.
In the example shown in Fig. 2.22.2, the compliance voltage is 10 V, and a constant cur-
rent of 10 A will be maintained into a load resistance ranging from zero to 1 7.
Constant-current supplies have limited applications. They will be used where currents
must be maintained constant over a limited range of variation in the load resistance. Typical
examples would be deflection and focusing coils for electron microscopes and gas spec-
Figure 2.22.3 shows the basic circuit for a constant-current linear supply. In this
example, a voltage-controlled current source is shown. This is an important concept, not
previously introduced. Just as constant-current supplies can be configured from voltage-
controlled current sources, so can constant-voltage supplies be configured from current-
controlled voltage sources. This concept has important implications for current sharing,
when constant-voltage supplies are to be operated in parallel.
FIG. 2.22.3 Example of a constant-current linear supply (basic circuit).
In this example, the load current returns to the supply via the low-value series resistor
. The current analogue voltage developed across this resistor is compared with the inter-
nal reference voltage by amplifier A1, and the series regulator transistor Q1 is adjusted
to maintain the voltage across R
constant. Thus the current in R
will be maintained
constant, and provided that the amplifier input current is negligible, the load current will
also be maintained constant, irrespective of load resistance, within the “compliance volt-
It should be noted from Fig. 2.22.3 that as the load resistance increases, the voltage
across the output terminals V
increases to approach the supply voltage V
. When V
Q1 is fully saturated and has no further control. Beyond this point, the current must start to
fall, and the output voltage will be defined by the characteristics of the header supply V
which is not regulated and hence is not well specified in this example.
Although it is possible to reconfigure a constant-voltage supply to give constant-current
performance, this is not recommended. To provide maximum efficiency and high perfor-
mance, the constant-current supply will have a very low reference voltage (typically less
than 100 mV), the internal current shunt must be highly stable, and internal current paths
must be well defined.
1. How do the general performance parameters of a constant-current power supply differ
from those of a constant-voltage power supply?
2. What is the meaning of the term “compliance voltage” in a constant-current supply?
3. What would be considered an overload condition for a constant-current supply? Compare
this with a constant-voltage supply.
4. Why is the output ripple and noise voltage a meaningless parameter for a constant-
5. How should output ripple and noise be defined in a constant-current supply?
The variable linear power supply, although perhaps somewhat out of place in a switching
power supply book, has been included here for several reasons.
First of all, when very low output noise is required, the linear regulator is still the best
technique available. Also, the “cascaded” linear system described here is a very useful and
somewhat neglected technique. Finally, the high dissipation and low efficiency of the dissi-
pative linear regulator serve to illustrate the advantages of the switchmode variable supply,
described in the next chapter.
In this section, we review the basic concepts of a linear variable supply for labora-
tory applications. The same general principles will apply to fixed-voltage linear regulators,
except that for the latter the losses would normally be much lower.
To its advantage, the linear regulator has inherently low noise levels, usually measured
in microvolts rather than the more familiar millivolts of switchmode systems. For applica-
tions in which the minimum electrical noise levels are essential (for example, sensitive
communications equipment and research and development activities), the advantage of
the very low noise levels of the dissipative linear regulator often outweighs the wish for
The transient response of a well-designed linear system may be of the order of 20 Ms for
full recovery, rather than the 500 Ms for the typical switchmode regulator.
The major disadvantage of the linear regulator is that it must dissipate as heat the power
difference between the used output power (volt-amperes) and the internally generated volt-
amperes. This dissipation can be very large. It is largest at high output currents and low
In the example to be considered here (a 60-V, 2-A variable supply), the unregulated
header voltage will be 70 V minimum. When the variable output voltage is set to zero at
2 A load (an output short circuit), normal series regulator dissipation would be 140 W mini-
mum. If this energy is all concentrated in series linear regulator transistors, then expensive
heat sinks and transistors will be required.
The following section describes a method of secondary preregulation which allows the
majority of the unwanted energy to be dissipated in passive resistors rather than in the series
regulator transistors. Dissipating the energy in resistors has major advantages. It should be
remembered that good-quality wirewound resistors are much more efficient at dissipat-
ing the unwanted energy, since they may run at much higher surface temperatures than
semiconductor devices can. Hence smaller air flow can efficiently carry away the excess
heat. Resistors are also much lower in cost than extra regulation transistors and heat sinks.
Documents you may be interested
Documents you may be interested