Cappels' project pages
A 5 volt to +24V
flyback output DC
to DC Converter
Return to HOME
A simple regulated +24 volts power
supply for low current applications.
The circuit was built on a small
piece of phenolic board
about 2 cm x 5.5 cm. It will be placed "piggyback" on a larger assembly.
I needed a few milliamps at +24
volts for a varactor bias circuit. Since my systems approach is still
to run all of the power
though a 5 volt regulator, this meant that I had to come up with a
5VDC-to-24VDC power supply. The resulting DC to DC converter and
regulation circuit are described here.
The intended use of the 24 volts is to power an amplifier that will
supply wave forms of up to a couple hundred Hz to bias a varactor or
pair of varactors in an RF circuit. This meant that the current
requirements would be low - probably a little over a milliamp. It also
made me worry about recovery spikes from the rectifier, and harmonics
from the switching itself, as they might affect the operation of the
circuit the power supply is meant to power. The circuit here only has
minimal filtering. Filtering to remove filter glitches will be added to
I first considered a switched capacitor power supply since this could
be made to switch "softly" at a low frequency, but the parts count
started looking pretty high considering I had to go from a regulated +5
volt power supply up to a regulated +24 volt output. This would
have taken five half wave multiplier stages.
I also looked at inductor based blocking oscillators, such as that used
in the 1.5 volt LED flashlight
The problem there was that I did
not have a suitable ferrite core. I also considered a forward converter
made with a pair of transistors connected as a multivibrator, driving a
center tapped coil with a center tapped base drive winding and a
secondary for output, but I didn't have an appropriate size core for
this either. My choice at the time was between
cores that are way too large and cores that are way too small.
Happily, I have a selection of molded chokes I picked up at a
surplus store and using a reliable old oscillator circuit, was able to
produce 24 volts DC at the required current levels, in a
size that seems reasonable. It is small enough that I can easily
enclose the entire circuit inside a copper shield.
(Above) The circuit can be seen
as an analog regulator with the voltage converter circuit in series
with the feedback path.
(Above) The basic converter was
once a popular code practice oscillator circuit.
The basic converter provides an output
voltage that is about 10 times the input voltage. The voltage boost
occurs when the 2N4401 switches off, and the magnetic flux that had
been supported by current flowing in the inductor, collapses. This
rapid change in flux is accompanied by the voltage on the collector of
the 2N4401 rapidly rising, limited for the most part by the
turnoff time of the
transistor, but ultimately being clamped to a diode drop above the
voltage across the output capacitor. This type of converter is often
referred to as a boost
converter or a flyback converter, the later term referring to the high
voltage power supply that was integrated with the horizontal deflection
circuits in cathode ray tube television sets. High voltages would be
developed when the horizontal scan circuit rapidly made the luminous
spot return (or "fly back") to the left side of the screen.
is a little under 300 kHz when under load. I had first used a 2N2222 as
the NPN because they are pretty common parts, but I could not resist
the improved efficiency of the faster switching 2N4401.
After measuring the input voltage required to maintain the output
voltage at +24 volts as the load was varied, the basic circuit was then
treated as a building block and placed inside an analog regulation
loop that could supply the range of voltages needed. From the graph
(further below) ,
the analog regulator's output, the emitter of the 2N2222, needs to be
able to cover the range from 2.2 volts to about 3.5 volts in order to
supply +24 Volts from no load to a 5 milliamp load.
After some refinement to the basic
converter, I determined that the converter's analog section should
also be able to provide about 40 milliamps to the converter section
when supplying a load current of 5 milliamps.
(Above) The Analog regulator is
made from a TL431 and a 2N222. Q2 and Q3
for an oscillator that develops +24 volts on the cathode of D3.
The completed circuit is shown
above. The analog regulator is a TL431 shunt regulator chip. It
is followed by Q1, a
2N2222 which both provides more output current, and shifts the voltage
one base-emitter drop lower. The TL431 adjusts the current through its
cathode, and thus the 500 Ohm resistor to
maintain 2.5 volts on the reference input. The voltage divider on the
output of the supply divides the +24 volts down to +2.5 volts. I added
a pot to adjust the output voltage since the build up of the tolerances
could result in more than a volt of error.
The voltage at
the cathode of the TL431 is a function of the current drawn by the
TL431, and this voltage is buffered by the 2N2222. The 10 Ohm resistor
in series with the 2N2222 limits the peak emitter current that is
available to change C1 during turn-on and in case there are large
swings in the load current. C4 and R10 reduce the closed loop gain at
high frequencies to keep the TL431 from oscillating.
(Above) Input voltage, in volts,
applied to the converter stage to
maintain +24 volts at the output
as a function of the load current in amps. Note that for this test, the
input to the regulator circuit was +9VDC. When powered from a +5VDC
supply, the maximum load current before dropping out of regulation is
Input current, in amps,
drawn by the converter stage while maintaining +24 VDC on the output,
as a function of load, in amps. The current determined by measuring the
voltage drop across the 10 Ohm resistor in series with the collector of
Since the TL431 cathode is designed to operate at +2.5 VDC, the base
emitter drop of the 2N2222 assured that the voltage supplied to the
converter could be lower than 2.0 volts, to accommodate no load. The
value of R1 was chosen so as to provide the minimum required current
through the TL431 when operating at the maximum output voltage, and
still provide sufficient base current to supply over 40 milliamps of
A casual glance will show some differences between the basic converter
stage and the one in the completed circuit.
The addition of D2 prevents the flyback pulse from avalanching the
base-emitter junction with the collector pulse from Q3. This might have
degraded the beta of Q2 over the long run. A significant amount of
power was also being dissipated in the avalanched base-emitter
junction. R4 helps Q2 turn off more quickly than
it would if R4 were not in the circuit.
A modification to the base drive circuit for Q3 helps Q3 turn off more
quickly, thus saving more power. When Q3 is driven by current from Q2
through R3, inductor L1 draws some of that base current. When Q2 turns
off, the current in L1 causes the base of Q3 to swing negative, thus
turning it off more quickly. R6, the 330 Ohm resistor across L1,
absorbs some of the energy from L1 so that the voltage on the base of
Q3 does not ring positive during the flyback pulse, thus saving several
milliamps of input current. In a higher power circuit, Q3 turning
on during the flyback pulse might well mean destruction of Q3.
The base drive enhancements are not required, as they only add to the
efficiency of the circuit. D2, R4, D1, and L1 can all be omitted, and
the original base drive circuits shown in the basic converter circuit
can be used instead, only the current drain will be a lot higher. The
input voltage and input current curves on this
web page were measured on the completed, improved circuit, inside the
The value for C3 was calculated to limit the maximum positive-going
voltage transient on the output in case the load became disconnected.
If all of
the energy stored in L2 was suddenly dumped into C3, how large would C3
have to be to keep the output voltage from rising more than 1 volt?
A safe and practical answer is found by quickly by ignoring the
resistance of the
circuit and noting that the energy stored in an inductor is 1/2 *
L*I^2, and the energy stored in a capacitor is 1/2 * C * V^2. Using 220
microhenries and taking the inductor's saturation current of 300
milliamps as the maximum current, C = (L*I^2)/V^2 = 20 uf.
I should mention that the 1N4141 is not normally my first choice as a
rectifier, but given the light usage, about 50 milliamps peak current,
and low demands on efficiency, it works very well in this application.
Contents ©2005, 2006 Richard Cappels All Rights Reserved. http://www.projects.cappels.org/
First posted in October, 2005. Revised
November 2005, September 2006.
You can send email to me at
"(at)" with "@" before mailing.
presented on this page is for personal, nonprofit educational and
use only. This material (including object files) is copyrighted by
Cappels and may not be republished or used directly for commercial
For commercial license, click
and intellectual property notice
(Summary: No warranties, use these pages at your
own risk. You may use the information provided here for personal and
educational purposes but you may not republish or use this information
for any commercial purpose without explicit permission.) I neither
express nor imply any warranty for the quality, fitness for any
particular purpose or user, or freedom from patents or other
restrictions on the rights of use of any
software, firmware, hardware, design, service,information, or advice
mentioned,or made reference to in these pages. By utilizing or relying
on software, firmware, hardware, design, service,information, or advice
provided, mentioned, or made reference to in these pages, the user
takes responsibility to assume all risk and associated with said
activity and hold Richard Cappels harmless in the event of any loss or
expense associated with said activity. The contents of this web site,
unless otherwise noted, is copyrighted by Richard
Cappels. Use of information presented on this site for personal,
educational and noncommercial use is encouraged, but unless explicitly
with respect to particular material, the material itself may not be
or used directly for commercial purposes. For the purposes of this
copying binary data resulting from program files, including assembly
code and object (hex) files into semiconductor memories for personal,
educational or other noncommercial use is not considered republishing.
desiring to use any material published in this pages for commercial
should contact the respective copyright holder(s).