The article presents a simple PWM controller design, with which you can easily convert a computer power supply assembled on a controller other than the popular TL494, in particular, DR-B2002, DR-B2003, SG6105 and others, into a laboratory one with adjustable output voltage and load current limitation. Also here I will share the experience of reworking computer PSUs and describe proven ways to increase their maximum output voltage.
In amateur radio literature, there are many schemes for converting obsolete computer power supplies (PSUs) into chargers and laboratory power supplies (IPs). But all of them relate to those PSUs in which the control unit is built on the basis of a TL494 type PWM controller chip, or its analogues DBL494, KIA494, KA7500, KR114EU4. We have converted more than a dozen of these PSUs. Chargers made according to the scheme described by M. Shumilov in the article "Computer power supply - charger" (Radio - 2009, No. 1) with the addition of a pointer measuring device for measuring the output voltage and charging current showed themselves well. Based on the same scheme, the first laboratory power supplies were made, until the “Universal control board for laboratory power supplies” came into view (Radio Yearbook - 2011, No. 5, p. 53). According to this scheme, it was possible to produce much more functional power supplies. Especially for this regulator circuit, a digital ampervoltmeter was developed, described in the article “A simple built-in ampervoltmeter on the PIC16F676”.
But all good things come to an end, and recently computer power supplies with other PWM controllers, in particular, DR-B2002, DR-B2003, SG6105, have become increasingly common. The question arose: how can these PSUs be used for the manufacture of laboratory IP? The search for circuits and communication with radio amateurs did not allow progress in this direction, although we managed to find a brief description and a diagram for including such PWM controllers in the article “SG6105 and DR-B2002 PWM controllers in computer IPs”. From the description, it became clear that these controllers are much more complicated than TL494 and it is hardly possible to try to control them from the outside to regulate the output voltage. Therefore, it was decided to abandon this idea. However, when studying the circuits of the "new" PSUs, it was noted that the construction of the control circuit for a push-pull half-bridge converter was performed similarly to the "old" PSUs - on two transistors and an isolation transformer.
An attempt was made to install the TL494 with its standard wiring instead of the DR-B2002 chip by connecting the collectors of the TL494 output transistors to the transistor bases of the power supply converter control circuit. As a strapping TL494 to ensure the regulation of the output voltage, the repeatedly tested above-mentioned M. Shumilov circuit was chosen. This inclusion of the PWM controller allows you to disable all the blocking and protection circuits available in the PSU, in addition, this circuit is very simple.
An attempt to replace the PWM controller was successful - the power supply unit worked, the output voltage adjustment and current limiting also worked, as in the converted "old" type power supply unit.
The PWM controller block is assembled on a printed circuit board made of one-sided foil fiberglass with a size of 40x45 mm. The drawing of the printed circuit board and the layout of the elements are shown in the figure. The drawing is shown from the installation side of the components.
The board is designed for the installation of output components. There are no special requirements for them. Transistor VT1 can be replaced by any other similar bipolar direct conduction transistor. The board provides for the installation of tuning resistors R5 of different sizes.
The board is fixed in a convenient place with one screw closer to the installation site of the PWM controller. The author found it convenient to attach the board to one of the PSU heatsinks. The PWM1, PWM2 outputs are soldered directly into the corresponding holes of the previously installed PWM controller - the outputs of which go to the bases of the converter control transistors (pins 7 and 8 of the DR-B2002 chip). The Vcc pin is connected to the point where there is an output voltage of the standby power circuit, the value of which can be in the range of 13 ... 24V.
The output voltage of the IP is adjusted by the potentiometer R5, the minimum output voltage depends on the value of the resistor R7. Resistor R8 can limit the maximum output voltage. The value of the maximum output current is regulated by the selection of the value of the resistor R3 - the lower its resistance, the greater the maximum output current of the PSU.
The work on altering the power supply unit is associated with work in high voltage circuits, therefore it is strongly recommended to connect the power supply unit to the network through an isolating transformer with a power of at least 100W. In addition, in order to avoid failure of key transistors in the process of setting up the IP, it should be connected to the network through a “safety” incandescent lamp for 220V with a power of 100W. It can be soldered to the PSU instead of the mains fuse.
Before proceeding with the alteration of a computer PSU, it is advisable to make sure that it is in good condition. Before switching on, 12V car bulbs with a power of up to 25 W should be connected to the output circuits + 5V and + 12V. Then connect the PSU to the network and connect the PS-ON output (usually green) to the common wire. If the PSU is in good condition, the “safety” lamp will briefly flash, the PSU will work and the lamps will light up in the load + 5V, + 12V. If, after turning on, the “safety” lamp lights up at full heat, a breakdown of power transistors, rectifier bridge diodes, etc. is possible.
Next, you should find on the PSU board the point at which there is an output voltage of the standby power circuit. Its value can be in the range of 13 ... 24V. From this point, in the future, we will take power for the PWM controller unit and the cooling fan.
Then you should unsolder the standard PWM controller and connect the PWM controller unit to the PSU board according to the diagram (Fig. 1). The P_IN input is connected to the 12-volt output of the PSU. Now you need to check the operation of the regulator. To do this, connect a load in the form of a car light to the P_OUT output, pull the resistor R5 to the left (to the position of minimum resistance) and connect the PSU to the network (again through a “safety” lamp). If the load lamp lights up, you should make sure that the adjustment circuit is working. To do this, carefully turn the slider of the resistor R5 to the right, while it is desirable to control the output voltage with a voltmeter so as not to burn the load lamp. If the output voltage is regulated, then the PWM controller unit is working and you can continue to upgrade the PSU.
We solder all the PSU load wires, leaving one wire in the +12 V circuits and a common one for connecting the PWM controller unit. We solder: diodes (diode assemblies) in circuits +3.3 V, +5 V; rectifier diodes -5 V, -12 V; all filter capacitors. The electrolytic capacitors of the +12 V circuit filter should be replaced with capacitors of the same capacity, but with an allowable voltage of 25 V or more, depending on the expected maximum output voltage of the manufactured laboratory power supply. Next, you should install a load resistor, shown in the diagram in Fig. 1 as R2, necessary to ensure stable operation of the IP without external load. The load power should be about 1W. The resistance of the resistor R2 can be calculated based on the maximum output voltage of the IP. In the simplest case, a 2-watt resistor with a resistance of 200-300 ohms will do.
Next, you can solder the binding elements of the old PWM controller and other radio components from the unused PSU output circuits. In order not to accidentally unsolder something “useful”, it is recommended to solder the parts not completely, but one by one, and only after making sure that the IP is working, remove the part completely. As for the L1 filter inductor, the author usually does nothing with it and uses the standard winding of the +12 V circuit. This is due to the fact that, for safety reasons, the maximum output current of a laboratory power supply is usually limited to a level not exceeding the passport rating for the +12 V power supply circuit .
After cleaning the installation, it is recommended to increase the capacitance of the filter capacitor C1 of the standby power supply, replacing it with a 50 V / 100 uF capacitor. In addition, if the VD1 diode installed in the circuit is low-power (in a glass case), it is recommended to replace it with a more powerful one, soldered from a -5 V or -12 V circuit rectifier. You should also select the resistance of the resistor R1 for comfortable operation of the M1 cooling fan.
The experience of reworking computer PSUs has shown that using various PWM controller control schemes, the maximum output voltage of the IP will be in the range of 21 ... 22 V. This is more than enough for the manufacture of chargers for car batteries, but still not enough for a laboratory power source. To obtain an increased output voltage, many radio amateurs suggest using a bridge circuit for rectifying the output voltage, but this is due to the installation of additional diodes, the cost of which is quite high. I consider this method irrational and use another way to increase the output voltage of the IP - upgrading the power transformer.
There are two main ways to upgrade the IP power transformer. The first method is convenient because its implementation does not require disassembly of the transformer. It is based on the fact that usually the secondary winding is wound in several wires and it is possible to “stratify” it. Schematically, the secondary windings of a power transformer are shown in fig. A). This is the most common scheme. Typically, a 5-volt winding has 3 turns wound in 3-4 wires (windings "3.4" - "general" and "general" - "5.6"), and a 12-volt winding has an additional 4 turns in one wire (windings "1" - "3.4" and "5.6" - "2").
To do this, the transformer is unsoldered, the taps of the 5-volt winding are carefully unsoldered and the “pigtail” of the common wire is untwisted. The task is to disconnect the 5-volt windings connected in parallel and connect all or part of them in series, as shown in the diagram of fig. b).
Selecting the windings is not difficult, but it is rather difficult to phase them correctly. The author uses for this purpose a low-frequency sinusoidal signal generator and an oscilloscope or an alternating current millivoltmeter. By connecting the output of the generator, tuned to a frequency of 30 ... 35 kHz, to the primary winding of the transformer, the voltage on the secondary windings is monitored using an oscilloscope or millivoltmeter. By combining the connection of 5-volt windings, an increase in the output voltage compared to the original one by the required amount is achieved. In this way, it is possible to achieve an increase in the output voltage of the PSU up to 30 ... 40 V.
The second way to upgrade a power transformer is to rewind it. This is the only way to get the output voltage of the PSU above 40V. The most difficult task here is the separation of the ferrite core. The author adopted the method of boiling the transformer in water for 30-40 minutes. But before boiling the transformer, one should think carefully about the way to disconnect the core, given the fact that after boiling it will be very hot, in addition, hot ferrite becomes very brittle. To do this, it is proposed to cut out two wedge-shaped strips from tin, which can then be inserted into the gap between the core and the frame, and with their help to separate the halves of the core. In case of breaking or chipping off parts of the ferrite core, you should not be particularly upset, since it can be successfully glued together with cyacrylane (the so-called "superglue").
After releasing the transformer coil, it is necessary to wind the secondary winding. Pulse transformers have one unpleasant feature - the primary winding is wound in two layers. First, the first part of the primary winding is wound on the frame, then the screen, then all the secondary windings, again the screen and the second part of the primary winding. Therefore, you need to carefully wind the second part of the primary winding, while remembering its connection and winding direction. Then remove the screen, made in the form of a layer of copper foil with a soldered wire leading to the transformer output, which must first be unsoldered. And finally, wind the secondary windings to the next screen. Now it is imperative to dry the coil well with a jet of hot air to evaporate the water that has penetrated into the winding during digestion.
The number of turns of the secondary winding will depend on the required maximum output voltage of the power supply at the rate of approximately 0.33 turns / V (that is, 1 turn - 3 V). For example, the author wound 2x18 turns of PEV-0.8 wire and got the maximum output voltage of the IP about 53 V. The cross section of the wire will depend on the requirement for the maximum output current of the IP, as well as on the dimensions of the transformer frame.
The secondary winding is wound in 2 wires. The end of one wire is immediately soldered to the first output of the frame, and the second is left with a margin of 5 cm to form a "pigtail" of the zero output. Having finished winding, the end of the second wire is soldered to the second terminal of the frame and a “pigtail” is formed in such a way that the number of turns of both half-windings must be the same.
Now you should restore the screen, wind the previously wound second part of the primary winding of the transformer, observing the original connection and winding direction, and assemble the transformer magnetic circuit. If the wiring of the secondary winding is soldered correctly (to the terminals of the 12-volt winding), then you can solder the transformer to the PSU board and check its performance.
Tell in:The article presents a simple PWM controller design, with which you can easily convert a computer power supply assembled on a controller other than the popular tl494, in particular, dr-b2002, dr-b2003, sg6105 and others, into a laboratory one with adjustable output voltage and load current limitation. Also here I will share the experience of reworking computer PSUs and describe proven ways to increase their maximum output voltage.
In amateur radio literature, there are many schemes for converting obsolete computer power supplies (PSUs) into chargers and laboratory power supplies (IPs). But they all relate to those PSUs in which the control unit is built on the basis of a tl494 type PWM controller chip, or its analogues dbl494, kia494, KA7500, KR114EU4. We have converted more than a dozen of these PSUs. Chargers made according to the scheme described by M. Shumilov in the article “A simple built-in ampervoltmeter on pic16f676” showed themselves well.
But all good things come to an end, and recently computer power supplies with other PWM controllers, in particular, dr-b2002, dr-b2003, sg6105, have become increasingly common. The question arose: how can these PSUs be used for the manufacture of laboratory IP? The search for circuits and communication with radio amateurs did not allow us to advance in this direction, although we managed to find a brief description and a diagram for including such PWM controllers in the article “sg6105 and dr-b2002 PWM controllers in computer IPs.” From the description, it became clear that these controllers much more complicated than tl494 and trying to control them from the outside to regulate the output voltage is hardly possible. Therefore, it was decided to abandon this idea. However, when studying the circuits of the "new" PSUs, it was noted that the construction of the control circuit for a push-pull half-bridge converter was carried out similarly to the "old" PSUs - on two transistors and an isolation transformer.
An attempt was made instead of the dr-b2002 microcircuit to install the tl494 with its standard harness, by connecting the collectors of the tl494 output transistors to the transistor bases of the power supply converter control circuit. As a strapping tl494 to ensure regulation of the output voltage, the above-mentioned M. Shumilov circuit, repeatedly tested above, was chosen. This inclusion of the PWM controller allows you to disable all the blocking and protection circuits available in the PSU, in addition, this circuit is very simple.
An attempt to replace the PWM controller was successful - the PSU started working, the output voltage adjustment and current limiting also worked, as in the converted "old" PSU.
The PWM controller block is assembled on a printed circuit board made of one-sided foil fiberglass with a size of 40x45 mm. The drawing of the printed circuit board and the layout of the elements are shown in the figure. The drawing is shown from the installation side of the components.
The board is designed for the installation of output components. There are no special requirements for them. Transistor vt1 can be replaced by any other bipolar direct conduction transistor of similar parameters. The board provides for the installation of trimming resistors r5 of different sizes.
The board is fixed in a convenient place with one screw closer to the installation site of the PWM controller. The author found it convenient to attach the board to one of the PSU heatsinks. The outputs pwm1, pwm2 are soldered directly into the corresponding holes of the previously installed PWM controller - the conclusions of which go to the bases of the converter control transistors (pins 7 and 8 of the dr-b2002 chip). The vcc pin is connected to a point where there is an output voltage of the standby power circuit, the value of which can be in the range of 13 ... 24V.
The output voltage of the IP is adjusted by the potentiometer r5, the minimum output voltage depends on the value of the resistor r7. Resistor r8 can limit the maximum output voltage. The value of the maximum output current is regulated by the selection of the value of the resistor r3 - the lower its resistance, the greater the maximum output current of the PSU.
The work on altering the power supply unit is associated with work in high voltage circuits, therefore it is strongly recommended to connect the power supply unit to the network through an isolating transformer with a power of at least 100W. In addition, in order to avoid failure of key transistors in the process of setting up the IP, it should be connected to the network through a “safety” incandescent lamp for 220V with a power of 100W. It can be soldered to the PSU instead of the mains fuse.
Before proceeding with the alteration of a computer PSU, it is advisable to make sure that it is in good condition. Before switching on, 12V car bulbs with a power of up to 25 W should be connected to the output circuits + 5V and + 12V. Then connect the PSU to the network and connect the ps-on output (usually green) to the common wire. If the PSU is in good condition, the “safety” lamp will briefly flash, the PSU will work and the lamps will light up in the load + 5V, + 12V. If, after turning on, the “safety” lamp lights up at full heat, a breakdown of power transistors, rectifier bridge diodes, etc. is possible.
Next, you should find on the PSU board the point at which there is an output voltage of the standby power circuit. Its value can be in the range of 13 ... 24V. From this point, in the future, we will take power for the PWM controller unit and the cooling fan.
Then you should unsolder the standard PWM controller and connect the PWM controller unit to the PSU board according to the diagram (Fig. 1). The p_in input is connected to the 12-volt output of the PSU. Now you need to check the operation of the regulator. To do this, connect a load in the form of a car light to the p_out output, bring the r5 resistor engine to the left to failure (to the position of minimum resistance) and connect the PSU to the network (again through a “safety” lamp). If the load lamp lights up, you should make sure that the adjustment circuit is working. To do this, you need to carefully turn the slider of the resistor r5 to the right, while it is desirable to control the output voltage with a voltmeter so as not to burn the load lamp. If the output voltage is regulated, then the PWM controller unit is working and you can continue to upgrade the PSU.
We solder all the PSU load wires, leaving one wire in the +12 V circuits and a common one for connecting the PWM controller unit. We solder: diodes (diode assemblies) in circuits +3.3 V, +5 V; rectifier diodes -5 V, -12 V; all filter capacitors. The electrolytic capacitors of the +12 V circuit filter should be replaced with capacitors of the same capacity, but with an allowable voltage of 25 V or more, depending on the expected maximum output voltage of the manufactured laboratory power supply. Next, you should install a load resistor, shown in the diagram in Fig. 1 as r2, necessary to ensure stable operation of the IP without external load. The load power should be about 1W. The resistance of the resistor r2 can be calculated based on the maximum output voltage of the IP. In the simplest case, a 2-watt resistor with a resistance of 200-300 ohms will do.
Next, you can solder the binding elements of the old PWM controller and other radio components from the unused PSU output circuits. In order not to accidentally unsolder something “useful”, it is recommended to solder the parts not completely, but one by one, and only after making sure that the IP is working, remove the part completely. As for the filter inductor l1, the author usually does nothing with it and uses the standard winding of the +12 V circuit. This is due to the fact that, for safety reasons, the maximum output current of a laboratory power supply is usually limited to a level not exceeding the passport rating for the +12 V power supply circuit .
After cleaning the installation, it is recommended to increase the capacitance of the filter capacitor C1 of the standby power supply, replacing it with a 50 V / 100 uF capacitor. In addition, if the vd1 diode installed in the circuit is low-power (in a glass case), it is recommended to replace it with a more powerful one, soldered from a -5 V or -12 V circuit rectifier. You should also select the resistance of the resistor r1 for comfortable operation of the M1 cooling fan.
The experience of reworking computer PSUs has shown that using various PWM controller control schemes, the maximum output voltage of the IP will be in the range of 21 ... 22 V. This is more than enough for the manufacture of chargers for car batteries, but still not enough for a laboratory power source. To obtain an increased output voltage, many radio amateurs suggest using a bridge circuit for rectifying the output voltage, but this is due to the installation of additional diodes, the cost of which is quite high. I consider this method irrational and use another way to increase the output voltage of the IP - upgrading the power transformer.
There are two main ways to upgrade the IP power transformer. The first method is convenient because its implementation does not require disassembly of the transformer. It is based on the fact that usually the secondary winding is wound in several wires and it is possible to “stratify” it. Schematically, the secondary windings of a power transformer are shown in fig. A). This is the most common scheme. Typically, a 5-volt winding has 3 turns wound in 3-4 wires (windings "3.4" - "general" and "general" - "5.6"), and a 12-volt winding has an additional 4 turns in one wire (windings "1" - "3.4" and "5.6" - "2").
To do this, the transformer is unsoldered, the taps of the 5-volt winding are carefully unsoldered and the “pigtail” of the common wire is untwisted. The task is to disconnect the 5-volt windings connected in parallel and connect all or part of them in series, as shown in the diagram of fig. b).
Selecting the windings is not difficult, but it is rather difficult to phase them correctly. The author uses for this purpose a low-frequency sinusoidal signal generator and an oscilloscope or an alternating current millivoltmeter. By connecting the output of the generator, tuned to a frequency of 30 ... 35 kHz, to the primary winding of the transformer, the voltage on the secondary windings is monitored using an oscilloscope or millivoltmeter. By combining the connection of 5-volt windings, an increase in the output voltage compared to the original one by the required amount is achieved. In this way, it is possible to achieve an increase in the output voltage of the PSU up to 30 ... 40 V.
The second way to upgrade a power transformer is to rewind it. This is the only way to get the output voltage of the PSU above 40V. The most difficult task here is the separation of the ferrite core. The author adopted the method of boiling the transformer in water for 30-40 minutes. But before boiling the transformer, one should think carefully about the way to disconnect the core, given the fact that after boiling it will be very hot, in addition, hot ferrite becomes very brittle. To do this, it is proposed to cut out two wedge-shaped strips from tin, which can then be inserted into the gap between the core and the frame, and with their help to separate the halves of the core. In case of breaking or chipping off parts of the ferrite core, you should not be particularly upset, since it can be successfully glued together with cyacrylane (the so-called "superglue").
After releasing the transformer coil, it is necessary to wind the secondary winding. Pulse transformers have one unpleasant feature - the primary winding is wound in two layers. First, the first part of the primary winding is wound on the frame, then the screen, then all the secondary windings, again the screen and the second part of the primary winding. Therefore, you need to carefully wind the second part of the primary winding, while remembering its connection and winding direction. Then remove the screen, made in the form of a layer of copper foil with a soldered wire leading to the transformer output, which must first be unsoldered. And finally, wind the secondary windings to the next screen. Now it is imperative to dry the coil well with a jet of hot air to evaporate the water that has penetrated into the winding during digestion.
The number of turns of the secondary winding will depend on the required maximum output voltage of the power supply at the rate of approximately 0.33 turns / V (that is, 1 turn - 3 V). For example, the author wound 2x18 turns of PEV-0.8 wire and got the maximum output voltage of the IP about 53 V. The cross section of the wire will depend on the requirement for the maximum output current of the IP, as well as on the dimensions of the transformer frame.
The secondary winding is wound in 2 wires. The end of one wire is immediately soldered to the first output of the frame, and the second is left with a margin of 5 cm to form a "pigtail" of the zero output. Having finished winding, the end of the second wire is soldered to the second terminal of the frame and a “pigtail” is formed in such a way that the number of turns of both half-windings must be the same.
Now you should restore the screen, wind the previously wound second part of the primary winding of the transformer, observing the original connection and winding direction, and assemble the transformer magnetic circuit. If the wiring of the secondary winding is soldered correctly (to the terminals of the 12-volt winding), then you can solder the transformer to the PSU board and check its performance.
ARCHIVE: Download
Section: [Power supplies (switched)]
Save article to:
do-it-yourself charger from a computer power supply
Different situations require different voltage and power IP. Therefore, many buy or make one so that it is enough for all cases.
And the easiest way is to take the computer as a basis. This laboratory power supply with characteristics 0-22 V 20 A redone with minor modifications from computer ATX to PWM 2003. For rework I used JNC mod. LC-B250ATX. The idea is not new and there are many similar solutions on the Internet, some have been studied, but the final one turned out to be different. I am very pleased with the result. Now I am waiting for a parcel from China with combined voltage and current indicators, and, accordingly, I will replace it. Then it will be possible to call my development LBP - charger for car batteries.
Diagram of an adjustable power supply:
First of all, I soldered all the wires of the output voltages +12, -12, +5, -5 and 3.3 V. I soldered everything except +12 V diodes, capacitors, load resistors.
I replaced the input high-voltage electrolytes 220 x 200 with 470 x 200. If there is, then it is better to put a larger capacity. Sometimes the manufacturer saves on the input power filter - accordingly, I recommend soldering it if it is missing.
+12 V output choke rewound. New - 50 turns with a wire with a diameter of 1 mm, removing the old windings. The capacitor was replaced with 4700 microfarads x 35 V.
Since the unit has a standby power supply with voltages of 5 and 17 volts, I used them to power the 2003rd and the voltage test unit.
To pin 4, I applied a direct voltage of +5 volts from the "duty room" (i.e. connected it to pin 1). With the help of a resistor 1.5 and 3 kOhm voltage divider from 5 volts of standby power, I made 3.2 and applied it to input 3 and to the right output of the resistor R56, which then goes to pin 11 of the microcircuit.
By installing the 7812 microcircuit on the output of 17 volts from the duty room (capacitor C15), I received 12 volts and connected it to a 1 Kom resistor (without a number on the diagram), which is connected to pin 6 of the microcircuit with the left end. Also, through a 33 ohm resistor, it powered the cooling fan, which was simply turned over so that it blew inward. The resistor is needed in order to reduce the speed and noise of the fan.
The entire chain of resistors and diodes of negative voltages (R63, 64, 35, 411, 42, 43, C20, D11, 24, 27) dropped out of the board, pin 5 of the microcircuit was shorted to ground.
Added adjustment voltage and output voltage indicator from a Chinese online store. It is only necessary to power the latter from the duty room +5 V, and not from the measured voltage (it starts working from +3 V). Power Supply Tests
Tests were carried out simultaneous connection of several car lamps (55+60+60) W.
This is about 15 Amperes at 14 V. I worked for 15 minutes without problems. Some sources recommend isolating the common 12 V output wire from the case, but then a whistle appears. Using the car radio as a power source, I did not notice any interference either on the radio or in other modes, and 4 * 40 W draws perfectly. Sincerely, Petrovsky Andrey.
Introduction
A big plus of a computer power supply is that it works stably when the mains voltage changes from 180 to 250 V, and some instances work even with a larger voltage spread. It is possible to obtain a useful load current of 15-17 A from a 200 W unit, and up to 22 A in a pulsed (short-term high load mode). and below, most often made on microcircuits 2003, AT2005Z, SG6105, KA3511, LPG-899, DR-B2002, IW1688. Such devices contain fewer discrete elements on the board, have a lower cost than those built on the basis of the popular PWM - TL494 microcircuits. In this article, we will look at several approaches to repairing the aforementioned power supplies and give some practical advice.
Blocks and diagrams
A computer power supply can be used not only for its intended purpose, but also as sources for a wide range of electronic designs for the home, requiring a constant voltage of 5 and 12 V for their operation. By a slight alteration described below, this is not at all difficult. And you can buy a PC PSU separately both in a store and used on any radio market (if you don’t have enough of your own “bins”) for a symbolic price.
In this way, the computer power supply compares favorably with all other industrial options in the future for use in the home laboratory of a radio master. For example, we will take JNC blocks of the LC-B250ATX and LC-B350ATX models, as well as InWin IP-P300AQ2, IP-P350AQ2, IP-P400AQ2, IP-P350GJ20, which use the 2003 IFF LFS 0237E chip in their design. Some others have BAZ7822041H or 2003 BAY05370332H. All these microcircuits are structurally different from each other in the purpose of the conclusions and the "stuffing", but the principle of operation is the same for them. So the 2003 IFF LFS 0237E chip (hereinafter referred to as 2003) is a PWM (pulse-width signal modulator) in a DIP-16 package. Until recently, most budget computer PSUs manufactured by Chinese companies were based on the TL494 PWM controller chip from Texas Instruments (http://www.ti.com) or its analogues from other manufacturers, such as Motorola, Fairchild, Samsung and others. The same microcircuit has a domestic analogue KR1114EU4 and KR1114EU3 (the pinout in the domestic version is different). Let's learn how to diagnose and test problems first.
How to change the input voltage
The signal, the level of which is proportional to the load power of the converter, is taken from the midpoint of the primary winding of the isolation transformer T3, then through the diode D11 and the resistor R35 it enters the correcting circuit R42R43R65C33, after which it is fed to the PR output of the microcircuit. Therefore, in this scheme, it is difficult to set the priority of protection for any one voltage. Here it would be necessary to change the scheme greatly, which is unprofitable in terms of time.
In other computer power supply circuits, for example, in LPK-2-4 (300 W), the voltage from the cathode of the dual Schottky diode type S30D40C, the +5 V output voltage rectifier, is fed to the UVac input of the U2 microcircuit and is used to control the input AC supply voltage BP. An adjustable output voltage is useful for the home lab. For example, to power electronic devices for a car from a computer power supply unit, where the voltage in the on-board network (with the engine running) is 12.5-14 V. The higher the voltage level, the greater the useful power of the electronic device. This is especially important for radio stations. For example, consider the adaptation of a popular radio station (transceiver) to our PSU LC-B250ATX - increasing the voltage on the 12 V bus to 13.5-13.8 V.
We solder a tuning resistor, for example, SP5-28V (preferably with the index "B" in the designation - a sign of linearity of the characteristic) with a resistance of 18-22 kOhm between terminal 6 of the U2 microcircuit and the +12 V bus. Install a car light bulb 5- on the +12 V output 12 W as a dummy load (you can also connect a constant resistor of 5-10 ohms with a dissipation power of 5 W or more). After the considered minor refinement of the PSU, the fan can not be connected and the board itself can not be inserted into the case. We start the PSU, connect a voltmeter to the +12 V bus and control the voltage. By rotating the variable resistor slider, we set the output voltage to 13.8 V.
Turn off the power and measure the resulting resistance of the tuning resistor with an ohmmeter. Now, between the +12 V bus and pin 6 of the U2 microcircuit, we solder a constant resistor of the corresponding resistance. In the same way, you can adjust the voltage at the +5 V output. The limiting resistor itself is connected to pin 4 of the 2003 IFF LFS 0237E microcircuit.
The principle of operation of the scheme 2003
The supply voltage Vcc (pin 1) to the U2 chip comes from the standby voltage source + 5V_SB. The negative input of the error amplifier IN of the microcircuit (pin 4) receives the sum of the output voltages of the IP +3.3 V, +5 V and +12 V. The adder is made, respectively, on resistors R57, R60, R62. The controlled zener diode of the U2 microcircuit is used in the optocoupler feedback circuit in the + 5V_SB standby voltage source, the second zener diode is used in the + 3.3V output voltage stabilization circuit. The control circuit of the output half-bridge converter BP is made according to a push-pull circuit on transistors Q1, Q2 (designation on the printed circuit board) of type E13009 and transformer T3 of type EL33-ASH according to the standard circuit used in computer units.
Interchangeable transistors - MJE13005, MJE13007, Motorola MJE13009 are produced by many foreign manufacturers, therefore, instead of the abbreviation MJE, the symbols ST, PHE, KSE, HA, MJF and others may be present in the transistor marking. To power the circuit, a separate winding of the standby transformer T2 type EE-19N is used. The more power the T3 transformer has (the thicker the wire is used in the windings), the greater the output current of the power supply itself. In some printed circuit boards that I had to repair, the swing transistors were named 2SC945 and H945P, 2SC3447, 2SC3451, 2SC3457, 2SC3460 (61), 2SC3866, 2SC4706, 2SC4744, BUT11A, BUT12A, BUT18A, BUV46, MJ E13005, and the designation on The board was listed as Q5 and Q6. And at the same time, there were only 3 transistors on the board! The 2003 IFF LFS 0237E chip itself was designated as U2, and at the same time, there is not a single U1 or U3 designation on the board. However, let's leave this strangeness in the designation of elements on printed circuit boards on the conscience of the Chinese manufacturer. The designations themselves are not important. The main difference between the LC-B250ATX type power supplies under consideration is the presence on the board of one microcircuit of the 2003 IFF LFS 0237E type and the appearance of the board.
The microcircuit uses a controlled zener diode (pins 10, 11), similar to TL431. It is used to stabilize the 3.3 V power supply circuit. I note that in my practice of repairing power supplies, the above circuit is the weakest point in a computer PSU. However, before changing the 2003 chip, I recommend that you first check the circuit itself.
Diagnostics of ATX power supplies on a chip 2003
If the power supply does not start, then you must first remove the case cover and check the oxide capacitors and other elements on the printed circuit board by external inspection. Oxide (electrolytic) capacitors obviously need to be replaced if their cases are swollen and if they have a resistance of less than 100 kOhm. This is determined by a "continuity" ohmmeter, for example, model M830 in the appropriate measurement mode. One of the most common PSU malfunctions based on the 2003 chip is the lack of a stable start. The launch is performed by the Power button on the front panel of the system unit, while the button contacts are closed, and pin 9 of the U2 chip (2003 and similar) is connected to the “case” with a common wire.
In the "braid" it is usually green and black wires. In order to quickly restore the device to working capacity, it is enough to disconnect pin 9 of the U2 chip from the printed circuit board. Now the PSU should turn on stably by pressing the key on the rear panel of the system unit. This method is good in that it allows you to continue to use an obsolete computer power supply without repairs, which are not always financially profitable, or when the unit is used for other purposes, for example, to power electronic structures in a home radio amateur laboratory.
If you hold down the “reset” button before turning on the power and release it after a few seconds, the system will simulate an increase in the delay of the Power Good signal. So you can check the reasons for the failure of data loss in CMOS (after all, the battery is not always the “guilty”). If data, such as time, is intermittently lost, the shutdown delay should be checked. To do this, "reset" is pressed before turning off the power and held for a few more seconds, simulating the acceleration of the removal of the Power Good signal. If data is saved during such a shutdown, the matter is a large delay during shutdown.
Power increase
The printed circuit board has two high-voltage electrolytic capacitors with a capacity of 220 microfarads. To improve filtering, attenuate impulse noise and, as a result, to ensure the stability of a computer power supply unit to maximum loads, these capacitors are replaced with analogues of a larger capacity, for example, 680 microfarads for an operating voltage of 350 V. Breakdown, loss of capacitance or breakage of the oxide capacitor in the power supply circuit reduces or nullifies the filtering of the supply voltage. The voltage on the plates of the oxide capacitor in power supply devices is about 200 V, and the capacitance is in the range of 200-400 microfarads. Chinese manufacturers (VITO, Feron and others) install, as a rule, the cheapest film capacitors, without much concern for either the temperature regime or the reliability of the device. The oxide capacitor in this case is used in the power supply device as a high-voltage power filter, therefore it must be high-temperature. Despite the operating voltage indicated on such a capacitor of 250-400 V (with a margin, as it should be), it still "surrenders" due to its poor quality.
For replacement, I recommend oxide capacitors from KX, CapXon, namely HCY CD11GH and ASH-ELB043 - these are high-voltage oxide capacitors specially designed for use in electronic power devices. Even if the external inspection did not allow us to find faulty capacitors, the next step is to solder the conders on the +12 V bus anyway and instead install analogs of a larger capacity: 4700 uF for an operating voltage of 25 V. The section of the PCB PCB itself with oxide capacitors for power supply, to be replaced is shown in Figure 4. We carefully remove the fan and install it the other way around - so that it blows in and not out. This upgrade improves the cooling of the radioelements and, as a result, increases the reliability of the device during long-term operation. A drop of machine or domestic oil in the mechanical parts of the fan (between the impeller and the axis of the electric motor) will not hurt. In my experience, it can be said that the noise of the supercharger during operation is significantly reduced.
Replacing diode assemblies with more powerful ones
On the printed circuit board of the power supply, diode assemblies are mounted on radiators. The UF1002G assembly is installed in the center (for 12 V power supply), on the right of this radiator there is a D92-02 diode assembly that provides -5 V power supply. If such a voltage is not needed in the home laboratory, this type assembly can be permanently soldered. In general, the D92-02 is designed for a current of up to 20 A and a voltage of 200 V (in a pulsed short-term mode many times greater), so it is quite suitable for installation instead of the UF1002G (current up to 10 A).
The Fuji D92-02 diode assembly can be replaced, for example, with S16C40C, S15D40C or S30D40C. All of them, in this case, are suitable for replacement. Diodes with a Schottky barrier have less voltage drop and, accordingly, heating.
The peculiarity of the replacement is that the “regular” output diode assembly (12 V bus) UF1002G has a completely plastic case made of composite, therefore it is attached to a common radiator or current-conducting plate using thermal paste. And the Fuji D92-02 diode assembly (and similar ones) has a metal plate in the case, which requires special care when installing it on a radiator, that is, through a mandatory insulating gasket and a dielectric washer under the screw. The reason for the failure of the UF1002G diode assemblies is the voltage spikes on the diodes with an amplitude that increases when the PSU is under load. At the slightest excess of the permissible reverse voltage, Schottky diodes receive an irreversible breakdown, therefore, the recommended replacement for more powerful diode assemblies in the case of a promising use of a power supply unit with a powerful load is fully justified. Finally, there is one tip that will allow you to check the performance of the protective mechanism. We will short-circuit with a thin wire, for example, MGTF-0.8, the +12 V bus to the case (common wire). Thus, the tension should be completely gone. To restore it, turn off the PSU for a couple of minutes to discharge high-voltage capacitors, remove the shunt (jumper), remove the load dummy and turn on the PSU again; it will work normally. Converted in this way, computer power supplies operate for years in 24-hour mode with full load.
Power output
Suppose you need to use the power supply for domestic purposes and you want to remove two terminals from the power supply. I did this using two (equal length) pieces of unnecessary computer power supply wire and connected all three pre-soldered wires in each conductor to the terminal block. To reduce power loss in the conductors going from the PSU to the load, another electric cable with a copper (less loss) multi-core cable is also suitable - for example, PVSN 2x2.5, where 2.5 is the cross section of one conductor. You can also not output wires to the terminal block, and connect the 12 V output in the PC power supply case to an unused connector of the PC monitor network cable.
Pin Assignment of the 2003 Chip
PSon 2 - PS_ON signal input that controls PSU operation: PSon=0, PSU is on, all output voltages are present; PSon=1, PSU is off, only +5V_SB standby voltage is present
V33-3 - Voltage input +3.3 V
V5-4 - Voltage input +5 V
V12-6 - Voltage input +12 V
OP1/OP2-8/7 - Control outputs for push-pull half-bridge converter BP
PG-9 - Testing. PG (Power Good) signal open collector output: PG=0, one or more output voltages are abnormal; PG=1, PSU output voltages are within specified limits
Vref1-11 - Controlled zener diode control electrode
Fb1-10 - Cathode controlled zener diode
GND-12 - Common wire
COMP-13 - Error amplifier output and PWM comparator negative input
IN-14 - Negative input of error amplifier
SS-15 - Positive input of the error amplifier, connected to an internal source Uref = 2.5 V. The output is used to organize a "soft start" of the converter
Ri-16 - Input for connecting an external resistor 75 kOhm
Vcc-1 - Supply voltage, connected to standby source + 5V_SB
PR-5 - Input for organizing PSU protection
Chip ULN2003 (ULN2003a) in essence, it is a set of powerful composite switches for use in inductive load circuits. It can be used to control loads of significant power, including electromagnetic relays, DC motors, electromagnetic valves, in various control circuits and others.
Brief description of ULN2003a. The ULN2003a is a high power output Darlington transistor assembly with protective diodes at the outputs, which are designed to protect control circuits from reverse voltage surge from an inductive load.
Each channel (Darlington pair) in the ULN2003 is rated for 500 mA and can handle a maximum current of 600 mA. The inputs and outputs are located opposite each other in the microcircuit housing, which greatly simplifies the layout of the printed circuit board.
ULN2003 belongs to the ULN200X family of chips. Different versions of this chip are designed for specific logic. In particular, the ULN2003 chip is designed to work with TTL logic (5V) and CMOS logic devices. ULN2003 is widely used in control circuits for a wide range of loads, as relay drivers, display drivers, line drivers, etc. ULN2003 is also used in stepper motor drivers.
Below is a list of what can replace ULN2003 (ULN2003a):
Often, the ULN2003 chip is used to control a stepper motor. Below is the connection diagram of ULN2003a and stepper motor.
