Diagram of an industrial laboratory power supply. Good DIY laboratory power supply
Hi all. Today is the final review, assembly of a laboratory linear power supply. Today there is a lot of metalwork, body fabrication and final assembly. The review is posted on the blog “DIY or Do It Yourself”, I hope I’m not distracting anyone here and preventing anyone from pleasing their eyes with the charms of Lena and Igor))). Anyone who is interested in homemade products and radio equipment - Welcome!!!
ATTENTION: Lots of letters and photos! Traffic!
Welcome radio amateur and DIY enthusiast! First, let's remember the stages of assembling a laboratory linear power supply. Directly to this review not relevant, so I posted it under a spoiler:
Assembly steps
Assembling the power module. Board, radiator, power transistor, 2 variable multi-turn resistors and a green transformer (from the Eighties®) As the wise one suggested kirich, I independently assembled a circuit that the Chinese sell in the form of a construction kit for assembling a power supply. At first I was upset, but then I decided that, apparently, the circuit is good, since the Chinese are copying it... At the same time, the childhood problems of this circuit (which were completely copied by the Chinese) came out; without replacing the microcircuits with more “high-voltage” ones, it is impossible to apply to the input more than 22 volts AC voltage... And several smaller problems that our forum members suggested to me, for which I thank them very much. Most recently, the future engineer" AnnaSun"suggested getting rid of the transformer. Of course, anyone can upgrade their power supply as they wish, you can also use a pulse generator as a power source. But any pulse generator (maybe except resonant ones) has a lot of noise at the output, and this interference will partially transfer to the LabBP output... What if there is pulse interference, then (IMHO) this is not LabBP. Therefore, I will not get rid of the “green transformer”. 
Because this linear block power supply, it has a characteristic and significant drawback: all the excess energy is released on the power transistor. For example, we supply 24V alternating voltage to the input, which after rectification and smoothing will turn into 32-33V. If a powerful load is connected to the output, consuming 3A at a voltage of 5V, all the remaining power (28V at a current of 3A), which is 84W, will be dissipated by the power transistor, turning into heat. One way to prevent this problem, and accordingly increase efficiency, is to install a manual or automatic switching windings This module was reviewed in: 
For convenience of working with the power supply and the ability to instantly turn off the load, a circuit was introduced additional module on a relay that allows you to turn the load on or off. This was dedicated to this. 
Unfortunately, due to the lack of the necessary relays (normally closed), this module did not work correctly, so it will be replaced by another module, on a D-trigger, which allows you to turn the load on or off using one button.
I'll tell you briefly about new module. The scheme is quite well known (sent to me in a private message): 
I slightly modified it to suit my needs and assembled the following board: 
WITH reverse side:
This time there were no problems. Everything works very clearly and is controlled with one button. When power is applied, the 13th output of the microcircuit is always logical zero, the transistor (2n5551) is closed and the relay is de-energized - accordingly, the load is not connected. When you press the button, a logical one appears at the output of the microcircuit, the transistor opens and the relay is activated, connecting the load. Pressing the button again returns the chip to its original state.
What is a power supply without a voltage and current indicator? That's why I tried to make an ampere-voltmeter myself. In principle, it turned out to be a good device, but it has some nonlinearity in the range from 0 to 3.2A. This error will not affect in any way when using this meter, say in charger for a car battery, but is unacceptable for a Laboratory power supply, therefore, I will replace this module with Chinese precision panel boards and displays with 5 digits... And the module I assembled will find use in some other homemade product. 
Finally, higher-voltage microcircuits arrived from China, as I told you about in. And now you can supply 24V to the input alternating current, without fear that it will break through the microcircuits... 
Now the only thing left to do is to make the body and assemble all the blocks together, which is what I will do in this final review on this topic.
Having searched for a ready-made case, I did not find anything suitable. The Chinese have good boxes, but, unfortunately, their price, and especially... 
The “toad” didn’t allow me to give the Chinese 60 bucks, and it’s stupid to give that kind of money for a body; you can add a little more and buy it. At least this PSU will make a good case.
So I went to the construction market and bought 3 meters of aluminum angle. With its help, the frame of the device will be assembled.
Preparing the details the right size. We draw out the blanks and cut off the corners using a cutting disc. . 
Then we lay out the blanks for the top and bottom panels to see what will happen. 
Trying to place the modules inside 
Assembly is carried out using countersunk screws (under the head with a countersink, a hole is countersunk so that the screw head does not protrude above the corner), and nuts on the reverse side. The outlines of the power supply frame are slowly appearing: 
And now the frame is assembled... It’s not very smooth, especially in the corners, but I think that painting will hide all the unevenness: 
Dimensions of the frame under the spoiler:
Dimensions
Unfortunately, there is little free time, so the plumbing work is progressing slowly. In the evenings, over the course of a week, I made a front panel from a sheet of aluminum and a socket for the power input and fuse. 
We draw out future holes for the Voltmeter and Ammeter. The seat size should be 45.5mm by 26.5mm
Paste over mounting holes masking tape: 
And with a cutting disc, using a Dremel, we make cuts (adhesive tape is needed so as not to go beyond the size of the sockets, and not spoil the panel with scratches) The Dremel quickly copes with aluminum, but it takes 3-4 for 1 hole 
There was a hitch again, it’s trivial, we ran out of cutting discs for the Dremel, a search in all the stores in Almaty did not lead to anything, so we had to wait for the discs from China... Fortunately, they arrived quickly in 15 days. Then the work went more fun and quickly...
I sawed holes for the digital indicators with a Dremel and filed them. 
We put a green transformer on the “corners” 
Let's try on a radiator with a power transistor. It will be isolated from the housing, since a transistor in a TO-3 housing is installed on the radiator, and there it is difficult to isolate the transistor collector from the housing. The radiator will be behind a decorative grille with a cooling fan. 
I sanded the front panel on a block. I decided to try on everything that would be attached to it. It turns out like this: 
Two digital meter, load switch button, two multi-turn potentiometers, output terminals and “Current Limit” LED holder. It seems like you forgot nothing? 
On the back of the front panel.
We disassemble everything and paint the power supply frame with black spray paint. 
We attach a decorative grille to the rear wall with bolts (purchased at the car market, anodized aluminum for tuning the radiator air intake, 2000 tenge (6.13USD)) 
This is how it turned out, view from the back of the power supply housing. 
We install a fan to blow the radiator with a power transistor. I attached it to plastic black clamps, it holds well, appearance does not suffer, they are almost invisible. 
We return the plastic base of the frame with the power transformer already installed. 
We mark the mounting locations for the radiator. The radiator is isolated from the device body, because the voltage across it is equal to the voltage at the collector of the power transistor. I think that it will be well blown by a fan, which will significantly reduce the temperature of the radiator. The fan will be controlled by a circuit that takes information from a sensor (thermistor) attached to the radiator. Thus, the fan will not “thresh” on empty, but will turn on when a certain temperature is reached on the radiator of the power transistor. 
We attach the front panel in place and see what happens. 
There was a lot of decorative grille left, so I decided to try to make a U-shaped cover for the power supply housing (in the manner computer cases), if you don’t like it, I’ll change it to something else. 
Front view. While the lattice is “baited” and does not yet fit tightly to the frame. 
It seems to be working out well. The grille is strong enough, you can safely put anything on top, but you don’t even need to talk about the quality of ventilation inside the case, the ventilation will be simply excellent compared to closed cases. 
Well, let's continue the assembly. We connect a digital ammeter. Important: do not step on my rake, do not use a standard connector, only solder directly to the connector contacts. Otherwise, it will be in place of the current in Amperes, showing the weather on Mars. 
The wires for connecting the ammeter, and all other auxiliary devices, should be as short as possible.
Between the output terminals (plus or minus) I installed a socket made of foil PCB. It is very convenient to draw insulating grooves in copper foil to create platforms for connecting all auxiliary devices (ammeter, voltmeter, load disconnect board, etc.) 
The main board is installed next to the heatsink of the output transistor. 
The winding switching board is installed above the transformer, which has significantly reduced the length of the wire loop. 
It's time to assemble the module additional food for winding switching module, ammeter, voltmeter, etc.
Since we have a linear analog power supply, we will also use the option on a transformer, no switching power supplies. :-)
We etch the board: 
Soldering the details: 
We test, install brass “legs” and build the module into the body: 
Well, all the blocks are built in (except for the fan control module, which will be manufactured later) and installed in their places. The wires are connected, the fuses are inserted. You can start the first time. We sign ourselves with the cross, close our eyes and give food...
There is no boom and no white smoke - that’s good... It seems like nothing is heating up at idle... We press the load switch button - the green LED lights up and the relay clicks. Everything seems to be fine so far. You can start testing. 
As they say, “soon the tale is told, but not soon the deed is done.” Swim up again underwater rocks. The transformer winding switching module does not work correctly with the power module. When the switching voltage from the first winding to the next occurs, a voltage jump occurs, i.e., when it reaches 6.4V, a jump occurs to 10.2V. Then, of course, you can reduce the tension, but this is not the point. At first I thought that the problem was in the power supply of the microcircuits, since their power is also from the windings power transformer, and accordingly increases with each subsequent connected winding. Therefore, I tried to supply power to the microcircuits from a separate power source. But it did not help.
Therefore, there are 2 options: 1. Completely redo the circuit. 2. Refuse the automatic winding switching module. I'll start with option 2. I can’t stay completely without switching the windings, because I don’t like putting up with the stove as an option, so I’ll install a toggle switch that allows you to select the supplied voltage to the power supply input from 2 options: 12V or 24V. This is, of course, a half-measure, but better than nothing at all.
At the same time I decided to change the ammeter to another similar one, but with green glow of the numbers, since the red numbers of the ammeter glow quite weakly and when sunlight they are hard to see. Here's what happened: 
It seems better this way. It is also possible that I will replace the voltmeter with another one, because... 5 digits in a voltmeter are clearly excessive, 2 decimal places are quite enough. I have replacement options, so there won't be any problems.
We install the switch and connect the wires to it. Let's check.
When the switch is in the down position - maximum voltage no load was about 16V 
When the switch is positioned up, the maximum voltage available for this transformer is 34V (no load) 
Now for the handles, I didn’t spend a long time coming up with options and found plastic dowels of a suitable diameter, both internal and external. 
We cut the tube to the required length and put it on the rods of the variable resistors: 
Then we put on the handles and secure them with screws. Since the dowel tube is quite soft, the handle is fixed very well; considerable effort is required to tear it off. 
The review turned out to be very large. Therefore, I will not take up your time and will briefly test the Laboratory power supply.
We already looked at interference with an oscilloscope in the first review, and since then nothing has changed in the circuitry.
Therefore, let’s check the minimum voltage, the adjustment knob is in the extreme left position: 
Now the maximum current 
Current limit 1A 
Maximum current limitation, current adjustment knob in the extreme right position: 
That's all for my dear radio destroyers and sympathizers... Thanks to everyone who read to the end. The device turned out to be brutal, heavy and, I hope, reliable. See you again on air!
UPD: Oscillograms at the output of the power supply when the voltage is turned on: 
And turn off the voltage: 
UPD2: Friends from the Soldering Iron forum gave me an idea on how to launch a winding switching module with minimal circuit modifications. Thank you all for your interest, I will finish the device. Therefore - to be continued. Add to favorites Liked +72 +134
Many different laboratory power supplies are presented on the Internet on radio engineering sites, although mostly simple designs. This same circuit is characterized by a fairly high complexity, which is justified by the quality, reliability and versatility of the power supply. We present in full homemade block bipolar power supply 2 x 30 V, with adjustable current up to 5 A and digital LED A/V meter.
Actually it's two identical blocks power supply in one housing, which significantly increases the functionality and capabilities of the device, allowing you to combine channel powers of up to 10 Amps. At the same time it is not typical symmetrical source power supply, although you can connect serial outputs here to get more high voltage or pseudo symmetry, considering common connection like a mass.
Diagrams of laboratory power supply modules
All power board circuits were designed from scratch, and all printed circuit boards are also independently developed. The first module “Z” is a diode bridge, voltage filtering, generating negative voltage to power operational amplifiers, 34 V positive voltage source direct current for operational amplifiers, powered by a separate auxiliary transformer, relay used to switch the main transformer windings controlled from another printed circuit board, and a 5V 1A power supply for power meters.
The "Z" modules of both units were designed to be nearly symmetrical (to fit better into the PSU case). Thanks to this, the ARK connectors were placed on one side to connect the wires and heatsink for the bridge rectifier, and the boards, as shown in the pictures, were placed symmetrically.
An 8-amp diode bridge is used here. The main transformers have dual secondary windings, each 14 V and a current of just over 5 A. The power supply was rated for 5 amps, but it turned out that at full voltage 30 V does not produce the full 5 A. However, there is no problem with a 5 amp load at lower voltage (up to 25 V).
The second module is an expanded version of the power supply with operational amplifiers.
Depending on whether the power supply is loaded or in standby mode, the voltage in the region of the amplifier U3, responsible for limiting the current, changes (with the same setting of the potentiometer limits). The circuit compares the voltage across potentiometer P2 with the voltage across resistor R7. Part of this voltage drop is applied to the inverse input of U4. Thereby output voltage depends on the potentiometer setting and is practically independent of the load. Almost because on a scale from 0 to 5 A the deviation is at the level of 15 mV, which in practice is enough to obtain a stable source for driving the LM3914 circuits that form the LED bar.
The visualization diagram is especially useful when multi-turn potentiometers are used for adjustment. It’s great that with the help of such a potentiometer you can easily set the voltage accurate to the third decimal place. Each LED in the line corresponds to a current of 0.25 A, so if the current limit is below 250 mA, the line is not displayed.
The way the ruler is displayed can be changed from a point to a ruler, but a point is selected here to avoid influencing too much large quantity light points and reduce energy consumption.
The next module is the winding switching system and fan control system that are installed on the radiators of old processors.
The circuits are powered by independent windings of an auxiliary transformer. Here we use m/s op-amp LM358, which contains two operational amplifiers inside. A BD135 transistor is used as a temperature sensor. After exceeding 55C, the fans turn on, and after cooling to approximately 50C, they automatically turn off. The winding switching system responds to the voltage value at the direct output terminals of the power supply and has a hysteresis of about 3 V, so the relay will not operate too often.
Measurement of load voltage and current is carried out using ICL7107 chips. The meter boards are double-sided and are designed such that for each power source there is a voltmeter and an ammeter on one board.
From the very beginning, the idea was to visualize the parameters of power supplies on seven-segment LED displays, because they are more readable than LCD. But nothing prevents you from measuring the temperature of radiators, winding switches and cooling systems on one Atmega MK, even for both power supplies at once. It's a matter of choice. Using a microcontroller will be cheaper, but as already mentioned above, this is a matter of taste.
All auxiliary systems are powered by a transformer that has been rewound by removing all windings except the 220V mains (primary). TS90/11 was used for this purpose.
The secondary winding is wound with 2 x 26 V AC to power the operational amplifiers, 2 x 8 V AC to power the indicators and 2 x 13 V to power the temperature control. A total of six independent windings were created.
Housing and assembly costs
The entire power supply is housed in a housing that was also designed from scratch. It was made to order. It is known that it is difficult to make a decent box (especially a metal one) at home.
Aluminum bezel used to mount all indicators and additional elements, was manufactured on a milling machine in accordance with the design.
Of course, this is not a low-budget implementation, given the purchase of two powerful toroidal transformers and custom-made housing. If you want something simpler and cheaper - .
The rest can be estimated based on prices in online stores. Of course, some elements were obtained from our own stock, but these too will need to be purchased, creating a power supply from scratch. The total cost was 10,000 rubles.
Assembly and configuration of LBP
- Assembling and testing a module with a bridge rectifier, filtering and relay, connecting to a transformer and activating a relay from an independent source to check the output voltages.
- Execution of the module for switching windings and monitoring radiator cooling. Running this module will make it easier to configure the future power supply. This will require a different power source to supply adjustable voltage to the input of the system responsible for controlling the relay.
- The temperature portion of the circuit can be tuned by simulating the temperature. For this purpose, a heat gun was used, which gently heated a radiator with a sensor (BD135). Temperature was measured using a sensor included in a multimeter (at that time there were no ready-made accurate temperature meters). In both cases, the setup comes down to selecting PR201 and PR202 or PR301 and PR302, respectively.
- We then run the power supply by adjusting RV1 to produce a 0V output, which is useful for setting current limiting. The limitation itself depends on the values of resistors R18, R7, R17.
- Regulation of A/V indicators comes down to adjusting the reference voltages between pins 35 and 36 of the ICL microcircuits. Voltage and current meters used an external reference source. In the case of temperature meters, such precision is not needed, and the display with a decimal point is still somewhat exaggerated. Transmission of temperature readings is carried out by one rectifier diode(there are three of them in the diagram). This is due to the PCB design. There are two jumpers on it.
- Directly at the output terminals, a voltage divider and a 0.01 Ohm / 5 W resistor are connected to the voltmeter, across which the voltage drop is used to measure the load current.
An additional element of the power supplies is a circuit that allows only one power supply to be turned on without the need for a second channel, despite the fact that the auxiliary transformer powers both channels of the power supply at once. On the same board there is a system for turning the power supply on and off using one low-current button (for each channel of the power supply).
The circuit is powered by an inverter, which in standby mode consumes about 1 mA from a 220 V network. All circuits in good quality you can
!
If you are looking for a simple and reliable linear power supply circuit, then this article is just for you. Here you will find full instructions on assembly and configuration of this power supply. The author of this homemade product is Roman (YouTube channel “Open Frime TV”).
First, a little background. Quite recently the author reworked his workplace and I wanted to install a linear unit as the third power supply, since sometimes he has to assemble circuits that cannot tolerate voltage ripple. And as we know, the linear unit has almost no voltage ripple at the output.
Until this moment, the author was not very interested in linear blocks, and somehow he did not particularly delve into this topic. When the idea to build such a block came, Roman immediately opened everyone’s favorite and widely known YouTube video hosting. As a result, after a lengthy search, the author was able to identify 2 schemes for himself. The author of the first is AKA KASYAN (author of the same name YouTube channel), and the second circuit is built on op-amps.
But since opamps can operate at voltages up to 32V, the output voltage accordingly could not exceed this limit, which means this circuit is no longer needed.
Okay, we can assemble the circuit from Kasyan, but even here we were disappointed. This scheme is afraid of static. This manifested itself as an explosion of transistors if you touched the output contacts.
This happened several times. And then the author decided to leave this diagram at rest. You will say that the Internet is full of linear power supply circuits.
Yes, this is undoubtedly true, but only these two schemes mentioned above had normally routed signets, which could simply be downloaded. Everything else is either without seals or assembled by hanging installation. And we (radio amateurs) are accustomed to the fact that everything is served on a silver platter.
The author decided to create a normal signet. The board turned out to be quite compact. After testing this scheme, surprisingly it performed well.
With such simplicity, the author liked it so much that he even decided to make a kit from this board. To do this, you need to convert the signet into a Gerber file (a file with the extension .gbr, which is a design of a printed circuit board for the subsequent production of photo masks on various equipment). Then you need to send the boards for manufacturing.
And now, a couple of weeks after ordering, we receive our long-awaited boards. Having opened the parcel and taken a closer look at the boards, we can make sure that everything turned out to be of very high quality and beautiful.
So, let's solder this board and check its operation. There are not so many components for installation; soldering takes about 20 minutes, no more.
We're done with soldering. We make the first switch on. And here we are in for a little disappointment. This board not without its problems. They manifested themselves in the fact that when the potentiometer knob is rotated to the left, the voltage and current increase, and when rotated to the right, a decrease occurs.
This happened because the author placed the resistors for this board on wires (for subsequent installation on the case) and there, without any problems, it was possible to change the direction of rotation simply by changing the side contacts. Well, okay, but everything else works as expected.
But nevertheless, the author corrected the signet, now when the potentiometer is rotated to the right, the voltage increases, everything is as it should be. So you can safely download and repeat this design (archive with this printed circuit board located in the description under the original video of the author, you must follow the SOURCE link at the end of the article).
Now let's move on to detailed consideration circuits and the board itself. You can see the diagram on your screens.
This power supply is equipped with a voltage and current regulator, as well as a short circuit protection system, which is simply necessary in such units.
Imagine for a moment what happens during a short circuit when the input voltage is 36V. It turns out that all the voltage is dissipated on the power transistor, which, of course, is unlikely to withstand such abuse.
Protection can be configured here. With the help of this trim resistor We set any operation current.
A 12V protection switch is installed here, and the input voltage can reach 40V. Therefore, it was necessary to obtain a voltage of 12V.
This can be implemented using a parametric stabilizer using a transistor and a zener diode. The zener diode is 13V, since there is a voltage drop across the collector-emitter transitions of the two transistors.
So, now you can start testing this linear power supply. We apply a voltage of 40V from laboratory block nutrition. On the load we hang a light bulb designed for a voltage of 36V, with a power of 100W.
Then we begin to slowly rotate the variable resistor.
As you can see, the voltage regulation works perfectly. Now let's try to regulate the current.
As can be observed, when the second resistor is rotated, the current decreases, which means that the circuit operates in normal mode.
Since this is a linear unit and all the “extra” voltage turns into heat, it needs quite a heatsink large sizes. Radiators from a computer processor have proven themselves to be excellent for these purposes. Such radiators have a large dissipation area, and if they are also equipped with a fan, then you can, in principle, completely forget about overheating of the transistor.
And now about how the protection works. We set the required current using a trimming resistor. In the event of a short circuit, the relay is activated. A pair of its contacts opens the output circuit and the transistor is safe.
To return to normal mode work there is such a button for opening, when pressed, the protection is removed.
Well, or you can simply disconnect the unit from the network and apply voltage again. This way the protection will also turn off. There are also 2 LEDs on the board. One signals about the operation of the unit, and the second about the activation of the protection.
To summarize, we can say that the unit turned out to be very cool and is suitable for both beginners and experienced radio amateurs. So download the archive and collect such a block for yourself.
Well, that's all. Thank you for attention. See you again!
Video:
To configure or repair radio devices, you must have several power sources. Many people already have such devices at home, but, as a rule, they have limited operational capabilities (the permissible load current is up to 1 A, and if current protection is provided, it is inertial or without the ability to regulate - trigger). In general, such sources cannot compete with their technical characteristics industrial blocks nutrition. It is quite expensive to purchase a universal laboratory industrial source.
Use of modern circuit technology and element base allow you to make a power source at home, which in its basic technical characteristics is not inferior to the best industrial designs. At the same time, it can be simple to manufacture and configure.
The basic requirements that such a power source must satisfy are: voltage regulation in the range 0...30 V; ability to provide load current up to 3 A with minimal ripple; trigger adjustment current protection. In addition, the current protection must operate quickly enough to prevent damage to the source itself in the event of a short circuit at the output.
The ability to smoothly adjust the current limit in the power supply allows you to configure external devices prevent their damage.
All these requirements are satisfied by the universal power supply circuit proposed below. In addition, this power supply allows you to use it as a source of stable current (up to 3 A).
Basic specifications power supply:
smooth voltage regulation in the range from 0 to 30 V;
ripple voltage at a current of 3 A is not more than 1 mV;
smooth adjustment of current limitation (protection) from 0 to 3 A;
voltage instability coefficient is not worse than 0.001%/V;
current instability coefficient is not worse than 0.01%/V;
The source efficiency is no worse than 0.6.
Electrical diagram of the power supply, fig. 4.10, consists of a control circuit (node A1), a transformer (T1), a rectifier (VD5...VD8), a power control transistor VT3 and a switching unit for transformer windings (A2).
The control circuit (A1) is assembled on two universal operational amplifiers (op-amps), located in one housing, and is powered by a separate winding of the transformer. This allows the output voltage to be adjusted from zero to more stable work the entire device. And to facilitate the thermal operation of the power control transistor, a transformer with a sectioned secondary winding. The taps are automatically switched to
depending on the output voltage level using relays K1, K2. This allows, despite the high current in the load, to use a heat sink for VT3 of small size, as well as increase the efficiency of the stabilizer.
The switching unit (A2), in order to ensure switching of four transformer taps using only two relays, connects them to next sequence: when the output voltage exceeds 7.5 V, K1 turns on; when the level exceeds 15 V, K2 is turned on; if 22 V is exceeded, K1 is switched off (in this case, the maximum voltage is supplied from the transformer windings). The specified thresholds are set by the zener diodes used (VD11...VD13). The relay is turned off when the voltage drops in the reverse order, but with a hysteresis of approximately 0.3 V, i.e. when the voltage drops to this value lower than when turned on, which eliminates chatter when switching windings.
The control circuit (A1) consists of a voltage stabilizer and a current stabilizer. If necessary, the device can operate in any of these modes. The mode depends on the position of the regulator "I" (R18).
The voltage stabilizer is assembled using elements DA1.1-VT2-VT3. The stabilizer circuit works as follows. The required output voltage is set by resistors “coarsely” (R16) and “finely” (R17). In voltage stabilization mode, the signal feedback voltage (-Uoc) from the output (X2) through a divider of resistors R16-R17-R7 is supplied to the non-inverting input operational amplifier DA1/2. A reference voltage of +9 V is supplied to the same input through resistors R3-R5-R7. When the circuit is turned on, the positive voltage at output DA1/12 will increase (it comes to control VT3 through transistor VT2) until the voltage at the output terminals X1-X2 will not reach the level set by resistors R16-R17. Due to the negative voltage feedback coming from output X2 to the input of amplifier DA1/2, the output voltage of the power supply is stabilized.
It took one day to develop this power supply, during the same day it was implemented, and the whole process was filmed on a video camera. A few words about the scheme. This is a stabilized power supply with output voltage regulation and current limitation. Schematic features allow you to reduce the minimum output voltage to 0.6 Volts, and the minimum output current to around 10 mA.
Despite the prostate design, this block even good laboratory power supplies with a cost of 5-6 thousand rubles are inferior! The maximum output current of the circuit is 14 Amperes, the maximum output voltage is up to 40 Volts - no longer worth it.
Quite smooth current limitation and voltage regulation. The block also has fixed protection against short circuits; by the way, the current protection can also be set (almost all industrial designs lack this function), for example, if you need the protection to operate at currents up to 1 Ampere, then you just need to set this current using trigger current setting regulator. The maximum current is 14A, but this is not the limit.
As a current sensor, I used several 5 watt 0.39 Ohm resistors connected in parallel, but their value can be changed based on the required protection current, for example - if you are planning a power supply with a maximum current of no more than 1 Ampere, then the value of this resistor is around 1 Ohm at power 3W.
At short circuits the voltage drop across the current sensor is sufficient to trigger transistor BD140. When it opens, the lower transistor - BD139 - also triggers, through open passage which supplies power to the relay winding, as a result of which the relay is activated and the working contact opens (at the output of the circuit). The circuit can remain in this state for any amount of time. Along with the protection, the protection indicator also works. In order to remove the block from protection, you need to press and lower button S2 according to the diagram.
Protection relay with a 24 Volt coil with a permissible current of 16-20 Amperes or more.
In my case, the power switches are my favorite KT8101 installed on the heat sink (there is no need to additionally isolate the transistors, since the key collectors are common). You can replace the transistors with 2SC5200 - a complete imported analogue or with KT819 with the GM index (iron), if desired, you can also use KT803, KT808, KT805 (in iron cases), but the maximum output current will be no more than 8-10 Amperes. If a unit is needed with a current of no more than 5 Amps, then one of the power transistors can be removed.
Low-power transistors like BD139 can be replaced with a complete analog - KT815G (you can also use KT817, 805), BD140 - with KT816G (you can also use KT814).
There is no need to install low-power transistors on heat sinks.
In fact, only a control (adjustment) and protection circuit (working unit) is presented. As a power supply I used modified computer blocks power (series connected), but you can use any network transformer with a power of 300-400 watts, a secondary winding of 30-40 Volts, a winding current of 10-15 Amps - this is ideal, but transformers of lower power are also possible.
Diode bridge- any, with a current of at least 15 Amps, voltage is not important. You can use ready-made bridges; they cost no more than 100 rubles.
In 2 months, over 10 such power supplies were assembled and sold - no complaints. I assembled exactly such a power supply for myself, and as soon as I didn’t torture it, it was indestructible, powerful and very convenient for any task.
If anyone wants to become the owner of such a power supply unit, I can make it to order, contact me at
