LM2596 is a step-down DC-DC voltage converter. Radio for everyone - LBP on lm2576 Lm2596 power supply circuit

Step-down DC-DC converters are increasingly finding their use in everyday life, households, automotive applications, and also as regulated power supplies in a home laboratory.

For example, on a heavy-duty vehicle, the voltage of the on-board cable network may be +24V, but you need to connect a car radio or other device with an input voltage of +12V, then such a step-down converter will be very useful to you.

Many people order step-down DC-DC converters from various Chinese sites, but their power is quite limited, due to the Chinese saving on the cross-section of the winding wire, semiconductor devices and inductor cores, because the more powerful the converter, the more expensive it is. Therefore, I suggest you assemble a step-down DC-DC yourself, which will surpass the Chinese analogues in power and will also be more economical. According to my photo report and the presented diagram, it is clear that assembly will not take much time.

The LM2596 chip is nothing more than a switching step-down voltage regulator. It is available in both fixed voltage (3.3V, 5V, 12V) and adjustable voltage (ADJ). Our step-down DC-DC converter will be built on the basis of an adjustable microcircuit.

Converter circuit

Basic parameters of the LM2596 regulator

Input voltage………. up to +40V

Maximum input voltage………. +45V

Output voltage………. from 1.23V to 37V ±4%

Generator frequency………. 150kHz

Output current………. up to 3A

Current consumption in Standby mode………. 80uA

Operating temperature from -45°С to +150°С

Housing type TO-220 (5 pins) or TO-263 (5 pins)

Efficiency (at Vin= 12V, Vout= 3V Iout= 3A).......... 73%

Although the efficiency can reach 94%, it depends on the input and output voltage, as well as on the quality of the winding and the correct selection of the inductor inductance.

According to the graph taken from, with an input voltage of +30V, an output voltage of +20V and a load current of 3A, the efficiency should be 94%.

Also, the LM2596 chip has current and overheat protection. I note that on non-original microcircuits these functions may not work correctly or may be completely absent. A short circuit at the output of the converter leads to failure of the microcircuit (tested on two LMs), although there is nothing surprising here; the manufacturer does not write in the datasheet about the presence of short circuit protection.

Schematic elements

All element ratings are indicated on the electrical circuit diagram. The voltage of capacitors C1 and C2 is selected depending on the input and output voltage (input (output) voltage + margin of 25%), I installed the capacitors with a margin of 50V.

Capacitor C3 is ceramic. Its denomination is selected according to the table from the datasheet. According to this table, the capacitance C3 is selected for each individual output voltage, but since the converter in my case is adjustable, I used a capacitor of average capacity 1nF.

Diode VD1 must be a Schottky diode, or another ultra-fast diode (FR, UF, SF, etc.). It must be designed for a current of 5A and a voltage of at least 40V. I installed a pulse diode FR601 (6A 50V).

Choke L1 must be rated for a current of 5A and have an inductance of 68 μH. To do this, take a core made of powdered iron (yellow-white), outer diameter 27mm, inner 14mm, width 11mm, your dimensions may vary, but the larger they are, the better. Next, we wind two wires (the diameter of each wire is 1 mm) 28 turns. I wound a single core with a diameter of 1.4 mm, but with a high output power (40W), the inductor got very hot, also due to the insufficient cross-section of the core. If you wind two wires, then you won’t be able to put the winding in one layer, so you need to wind it in two layers, without insulation between the layers (if the enamel on the wire is not damaged).

A small current flows through resistor R1, so its power is 0.25W.

Resistor R2 is tuning, but can be replaced with a constant one; for this, its resistance is calculated for each output voltage according to the formula:

Where R1 = 1kOhm (according to the datasheet), Vref = 1.23V. Then, let's calculate the resistance of resistor R2 for the output voltage Vout = 30V.

R2 = 1 kOhm * (30V/1.23V - 1) = 23.39 kOhm (reducing to the standard value, we get resistance R2 = 22 kOhm).

Also, knowing the resistance of resistor R2, you can calculate the output voltage.

Testing a step-down DC-DC converter on LM2596

During testing, a radiator with an area of ​​≈ 90 cm² was installed on the chip.

I carried out tests on a load with a resistance of 6.8 Ohms (a constant resistor lowered into water). Initially, I applied a voltage of +27V to the converter input, the input current was 1.85A (input power 49.95W). I set the output voltage to 15.5V, the load current was 2.5A ( output power 38.75W). The efficiency was 78%, which is very good.

After 20 min. During operation of the step-down converter, diode VD1 heated up to a temperature of 50°C, inductor L1 heated up to a temperature of 70°C, and the microcircuit itself heated up to 80°C. That is, all elements have a temperature reserve, except for the throttle, 70 degrees is too much for it.

Therefore, to operate this converter at an output power of 30-40W or more, it is necessary to wind the inductor with two (three) wires and select a larger core. The diode and microcircuit can maintain a temperature of 100-120°C for a long time without any fears (except for heating everything nearby, including the case). If desired, you can install a larger radiator on the microcircuit, and you can leave long leads on the VD1 diode, then heat will be dissipated better, or attach (solder to one of the leads) a small plate (radiator). You also need to tin the tracks of the printed circuit board as best as possible, or solder a copper core along them, this will ensure less heating of the tracks during long-term operation at a high output power.

Laboratory power supply based on the LM2576T-ADJ switching stabilizer with output voltage regulation 0-30V and current 0-3A , with the function of limiting the output current and indicating the limitation mode using an LED.

We have all been familiar with linear voltage stabilizers for a very long time, especially three-terminal ones in TO-220 packages such as 7805, 7812, 7824 and LM317. They are inexpensive and easily available. Their low noise and fast transient response make them ideal for many applications. But they have one drawback - inefficiency (very low efficiency). For example, when a voltage of 12V is applied to the 7805 stabilizer and a load current of 1A, the stabilizer will dissipate 7W of power with a load power of 5W. Therefore, a large radiator is required to cool the stabilizer itself. When efficiency is important, such as when running on battery power, a switching regulator must be selected. In fact, the most modern equipment uses switching power supplies and switching regulators or stabilizers. But many radio amateurs shy away from switching regulators, since, for example, using the popular LM3524 requires a large number of external parts and an external switching transistor. In addition, there are strict requirements for the inductor. How to choose the right one, and where to get them? Fortunately, the newer switching regulator type LM2576 from National Semiconductor's allows you to assemble a high-efficiency switching regulator as easily as using 7805, etc. The microcircuit is available in a five-pin conventional TO-220 package and a TO-263 package for surface mounting. Supply voltage range 7-40V direct current. Efficiency - up to 80%. Output current - up to 3A and for several voltages (3.3V, 5 V, 12V, 15V), as well as in an adjustable output voltage version, which is of particular interest to us.When designing using a switching stabilizer, the board is small in size; in addition, a radiator with a small surface area is required, usually no more than 100 cm2. The stabilizer conversion frequency is 52 kHz. There is a series of high-voltage stabilizers marked HV with an input voltage range of 7-60V and the ability to adjust the output voltage up to 55V.

The diagram shown in the figure of a laboratory power supply based on a switching stabilizer LM2576T-ADJ with adjustable output voltage in the range of 0-30V and the ability to limit the load current in the range of 0-3A was found on the Internet and discussed in detail on the forum of the site http://vrtp.ru . By the way, a wonderful site, I recommend visiting it :) The LED glow indicates that the output current limiting mode has been activated, which is very convenient when checking and repairing radio-electronic devices.

To facilitate the operation of the 7805 stabilizer (in the TO-92 housing) and to increase the upper limit of the voltage Uin, a zener diode VD1 is installed in series with U2. The current and voltage regulation circuit is assembled on a dual comparator LM393. The first half U3.1 houses a voltage regulator, and the second half U3.2 houses a current regulator. The transistor switch Q1 contains an assembly indicating the activation of the output current limiting mode. Rated current The choke must be selected at least as high as the load current. It is possible to power the low-current part of the circuit from a separate voltage source and supply it directly to input U2, while the zener diode VD1 is not installed. Works well with low resistance loads. Without changing the scheme, you can use it pulse stabilizers LM2596T-ADJ with a conversion frequency of 150 kHz and a supply voltage range of 4.5-40V. Output current - up to 3A. Efficiency - up to 90%.

Dimensions of the printed circuit board of the power supply 72x52 mm, distance between the axes variable resistors 30 mm:

A video of the stabilizer in action (without words) is given below. Since the assembly and testing of the device was carried out in Donetsk at a time when shells were exploding outside the window, there was no desire to tell anything. And I didn’t want to collect it, but I needed to somehow escape from reality. I hope you understand me.

Cost of a printed circuit board with mask and markings: finished :)

Cost of a set of parts with a printed circuit board for assembling a power supply (without radiator): temporarily out of stock :(

Cost of assembled and tested power supply board (without radiator): temporarily out of stock :(

Brief description, diagram and list of kit components

For purchase printed circuit boards, kits for assembly and ready-made blocks, please contact or

Good luck to everyone, peaceful skies, good luck, 73!

Someone might think: An old horse won’t spoil the furrows... And we will answer: but it won’t plow deep either.
Therefore, I offer you a review of a step-down voltage converter based on the MP1584 chip. The seller positions the finished boards as an improved alternative to LM2596 converters. In my previous review, I encountered a wild discrepancy with the stated parameters. The actual values ​​did not satisfy me and at the end of the review I mentioned that I ordered more advanced boards for testing.

So, we meet:

Delivery and appearance:
Considering the cheap cost of the order, I was not surprised to find a bag of baby bumps in my mailbox. Inside there were 2 boards sealed in an antistatic bag. Which was quite expected. I later signed it myself with a felt-tip pen so that I wouldn’t forget the stated parameters.


Board dimensions 22x17mm, height 4mm.
Solder pads. There are no holes for mounting.
There are no traces of flux, soldering is acceptable. I looked through a magnifying glass and found no defects; unfortunately, I myself am not able to solder like that. Under the microcircuit and the inductor there are metallized holes for better heat dissipation.

Comparison with LM2596:
The size difference is significant. True, due to the size of the board, the heat dissipation efficiency is lower, but the efficiency is stated to be up to 96%

Documentation and diagram:
Documentation in in electronic format you can see it here
An almost standard Schottky diode 40V, 3A is used, which, by the way, held up well on the tested board.
A choke with an inductance of 8.2 μH, which, according to Table 3 of the datasheet, indicates better operating efficiency of the converter at an output voltage of 3.3V and slightly worse at 5V. Resistor R3 on the board is 100 kOhm, according to the specification, the optimal output voltage is 1.8V. Once again I am convinced that all these boards are assembled from what was at hand, making production as cheap as possible.
Typical connection diagram:


Specific board diagram:


A break in the tuning resistor will produce at the output the maximum voltage for which the divider R1 R2 is configured. In this case, up to 20 Volts. And that's bad.

Initially I thought that the purchased board had ceramic capacitors at the input and output instead of electrolytic capacitors. But in fact it turned out that the electrolytes are 12-13 uF:


Also, instead of resistor R1, a tuning resistor is installed to adjust the output voltage. By the way, it is very unreliable, it is difficult to set the exact voltage. At the slightest mechanical load, the tension can “float away”. This problem can be solved in several ways: a drop of nail polish or enamel-type paint to fix the contact pads of the trimmer resistor


or replacing the “trimmer” with a constant resistor.
In a particular case, you can do this - set the tuning resistor to required voltage, unsolder it and install an equivalent constant resistance.

An interesting point: by controlling the input of microcircuit 2(EN) using a logic level, you can switch the microcircuit to stop-start mode, i.e. You can control the operation of the microcircuit from the outside and turn on or off the load accordingly.

An important fact is the conversion frequency: It is set by a resistor connected to pin 6 of the microcircuit and typically has a resistance of 200 kOhm, but 100 kOhm is installed on the board. Formula for setting the conversion frequency:

I asked at work to check the conversion frequency - they said about 950 KHz. An abundance of 104 resistors, unification, what to do. The frequency corresponds to the set resistance.

Efficiency:


The seller claims efficiency up to 96% and again it’s a lie. The maximum efficiency that can be squeezed out is no more than 88%. Moreover, it is maximum with a supply voltage of about 12 Volts and a load range of 0.5-2 Amperes.

Tests:
To begin with, measure the current consumption at idle 0.22mA. Not bad.


As a load I used 2 resistors 3.3 and 2.2 Ohms. Due to the strong heating, the latter were placed in a container with water during testing.


On this moment The thermal imager was not available, it was rented to another facility, so the temperature was measured with a fairly popular pyrometer.


Accuracy within a couple of degrees.

Test switching is carried out without a load to set the required output voltage to avoid failure of the board or load.


We give the load and leave it in operation:


In a couple of minutes I heard converter operation. Well, when I heard it, the radio connected to the same power supply began to hiss, and interference appeared. The voltage control began to show periodic drops in the output voltage by 10-15%. The thermal protection of the microcircuit worked and the converter periodically began to skip cycles. Computer geeks use thermite throttling
Thinking that a higher input voltage should make it easier for the converter to operate without interruption, I connected the converter to a 24 Volt power supply. The first switch on - a click and a hole appeared in the microcircuit (later, when I started studying the documentation, I realized that the efficiency had dropped a little and I simply finished off the microcircuit, which was already suffering from overheating).
There was no magic smoke. To the credit of the converter, there was no voltage at the output.

In order not to burn the second and last board, it was decided to use a radiator and install it using thermal sealant on reverse side fees.
Thermal sealant star 922 is familiar to many. I use it to fix LEDs. Not the best of course, but at least something.
Radiator:


On the reverse side, so that the radiator would not short-circuit the contacts on the board, I ground off a part with a file. For visual perception I painted over it with a marker:


This is what a board with a heatsink looks like (sawed off from a large one used in ATX power supplies)

Temperature measurements were summarized in a mini table:
For testing I chose the most common voltages in digital logic, 5V and 3.3V. The input voltage from the stand, taking into account the drop on the wires, is 11.5-11.7 Volts. Resistors are normal 5%. I rounded the current to tenths because I focused on temperature: t1 is the maximum temperature on the board from the parts side. t2 is the maximum temperature on the back side of the board.

Each time I let the board work for about 10 minutes, I measured the temperature. The measurements were taken repeatedly over the entire surface of the board at a distance of 1 cm, only the maximum value was taken into account. In 100% of cases, the hottest element on the board was the microcircuit.
With a load of 2.2 Ohms and an output voltage of 5V, measurements without a radiator were not carried out, since the microcircuit exploded on the first copy of the converter.


It was noticed that the output voltage increased under load at a given 3.3V (no load) to 3.45V. This was not observed when testing at a 5V output.

Unfortunately, an oscilloscope is not available and there is no way to view the output signal, but this drawback will be eliminated in the near future. Since I finally crushed my toad and ordered a kit oscilloscope DSO062.

Recommendations for use:
When the load current is higher than 1A, it is advisable to install a small radiator, maybe half of what I used. Quite enough. Fixing the trimmer resistor with varnish. When used in conjunction with a VHF receiver, use additional ceramic capacitors to filter out power supply noise.

Conclusions:
Pros:
Compactness. If you don’t “squeeze” it to the maximum from the converter, then it’s quite functional. Sufficiently high efficiency and wide voltage range. Turning on the converter can be controlled externally (minor alteration of the board is required - soldering the conductor). If the microcircuit fails, no input voltage is detected at the output of the converter (perhaps this is a special case).
Minuses:
I didn’t like the power supply marking only on the reverse side. The seller praised the board, it also does not meet the declared characteristics. Minor modification required efficient work. In addition, there is interference in the VHF FM range (noise and whistling can be heard on the radio, especially in marginal operating modes). The trimming resistor leaves much to be desired; it is optimal to replace it with a multi-turn or constant resistor (if you need one fixed output voltage).
UPD: I will continue to choose converters, which one do you recommend: KIS-3R33S, XM1584, MP2307 are other options, requirements are 5V output and 3A current without significant modifications?

Your comments on the review will be addressed in a timely manner and will help me in the future.

Quite a long time ago, while sitting in the car, I thought: why am I charging my phone through a car charger installed in the cigarette lighter. After all, there is often more than one “consumer”, and the cigarette lighter socket itself is sometimes needed. I formulated the specifications for myself: power supply from the on-board network through the ignition switch, output of 1-3 ports with a current of up to 2 A. I searched on the Internet and it turned out that I was far from the first who was puzzled by the problem and, even more, implemented it in various ways.

For my idea, I needed a voltage stabilizer that could withstand the on-board voltage and current up to 3 Amps. There are actually a huge number of implementation options, but they all boil down to one thing - a pulsed step-down converter. Why impulse? Because it has maximum efficiency. This means there will be almost nothing to heat up in the converter and the dimensions promise to be minimal.

A step-down converter is designed to reduce the voltage to the required value. Its power elements operate in key mode, simply on and off. At the moment of switching on, energy is accumulated by the inductor (coil on the core), at the moment when power element(transistor) is turned off, the inductor releases the stored energy to the load. As soon as the inductor releases the accumulated energy, the circuit that controls the output voltage will turn on the power transistor and the process will repeat.
IN currently All chargers for phones and tablets that are inserted into the cigarette lighter socket are made according to a circuit with a pulse step-down converter.

Delivery and appearance:
The board arrived in a sealed antistatic bag, which seems like a reason to rejoice, but in fact it should be taken for granted.
The soldering quality is quite good. Minor flux residue on the reverse side of the variable resistor terminals.
The variable multi-turn resistor allows you to accurately adjust the output voltage.


Mounting holes for screws are provided. There are no terminal blocks, the wires will have to be soldered. Under the chip there are holes with metallization for additional heat removal to the back side of the board.

The scheme couldn't be simpler:

The only thing is that the Chinese have different ratings for the inductor and capacitors. Apparently, whatever is available, they install it. It can't get any worse.

I quickly soldered the wires and the load in the form of a 2.2 Ohm 10 W wirewound resistor.
To limit the temperature during heating, the resistor was placed in water.


There are 2 voltages available at the stand: 12 Volts and 24 Volts. The first switch-on was carried out without a load, to adjust the output voltage so as not to burn the scarf. By rotating the screw of the resistor, I achieved an output voltage of 5 Volts.
A load of 2.2 Ohms implies a current of 2.27 Amps, which fits into the stated parameters of the board as well as my needs with a small margin, since I got hold of a dual connector from a dead motherboard:

1 Ampere per port.

10 minutes of work under load and the board heats up wildly. Photo from thermal imager:

back side

Achtung! Temperature is 115C on the diode and 110C on the microcircuit (side with parts) and 105C on the reverse side.
The throttle temperature is about 70C, a bit too much, but it does not reach saturation.
The maximum temperature for the diode is 150C, and for the microcircuit 125C.

Doesn't fit into any gates. I started to think that this was a defect or that I had once again bought cheap crap.
and discovered that this converter has lousy efficiency. And all due to the fact that the key element in the microcircuit is a bipolar transistor, which, although it operates in the key mode, when open, the voltage across it drops quite a bit.
Increasing the input voltage to 24 Volts did not help the situation.
Efficiency graph at a load current of 3 Amps:


Those. approximately 80% when powered from the vehicle's onboard network. The output on the microcircuit is released at a load of 3 A 3.7 W, and the diode and inductor also heat up. Replacing the diode (3A 40V) and inductor (47 μH), as well as installing a radiator, could solve the heating problem, but why such effort when you can get more advanced step-down converters for the same money.

An attempt to correct the situation:
I installed a small radiator on the reverse side through heat-conducting glue (I sawed the radiator from a faulty computer power supply).




I planned to take the diode from the “duty room” there. It’s a little more complicated with the inductor, but I think I could find one with a larger cross-section of the winding wire (taking into account the decent spread of inductance in the inductors used by the Chinese).
An attempt to turn on and take temperature readings led to a crash =) I mixed up the polarity and burned the microcircuit. I saved money; I had to immediately take 5 of them for experiments, but it would be better not to take them at all, because this ancient converter is so terrible that it doesn’t even work out 50% of the characteristics in the specific board used.

In the vastness of the network I discovered an atypical use of the LM2596 microcircuit - an amplifier audio frequency class D! The signal is supplied to input 4 " Feedback" The frequency of discrediting is really no more than 150 KHz. In no case is there a call to assemble an amplifier based on a converter; there are specialized microcircuits for this =)

The conclusions are disappointing:
The board as sold does not justify the stated characteristics. Moreover, the dependence on the load current is much higher than on the change in voltage. You can modify the board by replacing half of the parts, but what's the point?

However, if you need a buck converter ( step down), then the best alternative to the one under review would be converters assembled on microcircuits: LM2577, LM 2678 and similar. At the moment I have already ordered several boards to try

While I had been planning to install USB ports on the car for a very long time, my machine went to scrap :(


but still there was still a place where I would put the converter instead of the transformer power supply:
This time (where the creative inscription is):


These are two (front bar with USB ports plexiglass “case” walls torn from an old computer case):


Specially for the review I made a load plate for testing chargers(I even burned a couple, they couldn’t stand the load). On Ali these are sold ready-made for about $1:

It turns out that on the LM2596 microassembly you can easily assemble a full-featured stabilized power supply that can be used in almost any laboratory block power supply with protection against possible short circuit.

Maximum permissible characteristics and properties:

Foreign analogues: A complete analogue of this microcircuit is the MIC4576BU chip

Typical microcircuit connection circuit:

All components of the circuit used to assemble the structure in the first version correspond in nominal values ​​to those indicated in the datasheet (see the archive at the link above), only the tuning resistance of fifty kilo-ohms could not be found, so instead of it there is a resistance of 47 kilo-ohms. The advantage of this voltage stabilizer can be considered minimal heating at high currents, which is something that typical KRENOK and LM317 microassemblies cannot boast of.

Additionally, a signal can be sent to the fifth leg of the microassembly to turn off the device.

Option 2 - Adjustable voltage regulator based on the LM2596T chip

LM2596T, operating in pulse mode, has a fairly high efficiency and allows currents with a nominal value of up to 2 A to flow through itself, without requiring a heat sink. For high load currents, it is necessary to use a radiator with a surface area of ​​at least 100 cm2. In addition, the radiator should be fixed to the microassembly using heat-conducting paste type KPT-8.

The circuit can be configured for any other fixed output voltage, that is, use the stabilizer as a DC-DC converter. To do this, you need to replace resistance R2 with a resistor calculated using the following mathematical formula:

R 2 = R 1 ×(V out / V ref-1)
or R 2 = 1210×(V out /1.23 - 1)

If you connect this design to a network step-down transformer with