Fully automatic battery charger. Charger for car battery on TL494 Tl494 pulse charger simple

28.06.2021

If the whistling sound stops at low currents and with a further increase in the load current, the output transistor begins to heat up sharply, then the area of the magnetic core is insufficient to operate at the selected generation frequency - it is necessary to increase the operating frequency of the microcircuit by selecting resistor R4 or capacitor C3 or install a larger inductor. In the absence of a power transistor of the p-n-p structure, powerful transistors can be used in the circuit n-p-n structures, as it shown on the picture.

As a diode VD5 in front of inductor L1, it is advisable to use any available diodes with a Schottky barrier, rated for a current of at least 10A and a voltage of 50V; in extreme cases, you can use mid-frequency diodes KD213, KD2997 or similar imported ones. For the rectifier, you can use any powerful diodes with a current of 10A or a diode bridge, for example KBPC3506, MP3508 or the like. It is advisable to adjust the shunt resistance in the circuit to the required value. The range of adjustment of the output current depends on the ratio of the resistances of the resistors in the output circuit 15 of the microcircuit. In the lower position of the current control variable resistor slider in the diagram, the voltage at pin 15 of the microcircuit must coincide with the voltage on the shunt when the maximum current flows through it. The variable current control resistor R3 can be set with any nominal resistance, but you will need to select a fixed resistor R2 adjacent to it to obtain the required voltage at pin 15 of the microcircuit.
The variable output voltage adjustment resistor R9 can also have a wide range of nominal resistance 2 ... 100 kOhm. By selecting the resistance of resistor R10, the upper limit of the output voltage is set. The lower limit is determined by the ratio of the resistances of resistors R6 and R7, but it is undesirable to set it less than 1 V.

The microcircuit is installed on a small printed circuit board 45 x 40 mm, the remaining elements of the circuit are installed on the base of the device and the radiator.

The wiring diagram for connecting the printed circuit board is shown in the figure below.

The circuit used a rewound TS180 power transformer, but depending on the magnitude of the required output voltages and current, the power of the transformer can be changed. If an output voltage of 15 V and a current of 6 A is sufficient, then a power transformer with a power of 100 W is sufficient. The radiator area can also be reduced to 100...200 cm2. The device can be used as a laboratory power supply with adjustable output current limitation. If the elements are in good working order, the circuit starts working immediately and only requires adjustment.

Source: http://shemotechnik.ru

Who has not encountered in their practice the need to charge a battery and, disappointed in the lack of a charger with the necessary parameters, was forced to purchase a new charger in a store, or reassemble the necessary circuit?
So I have repeatedly had to solve the problem of charging various batteries when there was no suitable charger at hand. Accounted for a quick fix collect something simple, in relation to a specific battery.

The situation was tolerable until the need for mass preparation and, accordingly, charging the batteries arose. It was necessary to produce several universal chargers - inexpensive, operating in a wide range of input and output voltages and charging currents.

The charger circuits proposed below were developed for charging lithium-ion batteries, but it is possible to charge other types of batteries and composite batteries (using the same type of cells, hereinafter referred to as AB).

All presented schemes have the following main parameters:
input voltage 15-24 V;
charge current (adjustable) up to 4 A;
output voltage (adjustable) 0.7 - 18 V (at Uin=19V).

All circuits were designed to work with power supplies from laptops or to work with other power supplies with DC output voltages from 15 to 24 Volts and were built on widespread components that are present on the boards of old computer power supplies, power supplies of other devices, laptops, etc.

Memory circuit No. 1 (TL494)

The memory in Scheme 1 is a powerful pulse generator operating in the range from tens to a couple of thousand hertz (the frequency varied during research), with an adjustable pulse width.
The battery is charged by current pulses limited by feedback formed by the current sensor R10, connected between the common wire of the circuit and the source of the switch on the field-effect transistor VT2 (IRF3205), filter R9C2, pin 1, which is the “direct” input of one of the error amplifiers of the TL494 chip.

The inverse input (pin 2) of the same error amplifier is supplied with a comparison voltage, regulated by a variable resistor PR1, from a reference voltage source built into the chip (ION - pin 14), which changes the potential difference between the inputs of the error amplifier.
As soon as the voltage value on R10 exceeds the voltage value (set by variable resistor PR1) at pin 2 of the TL494 microcircuit, the charging current pulse will be interrupted and resumed again only at the next cycle of the pulse sequence generated by the microcircuit generator.
By thus adjusting the width of the pulses on the gate of transistor VT2, we control the battery charging current.

Transistor VT1, connected in parallel with the gate of a powerful switch, provides the necessary discharge rate of the gate capacitance of the latter, preventing “smooth” locking of VT2. In this case, the amplitude of the output voltage in the absence of a battery (or other load) is almost equal to the input supply voltage.

With an active load, the output voltage will be determined by the current through the load (its resistance), which allows this circuit to be used as a current driver.

When charging the battery, the voltage at the switch output (and, therefore, at the battery itself) will tend to increase over time to a value determined by the input voltage (theoretically) and this, of course, cannot be allowed, knowing that the voltage value of the lithium battery being charged should be limited to 4.1V (4.2V). Therefore, the memory uses a threshold device circuit, which is a Schmitt trigger (hereinafter - TS) on an op-amp KR140UD608 (IC1) or on any other op-amp.

When the required voltage value on the battery is reached, at which the potentials at the direct and inverse inputs (pins 3, 2 - respectively) of IC1 are equal, a high logical level (almost equal to the input voltage) will appear at the output of the op-amp, causing the LED indicating the end of charging HL2 and the LED to light up optocoupler VH1 which will open its own transistor, blocking the supply of pulses to output U1. The key on VT2 will close and the battery will stop charging.

Once the battery is charged, it will begin to discharge through the reverse diode built into VT2, which will be directly connected in relation to the battery and the discharge current will be approximately 15-25 mA, taking into account the discharge also through the elements of the TS circuit. If this circumstance seems critical to someone, a powerful diode (preferably with a low forward voltage drop) should be placed in the gap between the drain and the negative terminal of the battery.

The TS hysteresis in this version of the charger is chosen such that the charge will begin again when the voltage on the battery drops to 3.9 V.

This charger can also be used to charge series-connected lithium (and other) batteries. It is enough to calibrate the required response threshold using variable resistor PR3.
So, for example, a charger assembled according to scheme 1 operates with a three-section serial battery from a laptop, consisting of dual elements, which was mounted to replace the nickel-cadmium battery of a screwdriver.
The power supply from the laptop (19V/4.7A) is connected to the charger, assembled in the standard case of the screwdriver charger instead of the original circuit. The charging current of the “new” battery is 2 A. At the same time, transistor VT2, working without a radiator, heats up to a maximum temperature of 40-42 C.
The charger is switched off, naturally, when the battery voltage reaches 12.3V.

The TS hysteresis when the response threshold changes remains the same as a PERCENTAGE. That is, if at a shutdown voltage of 4.1 V, the charger was turned on again when the voltage dropped to 3.9 V, then in this case the charger was turned on again when the voltage on the battery decreased to 11.7 V. But if necessary, the hysteresis depth can change.

Charger Threshold and Hysteresis Calibration

Calibration occurs using an external voltage regulator (laboratory power supply).
The upper threshold for triggering the TS is set.
1. Disconnect the upper pin PR3 from the charger circuit.
2. We connect the “minus” of the laboratory power supply (hereinafter referred to as the LBP everywhere) to the negative terminal for the battery (the battery itself should not be in the circuit during setup), the “plus” of the LBP to the positive terminal for the battery.
3. Turn on the charger and LBP and set the required voltage (12.3 V, for example).
4. If the end of charge indication is on, rotate the PR3 slider down (according to the diagram) until the indication goes out (HL2).
5. Slowly rotate the PR3 engine upward (according to the diagram) until the indication lights up.
6. Slowly reduce the voltage level at the output of the LBP and monitor the value at which the indication goes out again.
7. Check the level of operation of the upper threshold again. Fine. You can adjust the hysteresis if you are not satisfied with the voltage level that turns on the charger.
8. If the hysteresis is too deep (the charger is switched on at a too low voltage level - below, for example, the battery discharge level), turn the PR4 slider to the left (according to the diagram) or vice versa - if the hysteresis depth is insufficient, - to the right (according to the diagram). When changing depth of hysteresis, the threshold level may shift by a couple of tenths of a volt.
9. Make a test run, raising and lowering the voltage level at the LBP output.

Setting the current mode is even easier.
1. We turn off the threshold device using any available (but safe) methods: for example, by “connecting” the PR3 engine to the common wire of the device or by “shorting” the LED of the optocoupler.
2. Instead of the battery, we connect a load in the form of a 12-volt light bulb to the output of the charger (for example, I used a pair of 12V 20-watt lamps to set up).
3. We connect the ammeter to the break of any of the power wires at the input of the charger.
4. Set the PR1 engine to minimum (to the maximum left according to the diagram).
5. Turn on the memory. Smoothly rotate the PR1 adjustment knob in the direction of increasing current until the required value is obtained.
You can try to change the load resistance towards lower values of its resistance by connecting in parallel, say, another similar lamp or even “short-circuiting” the output of the charger. The current should not change significantly.

During testing of the device, it turned out that frequencies in the range of 100-700 Hz were optimal for this circuit, provided that IRF3205, IRF3710 were used (minimum heating). Since the TL494 is underutilized in this circuit, the free error amplifier on the IC can be used to drive a temperature sensor, for example.

It should also be borne in mind that if the layout is incorrect, even a correctly assembled pulse device will not work correctly. Therefore, one should not neglect the experience of assembling power pulse devices, described repeatedly in the literature, namely: all “power” connections of the same name should be located at the shortest distance relative to each other (ideally at one point). So, for example, connection points such as the collector VT1, the terminals of resistors R6, R10 (connection points with the common wire of the circuit), terminal 7 of U1 - should be combined almost at one point or through a straight short and wide conductor (bus). The same applies to drain VT2, the output of which should be “hung” directly onto the “-” terminal of the battery. The terminals of IC1 must also be in close “electrical” proximity to the battery terminals.

Memory circuit No. 2 (TL494)

Scheme 2 is not very different from Scheme 1, but if the previous version of the charger was designed to work with an AB screwdriver, then the charger in Scheme 2 was conceived as a universal, small-sized (without unnecessary adjustment elements), designed to work with composite, sequentially connected elements up to 3, and with singles.

As you can see, to quickly change the current mode and work with different numbers of elements connected in series, fixed settings have been introduced with trimming resistors PR1-PR3 (current setting), PR5-PR7 (setting the end of charging threshold for a different number of elements) and switches SA1 (current selection charging) and SA2 (selecting the number of battery cells to be charged).
The switches have two directions, where their second sections switch the mode selection indication LEDs.

Another difference from the previous device is the use of a second error amplifier TL494 as a threshold element (connected according to the TS circuit) that determines the end of battery charging.

Well, and, of course, a p-conductivity transistor was used as a key, which simplified the full use of the TL494 without the use of additional components.

The method for setting the end of charging thresholds and current modes is the same, as for setting up the previous version of the memory. Of course, for a different number of elements, the response threshold will change multiples.

When testing this circuit, we noticed stronger heating of the switch on the VT2 transistor (when prototyping I use transistors without a heatsink). For this reason, you should use another transistor (which I simply didn’t have) of appropriate conductivity, but with better current parameters and lower open-channel resistance, or double the number of transistors indicated in the circuit, connecting them in parallel with separate gate resistors.

The use of these transistors (in a “single” version) is not critical in most cases, but in this case, the placement of the device components is planned in a small-sized case using small radiators or no radiators at all.

Memory circuit No. 3 (TL494)

In the charger in diagram 3, automatic disconnection of the battery from the charger with switching to the load has been added. This is convenient for checking and studying unknown batteries. The TS hysteresis for working with a battery discharge should be increased to the lower threshold (for switching on the charger), equal to the full battery discharge (2.8-3.0 V).

Charger circuit No. 3a (TL494)

Scheme 3a is a variant of scheme 3.

Memory circuit No. 4 (TL494)

The charger in diagram 4 is no more complicated than the previous devices, but the difference from the previous schemes is that the battery here is charged with direct current, and the charger itself is a stabilized current and voltage regulator and can be used as a laboratory power supply module, classically built according to “datasheet” to the canons.

Such a module is always useful for bench tests of both batteries and other devices. It makes sense to use built-in devices (voltmeter, ammeter). Formulas for calculating storage and interference chokes are described in the literature. Let me just say that I used ready-made various chokes (with a range of specified inductances) during testing, experimenting with a PWM frequency from 20 to 90 kHz. I didn’t notice any particular difference in the operation of the regulator (in the range of output voltages 2-18 V and currents 0-4 A): minor changes in the heating of the key (without a radiator) suited me quite well. The efficiency, however, is higher when using smaller inductances.
The regulator worked best with two series-connected 22 µH chokes in square armored cores from converters integrated into laptop motherboards.

Memory circuit No. 5 (MC34063)

In diagram 5, a version of the PWM controller with current and voltage regulation is made on the MC34063 PWM/PWM chip with an “add-on” on the CA3130 op amp (other op amps can be used), with the help of which the current is regulated and stabilized.
This modification somewhat expanded the capabilities of the MC34063, in contrast to the classic inclusion of the microcircuit, allowing the function of smooth current control to be implemented.

Memory circuit No. 6 (UC3843)

In diagram 6, a version of the PHI controller is made on the UC3843 (U1) chip, CA3130 op-amp (IC1), and LTV817 optocoupler. The current regulation in this version of the charger is carried out using a variable resistor PR1 at the input of the current amplifier of the U1 microcircuit, the output voltage is regulated using PR2 at the inverting input IC1.
There is a “reverse” reference voltage at the “direct” input of the op-amp. That is, regulation is carried out relative to the “+” power supply.

In schemes 5 and 6, the same sets of components (including chokes) were used in the experiments. According to the test results, all of the listed circuits are not much inferior to each other in the declared range of parameters (frequency/current/voltage). Therefore, a circuit with fewer components is preferable for repetition.

Memory circuit No. 7 (TL494)

The memory in diagram 7 was conceived as a bench device with maximum functionality, therefore there were no restrictions on the volume of the circuit and the number of adjustments. This version of the charger is also made on the basis of a PHI current and voltage regulator, like the option in diagram 4.
Additional modes have been introduced into the scheme.
1. “Calibration - charge” - for pre-setting the end voltage thresholds and repeating charging from an additional analog regulator.
2. “Reset” - to reset the charger to charge mode.
3. “Current - buffer” - to switch the regulator to current or buffer (limiting the output voltage of the regulator in the joint supply of the device with battery voltage and the regulator) charge mode.

A relay is used to switch the battery from the “charge” mode to the “load” mode.

Working with the memory is similar to working with previous devices. Calibration is carried out by switching the toggle switch to the “calibration” mode. In this case, the contact of the toggle switch S1 connects the threshold device and a voltmeter to the output of the integral regulator IC2. Having set the required voltage for the upcoming charging of a specific battery at the output of IC2, using PR3 (smoothly rotating) the HL2 LED lights up and, accordingly, relay K1 operates. By reducing the voltage at the output of IC2, HL2 is suppressed. In both cases, control is carried out by a built-in voltmeter. After setting the PU response parameters, the toggle switch is switched to charge mode.

Scheme No. 8

The use of a calibration voltage source can be avoided by using the memory itself for calibration. In this case, you should decouple the TS output from the SHI controller, preventing it from turning off when the battery charge is complete, determined by the TS parameters. The battery will one way or another be disconnected from the charger by the contacts of relay K1. The changes for this case are shown in Figure 8.

In calibration mode, toggle switch S1 disconnects the relay from the positive power supply to prevent inappropriate operations. In this case, the indication of the operation of the TC works.
Toggle switch S2 performs (if necessary) forced activation of relay K1 (only when calibration mode is disabled). Contact K1.2 is necessary to change the polarity of the ammeter when switching the battery to the load.
Thus, a unipolar ammeter will also monitor the load current. If you have a bipolar device, this contact can be eliminated.

Charger design

In designs it is desirable to use as variable and tuning resistors multi-turn potentiometers to avoid suffering when setting the necessary parameters.

Design options are shown in the photo. The circuits were soldered impromptu onto perforated breadboards. All the filling is mounted in cases from laptop power supplies.
They were used in designs (they were also used as ammeters after minor modifications).
The cases are equipped with sockets for external connection of batteries, loads, and a jack for connecting an external power supply (from a laptop).

Over 18 years of work at North-West Telecom, I have made many different stands for testing various equipment being repaired.
He designed several digital pulse duration meters, different in functionality and elemental base.

More than 30 improvement proposals for the modernization of units of various specialized equipment, incl. - power supply. For a long time now I have been increasingly involved in power automation and electronics.

Why am I here? Yes, because everyone here is the same as me. There is a lot of interest here for me, since I am not strong in audio technology, but I would like to have more experience in this area.

Reader vote

The article was approved by 77 readers.

To participate in the voting, register and log in to the site with your username and password.

Another charger is assembled according to the circuit of a key current stabilizer with a unit for monitoring the achieved voltage on the battery to ensure that it is turned off at the end of charging. To control the key transistor, a widely used specialized microcircuit TL494 (KIA491, K1114UE4) is used. The device provides regulation of the charge current within 1 ... 6 A (10 A max) and output voltage 2 ... 20 V.

The key transistor VT1, diode VD5 and power diodes VD1 - VD4 through mica spacers must be installed on a common radiator with an area of 200 ... 400 cm2. The most important element in the circuit is inductor L1. The efficiency of the circuit depends on the quality of its manufacture. As a core, you can use a pulse transformer from a 3USTST TV power supply or similar. It is very important that the magnetic core has a slot gap of approximately 0.5 ... 1.5 mm to prevent saturation at high currents. The number of turns depends on the specific magnetic circuit and can be in the range of 15 ... 100 turns of PEV-2 2.0 mm wire. If the number of turns is excessive, then a soft whistling sound will be heard when the circuit operates at rated load. As a rule, the whistling sound occurs only at medium currents, and with a heavy load, the inductance of the inductor due to the magnetization of the core drops and the whistling stops. If the whistling sound stops at low currents and with a further increase in the load current, the output transistor begins to heat up sharply, then the area of the magnetic core is insufficient to operate at the selected generation frequency - it is necessary to increase the operating frequency of the microcircuit by selecting resistor R4 or capacitor C3 or install a larger inductor. If there is no power transistor of the p-n-p structure in the circuit, you can use powerful transistors of the n-p-n structure, as shown in the figure.

As a diode VD5 in front of inductor L1, it is advisable to use any available diodes with a Schottky barrier, rated for a current of at least 10A and a voltage of 50V; in extreme cases, you can use mid-frequency diodes KD213, KD2997 or similar imported ones. For the rectifier, you can use any powerful diodes with a current of 10A or a diode bridge, for example KBPC3506, MP3508 or the like. It is advisable to adjust the shunt resistance in the circuit to the required value. The range of adjustment of the output current depends on the ratio of the resistances of the resistors in the output circuit 15 of the microcircuit. In the lower position of the current adjustment variable resistor slider in the diagram, the voltage at pin 15 of the microcircuit must coincide with the voltage on the shunt when the maximum current flows through it. The variable current control resistor R3 can be set with any nominal resistance, but you will need to select a fixed resistor R2 adjacent to it to obtain the required voltage at pin 15 of the microcircuit.
The variable output voltage adjustment resistor R9 can also have a wide range of nominal resistance 2 ... 100 kOhm. By selecting the resistance of resistor R10, the upper limit of the output voltage is set. The lower limit is determined by the ratio of the resistances of resistors R6 and R7, but it is undesirable to set it less than 1 V.

The microcircuit is installed on a small printed circuit board 45 x 40 mm, the remaining elements of the circuit are installed on the base of the device and the radiator.

The wiring diagram for connecting the printed circuit board is shown in the figure below.

PCB options in lay6

We say thank you for the seals in the comments Demo

Source: http://shemotechnik.ru

So. We have already looked at the half-bridge inverter control board; it’s time to put it into practice. Let's take a typical half-bridge circuit; it does not cause any particular difficulties in assembly. The transistors are connected to the corresponding terminals of the board, a standby power supply of 12-18 volts is supplied. If 3 diodes are connected in series, the voltage at the gates will drop by 2 volts and we will get exactly the required 10-15 volts.

Let's look at the diagram:
The transformer is calculated by the program or simplified using the formula N=U/(4*pi*F*B*S). U=155V, F=100000 hertz with RC ratings of 1nf and 4.7kOhm, B=0.22 T for the average ferrite, regardless of permeability, the only variable parameter that remains is S - the cross-sectional area of the side of the ring or the middle rod Ш of the magnetic circuit in square meters.

The throttle is calculated using the formula L=(Uppeak-Ustab)*Тdead/Imin. However, the formula is not very convenient - the dead time depends on the difference between the peak and stabilized voltage. The stabilized voltage is the arithmetic mean of the sample from the output pulses (not to be confused with the root mean square). For a power supply that is regulated over the full range, the formula can be rewritten as L= (Upeak*1/(2*F))/Imin. It can be seen that, in the case of complete voltage regulation, the more inductance is needed, the lower the minimum current value. What will happen if the power supply is loaded with less than a current Imin. And everything is very simple - the voltage will tend to the peak value, it seems to ignore the inductor. In the case of feedback regulation, the voltage will not be able to rise; instead, the pulses will be suppressed so that only their fronts remain, stabilization will occur due to heating of the transistors, essentially a linear stabilizer. I think it is correct to take Imin such that the linear mode losses are equal to the losses at maximum load. Thus, the adjustment remains in full range and is not dangerous for the power supply.

The output rectifier is built according to a full-wave circuit with midpoint. This approach allows you to halve the voltage drop across the rectifier and allows you to use ready-made diode assemblies with a common cathode, which are no more expensive than a single diode, for example MBR20100CT or 30CTQ100. The first digits of the marking mean a current of 20 and 30 amperes, respectively, and the second digits mean a voltage of 100 volts. It is worth considering that the diodes will have double voltage. Those. we get 12 volts at the output, and at the same time there will be 24 on the diodes.

Half-bridge transistors.. And here it’s worth thinking about what we need. Relatively low-power transistors like IRF730 or IRF740 can operate at very high frequencies, 100 kilohertz is not the limit for them, and besides, we do not risk a control circuit built on not very powerful parts. For comparison, the gate capacitance of the 740 transistor is only 1.8 nf, and the IRFP460 is as much as 10 nf, which means 6 times more power will be used to transfer the capacitance each half-cycle. Plus, this will tighten the fronts. For static losses, you can write P=0.5*Ropen *Itr^2 for each transistor. In words - the resistance of an open transistor multiplied by the square of the current through it, divided by two. And these losses are usually several watts. Another thing is dynamic losses, these are losses on the fronts, when the transistor passes through the hated mode A, and this evil mode causes losses, roughly described as the maximum power multiplied by the ratio of the duration of both fronts to the duration of the half-cycle, divided by 2. For each transistor. And these losses are much more than static. Therefore, if you take a more powerful transistor when
you can get by with an easier option, you can even lose in efficiency, so don’t overuse it.

Looking at the input and output capacitances, you may want to make them too large, and this is quite logical, because despite the operating frequency of the power supply of 100 kilohertz, we still rectify the mains voltage of 50 hertz, and in case of insufficient capacitance we will get the same output rectified sine wave, it is remarkably modulated and demodulated back. So you should look for pulsations at a frequency of 100 hertz. For those who are afraid of “HF noise”, I assure you that there is not a drop of it, it was checked with an oscilloscope. But increasing the capacitances can lead to huge inrush currents, and they will certainly cause damage to the input bridge, and inflated output capacitances will also lead to an explosion of the entire circuit. To correct the situation, I made some additions to the circuit - a relay for monitoring the charge of the input capacitance and a soft start on the same relay and capacitor C5. I’m not responsible for the ratings, I can only say that C5 will be charged through resistor R7, and the charging time can be estimated using the formula T=2pRC, the output capacitance will be charged at the same speed, charging with a stable current is described by U=I*t/C, although not exactly, but it is possible to estimate the current surge depending on time. By the way, without a throttle it makes no sense.

Let's look at what came out after modification:

Let's imagine that the power supply is heavily loaded and at the same time turned off. We turn it on, but the capacitors do not charge, the charging resistor just lights up and that’s it. It's a problem, but there is a solution. The second contact group of the relay is normally closed, and if the 4th input of the microcircuit is closed with a built-in 5 volt stabilizer on the 14th leg, then the pulse duration will drop to zero. The microcircuit will be turned off, the power switches will be locked, the input capacitance will be charged, the switch will click, capacitor C5 will begin to charge, the pulse width will slowly rise to the operating level, the power supply will be completely ready for operation. If the voltage in the network decreases, the relay will turn off, this will lead to the control circuit being turned off. When the voltage is restored, the starting process will be repeated again. It seems like I did it correctly, if I missed something, I will be glad for any comments.

Current stabilization here plays more of a protective role, although adjustment is possible with a variable resistor. It was implemented through a current transformer, because it was adapted to a power supply with a bipolar output, but it’s not all that simple. The calculation of this transformer is very simple - a shunt with a resistance of R Ohm is transferred to the secondary winding with the number of turns N as resistance Rнт=R*N^2, you can express the voltage from the ratio of the number of turns and the drop on the equivalent shunt, it should be greater than the drop voltage diode. The current stabilization mode will begin when the voltage at the + input of the op-amp tries to exceed the voltage at the - input. Based on this calculation. The primary winding is a wire pulled through a ring. It is worth considering that a break in the load of a current transformer can lead to the appearance of huge voltages at its output, at least sufficient to break down the error amplifier.

Capacitors C4 C6 and resistors R10 R3 form a differential amplifier. Due to the chain R10 C6 and the mirrored R3 C4, we obtain a triangular decline in the amplitude-frequency response of the error amplifier. This looks like a slow change in pulse width depending on the current. On the one hand, this reduces the feedback speed, on the other hand, it makes the system stable. The main thing here is to ensure that the frequency response goes below 0 decibels at a frequency of no more than 1/5 of the switching frequency; such feedback is quite fast, in contrast to feedback from the output of the LC filter. The start frequency of the cutoff at -3dB is calculated as F=1/2pRC where R=R10=R3; C=C6=C4, I’m not responsible for the values in the diagram, I didn’t count them. Own gain

The circuit is calculated as the ratio of the maximum possible voltage (dead time tends to zero) on capacitor C4 to the voltage of the saw generator built into the chip and converted to decibels. It raises the frequency response of the closed system upward. Considering that our compensating chains give a decline of 20 dB per decade starting from a frequency of 1/2pRC and knowing this rise, it is not difficult to find the point of intersection with 0 dB, which should be no more than at a frequency of 1/5 of the operating frequency, i.e. 20 kilohertz. It is worth noting that the transformer should not be wound with a huge power reserve, on the contrary, the short-circuit current should not be particularly large, otherwise even such a high-frequency protection will not be able to work on time, and what if a kiloampere jumps out there .. So we don’t abuse this either .

That's all for today, I hope the diagram will be useful. It can be adapted for a power screwdriver, or a bipolar output can be made to power an amplifier; it is also possible to charge batteries with a stable current. For complete wiring of the tl494 we refer to the last part; the only additions to it are the soft start capacitor C5 and the relay contacts on it. Well, an important note - monitoring the voltage on the half-bridge capacitors forced us to connect the control circuit with power in such a way that this will not allow the use of standby power with a quenching capacitor, at least with bridge rectification. A possible solution is a half-wave rectifier such as a diode half-bridge or a transformer in the duty room.

ID: 1548

What do you think of this article?

TL494 in a full power supply

More than a year has passed since I seriously took up the topic of power supplies. I read the wonderful books “Power Supplies” by Marty Brown and “Power Electronics” by Semenov. As a result, I noticed a lot of errors in circuits from the Internet, and lately all I see is a cruel mockery of my favorite TL494 microcircuit.

I love the TL494 for its versatility; there is probably no power supply that cannot be implemented on it. In this case, I want to look at the implementation of the most interesting half-bridge topology. The control of the half-bridge transistors is done galvanically isolated, this requires a lot of elements, in principle a converter inside a converter. Despite the fact that there are many half-bridge drivers, it is still too early to write off the use of a transformer (GDT) as a driver; this method is the most reliable. Bootstrap drivers exploded, but I have not yet seen a GDT explosion. The driver transformer is a regular pulse transformer, calculated using the same formulas as the power transformer, taking into account the drive circuit. Often I have seen the use of high power transistors in GDT drives. The outputs of the microcircuit can produce 200 milliamps of current, and in the case of a well-designed driver, this is a lot; I personally drove the IRF740 and even the IRFP460 at a frequency of 100 kilohertz. Let's look at the diagram of this driver:

T
This circuit is connected to each output winding of the GDT. The fact is that at the moment of dead time, the primary winding of the transformer is open-circuited, and the secondary windings are not loaded, so the discharge of the gates through the winding itself will take an extremely long time, the introduction of a supporting, discharge resistor will prevent the gate from quickly charging and waste a lot of energy. The diagram in the figure is free from these shortcomings. The edges measured on a real prototype were 160ns rising and 120ns falling on the gate of the IRF740 transistor.

The transistors complementing the bridge in the GDT drive are constructed similarly. The use of bridge swing is due to the fact that before the tl494 power trigger operates upon reaching 7 volts, the output transistors of the microcircuit will be open; if the transformer is turned on as a push-pull, a short circuit will occur. The bridge is working stably.

The VD6 diode bridge rectifies the voltage from the primary winding and if it exceeds the supply voltage, it returns it back to capacitor C2. This happens due to the appearance of reverse voltage; after all, the inductance of the transformer is not infinite.

The circuit can be powered through a quenching capacitor; now a 400 volt K73-17 at 1.6 uF is working. diodes KD522 or much better 1n4148, replacement with more powerful 1n4007 is possible. The input bridge can be built on 1n4007 or use a ready-made kts407. On the board, Kts407 was mistakenly used as VD6, it should not be placed there under any circumstances, this bridge must be made on RF diodes. Transistor VT4 can dissipate up to 2 watts of heat, but it plays a purely protective role; you can use KT814. The remaining transistors are KT361, and replacement with low-frequency KT814 is highly undesirable. The tl494 master oscillator is configured here at a frequency of 200 kilohertz, which means that in push-pull mode we get 100 kilohertz. We wind the GDT on a ferrite ring 1-2 centimeters in diameter. Wire 0.2-0.3mm. There should be ten times more turns than the calculated value, this greatly improves the shape of the output signal. The more it is wound, the less you need to load the GDT with resistor R2. I wound 3 windings of 70 turns on a ring with an outer diameter of 18mm. The overestimation of the number of turns and the mandatory loading are associated with the triangular component of the current; it decreases with an increase in turns, and loading simply reduces its percentage influence. The printed circuit board is included, but it does not exactly correspond to the diagram, but the main blocks are there, plus a body kit for one error amplifier and a series stabilizer for power supply from a transformer have been added. The board is made for installation into the section of the power section board.

The microcircuit is installed on a small printed circuit board 45 x 40 mm, the remaining elements of the circuit are installed on the base of the device and the radiator.

The wiring diagram for connecting the printed circuit board is shown in the figure below.

PCB options in lay6

We say thank you for the seals in the comments Demo

Source: http://shemotechnik.ru