Switching power supply for TDA7294 on IR2153
One day I had to build an amplifier for TDA7294. And it was necessary to build it as soon as possible. I had a birthday coming up, and it was supposed to celebrate it outdoors, listening to the sounds of my rare Radiotech S30 speakers.
The amplifier was assembled immediately. For anyone interested, read the article “TDA7294 Bass Amplifier”. It was time to assemble the switching power supply. It was extremely important to keep the power supply small.
So I chose the simplest switching power supply circuit on the ir2153.
I found many similar circuits on the internet with slight differences. Not all of the circuits on the net work. It took me a while to figure that out. The circuit I showed you is fully working. If you follow the nominal values of this circuit and use my printed circuit board you can save time in correcting your own and other people’s mistakes.
A more complex analog of this circuit is described in the article “The Impulse Power Supply for an ir2153 300W LF Amplifier”. This circuit has an overload protector and a soft start.
The simplicity of the circuit for TDA7294 on ir2153 allows beginners to repeat it with ease. Another plus is the dimensions. The dimensions of the switching power supply board are 80 mm wide and 80 mm high.
The circuit working principle. You can read about how the ir2153 switching power supply works in the article “The switching power supply for an ir2153 LF amplifier with 300W power”.
The schematic diagram does not show a varistor, but the printed circuit board does. Basically it can be put down, because it plays almost no role, it serves as a protection against power surges (no need to solder any jumpers, just do not put varistor and that’s all).
NTC thermistor at first turn on limits current surge, when you charge the mains and output electrolytes, after some time it heats up and its resistance decreases. Simple, but not really reliable protection. When you turn it on again, when the thermistor is hot, the protection is no longer effective. But as practice has shown, the power supply is reliable and does not fail, as some people write in the comments.
The timing elements R2 and C3 were chosen so that the driver would provide pulse generation with a frequency of about 70 kHz. You can find a program for calculating R2 and C3 under the article, you can calculate for the frequency you want.
Elements.
| DESCRIPTION | TYPE | NUMBER | NUMBER | COMMENT |
| Power Driver | IR2153 | 1 | ||
| VT1,VT2 | MOSFET transistor | IRF740 | 2 | |
| VDS1 | Diode bridge | RS607 | 1 | 6A 1000V |
| VDR1 | Varistor | MYG14-431 | 1 | Can be omitted |
| NTC | Thermistor | 5D-9 | 1 | Or other 5Ohm resistor |
| R1 | 2W resistor | 18kOhm | 1 | |
| R2 | 0.25W resistor | HER108 | 10kOhm | |
| R3,R4 | 0.25W resistor | 33 Ohm | 2 | |
| C1,C2 | Electrolyte | 220μF 220V | 2 | |
| C3 | Non-polar capacitor | 1nF | 1 | Ceramics any voltage |
| C4 | Non-polar capacitor | 0,1 uF | 1 | Ceramics any voltage |
| C5 | Electrolyte | 220μF 16V | 1 | |
| C6 | Non-polar capacitor | 0.33 uF | 1 | Ceramics any voltage |
| C7 | Non-polar capacitor | 1µF 400V | 1 | Film |
| C8-C9 | Electrolyte | 470 uF 50V | 2 | |
| C10-C11 | Non-polar capacitor | 0,1 uF | 2 | Film |
| VD1 | Diode | HER108 | 1 | |
| VD2 | Impulse diode | FR107,FR157 | 1 | Any other pulsed |
| VD3-VD6 | Schottky diode | KD213A | 4 |
Components list in PDF format DOWNLOAD
Transformer. The most difficult part of the assembly is the calculation and the winding of the pulse transformer. I will not tell in detail about the technology of calculation and winding of the transformer, as I have already told you before, read the article “Calculation and winding of the pulse transformer”. I also recommend to read the article “How to rewind a transformer from a PC power supply”.
At this point I will share a little bit of my experience. The article, the link to which is located above, describes a method of winding the secondary with a tap from the middle, double wire (if the calculation of the secondary has one strand) and then connect them to the midpoint. This gives synchronism, i.e. both shoulders will have the same voltage. The secondary winding of the transformer for this device should have two strands of 0.85 mm diameter, to provide the power we need (according to my calculations, you can have one strand).
So, if you were to use the method from the article above, you would have to wind with 4 wires at once, which is very inconvenient.
I decided to use two wires, i.e. first I wound one shoulder with two wires, then the insulation, and then the second shoulder with two wires.
They advised not to do it this way, because of not synchronous winding will be different voltage. I had the same voltage, and it was easier to wind, because the bagel is small.
Below I will give some winding data.
The wire diameter of both the primary and secondary windings is 0.85 mm. The magnet wire is glued from two rings of 28mm*16mm*9mm and magnetic permeability 2000NM.
The primary winding has 39 turns, although according to the calculations it was forty and a penny, but they did not fit. As a consequence, I had to reduce the number of turns of the secondary winding, relative to the calculations.
So, the secondary winding has 8 + 8 turns. This means 8 turns, then the tap (this will be the middle point), insulation, then another 8 turns.
The secondary winding is wound with two strands of 0.85 mm diameter.
(8 turns of the secondary winding)
(connect the end of the 8th loop to the wire to make a tap and wind another 8 turns in the same direction).
Insulation take to taste (rag tape, duct tape or FUM tape, lavsan film or tape). I use lavsan film from the scraps of twisted pair.
All windings should be wound in the same direction you choose.
Cooling.
My key heater is the front panel of my amplifier. It is made of dural, 47mm high, 92mm wide, 7mm thick. When testing and further operating one channel of TDA7294, the keys are warm, not hot.
The keys are mounted on the heat sink through silicone gaskets and dielectric bushings.
The Schottky’s are without heat sinks. They are not very hot, again when operating one channel, the transformer is not hot.
The assembly of this circuit on a transformer from a personal computer power supply is described in the article “The simplest bipolar power supply”.
List of components for an IR2153-based power supply DOWNLOAD
Printed circuit board of IR2153 from DOWNLOAD
Calculation calculator for IR2153 time suppressor elements DOWNLOAD
Car voltage converter on IR2153
| Current time: Thu Sep 15, 2022 18:28:54 |
Time Zone: UTC + 3 hours
Half-bridge on IR2153, help achieve stability.
| Page 10 of 10 | [ Messages: 197 ] | To page Pre. 1 . 6 , 7 , 8 , 9 , 10 |
|
Colleagues, why do you propose to entrust the output voltage stabilization system of the power supply to compensate for voltage sags on the line rectifier bridge? It seems to me that the task of compensating the mains voltage ripple should be solved by the electrolytic capacitors on the input and output of the power supply. How do those hams live, who built their IPI on IR2153 without voltage stabilization? Are they tormented by the AC background? _________________ Regards, VictorS. One of the first tasks of any stabilizer is to cut the ripple voltage. A capacitor can only reduce them. So how do hams live who built their IPI on IR2153 without voltage regulator? Are they bothered by the AC background? There are devices which are not sensitive to ripple, such as power amplifiers of audio frequency, heaters. But they usually do not need stabilization. If the regulator will not cut ripple, it usually will not stabilize the voltage. _________________ If you are dissatisfied with your standard of living, the laws of our country, the level of prices, then remember all this at the next election. And why do you propose to put on the system of stabilization of the output voltage of the power supply, to compensate the voltage sag on the mains rectifier bridge? It seems to me that the task of compensating the mains voltage ripple should be solved by the electrolytic capacitors on the input and output of the power supply. This is a misconception. My favorite picture already. The voltages are slightly different, but the point, I hope, is clear. In order to avoid this, you need to install capacitors of infinitely high capacitance. But in this case, the amplitude of current pulses through diodes also becomes infinitely large, and the PowerFactor, on the contrary, drops to the level of the plinth. So it is the primary task of the stabilizer to parry these pulsations. It does not matter at all whether it is pulsed, linear or whatever. PCB assembly from $30 + FREE shipping worldwide + stencil Thanks so much for the detailed explanation. I take the fact that the voltage ripple suppression under load, by a stabilizer, for granted. I thought, that we are speaking about some special circuit tricks, not to put capacitors at all, or minimum capacitance. _________________ Regards, VictorS. High quality at a competitive price allows MORNSUN DC/DC converters to compete with counterparts from the world’s leading manufacturers. Products of this brand, such as the UWTH1D family, can be successfully used in railway applications. For telecommunications equipment are suitable DC/DC-converters of VCB and VCF families, for distributed power supply systems – K78 family small-sized pulsed PoL-stabilizers, and for automated production systems and robotics – KUB series is indispensable. There are also unique solutions, such as miniature DC/DC converter B0505ST16-W5 in chip form, designed for medical devices. _________________ Z Wisdom(Experience and endurance) comes with the years. All your troubles and problems are from lack of knowledge. A wise man can learn from a fool, but a fool can learn from … Albert Einstein will not help and the GDP will not be saved. High quality at a competitive price allows MORNSUN DC/DC converters to compete with counterparts from the world’s leading manufacturers. Products of this brand, such as the UWTH1D family, can be successfully used in railway applications. For telecommunications equipment are suitable DC/DC-converters of VCB and VCF families, for distributed power supply systems – K78 family small-sized pulsed PoL-stabilizers, and for automated production systems and robotics – KUB series is indispensable. There are also unique solutions, such as miniature DC/DC converter B0505ST16-W5 in chip form, designed for medical devices. But on this chip you can stabilize it using resonant topology. correspond to the operation of a resonant converter. and without using resonance, you should FORGET about voltage stabilization on the IR2153 chip. _________________ Wisdom comes with impotence. _________________ If you are dissatisfied with your standard of living, the laws of our country, the level of prices, then remember all this at the next election. Thank you for your attention to the discussion, in the wake of my question. The work of quasiresonant converter was described in Radio magazine №2 of 1996, and there, it was suggested to introduce LC-chain between the connection point of key transistors and the primary of the power transformer, and at constant frequency conversion, the LC-chain parameters were simply calculated. In the disassembled converter on the IR2153, it is proposed to change the frequency in a wide range – up to three times, for example. Now it is not clear – at what frequency to calculate the resonant LC circuit – on the average – corresponding to the optimal parameters of input and output voltage, or the resonant LC circuit is calculated by other input data? Thank you in advance for your informative answer. _________________ Regards, VictorS. Why do you have to reinvent the lisapedad? You know L6598 or IRS27951 are a bit more expensive than 2153, tuned to work with resonant controller, have protection, soft start, OS on the usual optocoupler, etc. And for IRS the calculation methodology is described in detail by the manufacturer. VictorS, find and download the an-1160 upnote. there is all the theory and calculation of the classical resonant converter. and there you will find the answers to your questions. _________________ Wisdom comes with impotence. Thank you very much for the tip on AN-1160 – you can’t read too much. As I see it, the IR2153 under discussion is very similar to the IRS2795 from upnote, so we can try the same approach. For those interested, here is a link to the AN-1160: Design of Resonant Half-Bridge converter using IRS2795(1,2) Control IC http://files.domcxem.ru/infocenter/%D0% . N-1160.pdf _________________ Regards, VictorS. Hi all, Continuing my searches on the subject of resonant, stabilized IIP on IR2153, led to such a scheme:
With the resonant link from Lr – C6, everything is clear. But I don’t know why they put a knot on the BUX85 transistor and diode bridge on the left side of the circuit. It looks like some kind of softstart, intricate, working in reverse – firstly current pulse, through discharged capacitor 0,22 uF, and then current lowering, through resistor 100 ohm, but then why the choke 4,7 mG is put. Or this node works differently? Who can clearly explain? Thank you in advance. _________________ Regards, VictorS. For resonators there are a lot of microcircuits in the same package specially sharpened for this purpose. Why do you need 2153 in it? _________________ If you are dissatisfied with your standard of living, the laws of our country, the level of prices, then remember all this at the next election. Dear Telecott, I have 10 pieces of IR2153, a bunch of IRF740 and a lot of spare parts from computer power supplies. And it seems to me that engineering is about being able to make what you need from what is available. And if the experiments to build a stabilized IC on the IR2153 are unsuccessful, you will have to build the right one on the TL494. I don’t need any specialized resonant ICs yet. _________________ Regards, VictorS. I tested just such a frequency changer about 2-3 years ago, it works through the 5th point. The frequency is of course regulated, but the symmetry of the pulses is lost, especially when changing. I always thought it was necessary to choose or design the circuit correctly. You can of course use pliers to hammer nails and some people can do it, but I prefer a hammer. If you want a stabilized 2153 block you have to make a standard circuit and stabilize the secondary side with a step down regulator or 3.3V magnetic amplifiers in comp units with only 2 chokes. _________________ If you are dissatisfied with your standard of living, the laws of our country, the level of prices, then remember all this at the next election. Last edited by Telecott Tue Jan 23, 2018 12:47:31 pm, edited 1 time in total. An aspiring composer asked Mozart how to write symphonies. – You are still young, you should start with ballads,” replied the composer. – Excuse me, but you began to write symphonies before you were ten years old. – Yes, but I never asked anyone how they should write them. What I’m getting at is that every problem has its own tool to solve it. The IR2153 is no exception, and it solves this problem quite well. But you want more from it. And other comrades tell you that the tool is not quite suitable, but there is another, much better one which does “more”. But you disagree, and want to get something workable with your version of the tool from the comrades telling you off. Don’t you find that somewhat contradictory? On the other hand, I can in no way discourage you from inventing something, but only if you do the invention yourself, and the process and, if they do, share the results with your objectors, so they can be sure that you really have it working. Of course, if you want to. Point one: to build a resonant power supply you need to learn how to control the frequency of the master oscillator in a linear and predictable manner. Your move: research the IR2153 and document ways, dependencies. Hi all, Continuing my searches on the subject of resonant, stabilized IIP on IR2153, led to such a scheme:
But I don’t know why I made the BUX85 junction and diode bridge on the left side of the circuit. It looks like some kind of softstart, intricate, working in reverse – Looks like some kind of prototype, trying to find the source code, but no. Judging by the fact that vcc goes into the socket, the circuit is not collected. Generally need a PSU for soldering irons at 24V about 150W, 5V and with the Krekin can be taken. But considering that there are no stabilization schemes, it may be easier to make it so that at 250v in the socket, the output was not much more than 25v. Although I built a transformer one and there it’s like an extra voltage isn’t a problem, and there’s no stabilization there either, so maybe I shouldn’t bother. Time Zone: UTC + 3 hours Who’s on the forum nowThis forum is currently viewed by: no registered users or guests: 32 Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group Russian Support phpBB Extended by Karma MOD © 2007-2012 m157y Extended by Topic Tags MOD © 2012 m157y Site tested in browsers: IE8.0, Opera 9.0, Netscape Navigator 7.0, Mozilla Firefox 5.0 Adapted to work with screen resolutions from 1280×1024 and higher. At lower resolutions horizontal scrolling is possible. If you have any questions, please contact Cat: kot@radiokot.ru ©2005-2022 Power Car Inverter Transformer In this article I want to tell about the winding
Three Schematics for 12 volt – 220 volt (100 watt) pulse inverters Sometimes, when
How to Make 12-220V Inverters To connect household appliances to the onboard electrical system
Assembling a home-made converter from 12V to 220V To connect an electrical appliance to
|
