Сравним питание предов от сети и DC-DC конверторов
Hello, hello audio titans. Today's issue I want to dedicate to some issues of power supply of radio-electronic equipment, especially assembled by own hands. These studies of mine do not pull on a scientific approach, it is absolutely garage type, but absolutely practical research, which will certainly seem useful to you in case if you are going to assemble something from audio devices and power it accordingly from some power source, presumably from the network of course. I just faced this problem myself and as a perfectionist, I wanted to do better than possible right away. As best as possible, to put it simply. Therefore, I bought these devices like a fool of candy wrappers, I have them now. And compared them with each other, which I want
to show you. Here are actually the results of these studies, but first things first . Powering an audio device must be done carefully, having thought out the concept. Because some devices require specially filtered power. Many people power low-current devices from batteries ... This is also normal, of course, but impractical in my opinion. Of course, you need to power it from the network, with network
voltage, while many audio devices, most devices, require a reduced supply voltage. That is, some kind of transformer, in a traditional rectifier and stabilizer circuit as an option. I will show all this using the example of a device that is most demanding of supply voltage pulsations. This is a preamplifier for vinyl..
This is Sukhov's preamplifier, Nikolai Sukhov, about whom I have already made many issues and you probably know them. This is what these amplifiers actually look like. They are assembled using smd technology. Small, tiny, but at the same time they have a huge gain factor and, accordingly, supply pulsations, if they appear, that is, in any case, they will appear ...
These supply pulsations will negatively affect the results of the work, especially if you are chasing some premiere parameters, which, for example, this preamplifier is capable of ... This is the signal-to-noise ratio, accordingly, the dynamic range, and so on ... All this is important to consider when building a power supply. Traditionally, such a device, or rather such a board, is powered, it requires power supply, say, with a bipolar voltage of plus or minus 18 volts, as Nikolay insists, at a current of approximately 20 mA per arm, maybe 25. This is a small current, but not tiny. Practice has shown that not even all devices have coped with this task. But
of course, you need to take into account the pulse surge when turning on the power. Many circuits require unipolar power supply. It is somewhat simpler there, but the ordinal complexity does not change. So what are the traditional solutions to this problem? By the way, many say: "Here, sell me, or assemble such a preamplifier for me" and if I fulfill their request somehow, they are happy, not realizing that this is only half the work. You also need to power it correctly. In order to get these parameters ...
Looking ahead, I will say that of course the preamplifier must be shielded to get even better parameters that are achievable, for example, in modeling, in emulator programs ... That is, you need to take a full range of measures. Today I will talk only about devices that supply voltage to these devices and that you will probably use, in case of building your own acoustic system yourself... If you are a handy, smart, normal person from a past life, who has an education, some money and an understanding of what all this is for and how best to do it. Yesterday I was amazed by one citizen. He asked me on the channel in the comments where to buy
such a Sukhov amplifier, a high-fidelity final one - for a car. Honestly, he crushed me with this question. Why do you need one like this for a car? I don't understand. In short, there are some absurd things and in order to avoid one of these absurdities of power supply - a crooked power supply... for a basically good
device, I started this whole issue today. So, bipolar power supply is needed plus or minus 18 V. Classic circuit: step-down transformer, rectifier and some kind of stabilizer... You can do without a stabilizer, too, but the parameters will be a couple of decibels worse. Well, in the pulse there by 10 dB...
the parameter may be worse, but it will still work. In general, you can't tell the difference just by ear. It's a shame, right? But we'll still talk about how it should be. So transformer. The step-down transformer that I used. I used a toroidal transformer wound by the Tornado company, the good Sharshov makes them. I did everything here to show it to you closer. This is what this transformer looks like. These are the parameters it provides. That is, the primary winding is 230 volts, with this current, and the secondary is two at
21 V and it can withstand 240 mA current... temperature... 10 VA, this is the transformer. It also has a cap to attach it to. That's how it is. Next, you need to straighten it all out. For this, a single-board solution is used - a rectifier and stabilizer are installed here at once... You see. Such a simple circuit costs, I think, 3 euros somewhere on AliExpress. I bought a lot of them and have already used them. Reliable, good, it doesn't grab stars from the sky in terms of pulsations, but in principle
it's a completely normal working circuit. Here is a slightly more complicated circuit with advanced converters, with advanced stabilizers. This requires a rectifier board with smoothing capacitors, of course, initially, and then the stabilizer board itself is required. I can't show it live now, because
I have only one and I built it into this box. This is also a preamplifier for a player. Moreover, several of them are built in there at once for comparison, it doesn't matter. It is there. I will attach pictures here for you to look at with the parameters. They are much better than the parameters of this simple stabilizer that I showed you before. It is assembled on TPS7A4701 and TPS7A 3301. Accordingly, plus and minus and provides quite
stable operation. By the way, it is a reliable board. It is very difficult to buy, it costs about 25 euros, that is, not cheap. If that one costs 3 euros, then this is 25. That is, you understand. That is, there is some difference in all this and without a rectifier. That is, I also need a rectifier with capacitors, like the one I showed you. That's how it is. Again, you won't catch anything by ear. But in the pursuit of perfectionism, it kind of makes sense. And everything was fine until the same Sukhov
threw in the idea of powering it all from some kind of unipolar power supply in a wide range - from 3 to 30 V, for example, or even from a USB outlet. Which is what I actually fell for. I found it interesting. All this is done... all this is done on a certain XL 6007 microcircuit. I'll show it to you now. Here are its parameters . You can see what it is. At 400 kHz, it converts to a bipolar supply voltage, that is, it works with an input voltage range from 3.6 to 24 V, for example, and it converts all this...
This is what it looks like externally and it has certain parameters that are completely satisfactory. Here is its structural diagram, block diagram, block chart, which shows how it works. By the way, here is the basic diagram. That is, one voltage is applied to the input, two voltages are removed from the output, which are quite capable of powering a wide range of these electronic devices, by the way. That is, it holds
a pretty decent current. Well, so to speak... I don't remember how much it holds there exactly... You can look... Let's say that in this connection all this allows it to hold, well, very well. Let's say there is a diagram like this, many different implementations are assembled on it based on this microcircuit. There are other microcircuits, I just gave you this one as the most illustrative, providing the necessary sufficient parameters for powering a large number of electronic devices. I hope you understand that we are not talking about power amplifiers, but about some preamplifiers , DACs, converters beloved by many, some streamers, shmimmers, well, things that consume from roughly to half an Ampere. Although this is an excessive current... Well, let's say up to 100 mA - this is considered low-current equipment, low-current, low-power. I bought three types of such converters in order
to compare and identify the winner, to recommend this to my partners, friends, and of course to myself. Why do I personally need this? Look. Here I have a Thorens, it is powered by an adapter with a DC voltage of 16-12 V, it doesn't matter. It is plugged into a socket. That is, in order for me to build in this Sukhov preamplifier here, or any other one to build in, I will have to assemble a network power supply, bring it out separately, drill holes, to bring out the wires and place the switch somewhere. There is no switch here. There is a switch right here, which turns on these 16 V there and I cannot supply 220 there. Okay, if my preamplifier were powered by 16 V or 12 V, there are diagrams, a lot of diagrams. I even have such a preamplifier. Here. But at the same time, the parameters of this voltage in terms of pulsation are not top-notch,
that is, this will be a bad power supply. This is not accepted. Still, the preamplifier needs to be powered from something, well, solid ... In short, I would like to build a Sukhov board there, for example. Closer to the tonearm ... well, as it should be. I talked about this in the video about this preamplifier. So, I bought three boards. Now I will show them to you and tell you about them at the same time. Here is the first board that I bought, it looks like this. From now on, I will call it the converter one board.
Converter 2 looks like this. There is also one input voltage, and two output voltages. You see the minus and plus voltage, this is number two. It is a little larger, but much more powerful, I will say right away. And converter number 3, miniature, like this one. But functionally it is exactly the same... That is, it has a unipolar input voltage and a bipolar output voltage.
I bought three boards like this, each of them is described by some of its own parameters, of course... All this is very good, but it provides functionally the same thing. That is, power supply of electronic equipment from a bipolar voltage. First, I will immediately note that I bought the first board in two versions. There you can set the jumpers... Let me
show you. You can set the voltage with jumpers, of course, here, with jumpers. But I immediately bought several of them in order to change the voltage and supply less. I don’t need plus or minus 18 V for a Sukhov fuse in general. There, plus or minus 12 V is more than enough because... But I stopped
at plus or minus 15 V, I will say right away because the preamplifier generally heats up firstly and some people even make radiators on the chips that are there. Which makes sense in a confined space. For example, here is a typical amplifier. Also based on the same Sukhov's accessory. Here I even made ventilation, you see a through one here, so that there would be some convection. And here it will still heat up, so I
made it a little, say plus or minus 15 V. It will not affect the final characteristics in any way... Well, theoretically there are fractions of a decibel, but the integrity is more important, so that it does not burn out simply from overheating. The temperature regime also needs to be taken into account, by the way, this is also one of the important parameters of the design of radio-electronic equipment, which many audio enthusiasts omit as generally insignificant. In fact, it can bring quite serious problems. For example, the same Sukhov preamplifier will be connected to the MC head, let me remind you that it has a
very low resistance, about 100 Ohm, I think, in the case of the same thermal breakdown of , say, the field-effect input transistors burning out, it supplies all these pluses and minuses to the input, well, some of the 18 V voltages. Well, and the head safely burns out, which happened to one of the people I contacted. That's it. And this is very bad. That is, it is better to supply less voltage and protect yourself in this way. There is another way, which I do not want to talk about, so as not to irritate Nikolay Sukhov. So, in order to protect the owner of the MC head from losing quite expensive equipment, that is, at least 300 euros flew down the drain for the person... Well, plus the corrector itself, which also had to be re-soldered and so on, that is, this is an important thing - the temperature regime. And here is
the supply voltage reduced to plus or minus 15 V or even 12 V, - I think this is laconic, elegant and correct. In short, the optimal way to achieve this disaster ... So, I showed you this, these three converters. Yes, they are sold in these little bags on Alibaba, which I showed you there. So there were options here that these 400 kHz, at which they work, this will somehow go to the input and there will be some harmonics, some modulations and so on ... It's a shame, a shame, but no good. Therefore, presumably we planned, I planned to solder, for example, chokes to the output of these boards, to the output of this board . According to the calculations, 2.2 mH are obtained on the plus and minus. So that high-frequency interference lies there. For this purpose, I bought, like a fool, candy wrappers again, different
ferrite products. I don't know what to call them - chokes. Well, how chokes? Well, okay, chokes. When I didn't have them, at the very beginning, when I was struggling with this, I used ferrite coils from speakers. Thank God I have them. In short, I bought three types of different ratings. I changed them all. Here I showed only an insignificant part. But let's go through them one by one, what do I have here? First, these are these coils, I bought a rating of 2 mH, as you can see it's written here. By the way, a resistance of 0.1 Ohm is one of the best indicators, and 0.6, excuse me, 0.6. Here I looked closely - 0.6 Ohm to direct current. Resistance is meant.
This is kind of very little because all these turned out to be.. for example, 26 Ohm... This is 4.7 mH. I also bought 2.2 mH. They are under them. These are the type. But I want to say this 1 W right away. This is very important because I first bought a quarter of a watt, I think, or half a watt, I don’t remember, one of them burned out, the second one burned out, that is, its rating was not 26 Ohm DC , but had a different resistance. This is not good. So the third format is these coils. Well, I bought several of them of different ratings, too, 0.6 Ohm, you see, 3.3 mH, like this.
I called them black or something, maybe I just called them black when constructing these graphs. I fed all three types of converters - the first, second and third - from two types of input supply voltage. The first with the help of a USB-C connector, to which I supplied voltage from a regular charger. I will say right away that the simplest charger, not designed for high current, did not pull. It laid down the output voltage. There was nonsense and the device did not work, and from USB-C, that's where it all worked. And the second thing is from a 12 V adapter,
I simply didn't have another one. The ones I took were the first ones I came across, I have several of them. I tried with several. It works about the same because they provide a noticeably large current, sufficient, let's say, to power such a bundle. What did I get... Let's first analyze this noise floor. That is, the background of my sound card, so that it is clear at what level all these measurements are made. In this graph, you see the included converter number one 10 minutes after the first switch-on...
Here is the first switch-on, here is the second switch-on. That is, with warming up, in principle, its parameters improve. I cannot be sure that all the data that I took was taken with a warmed-up device. Therefore, some drift will still be in all these readings. The following graph is when powering converter
number one from a USB adapter via a USB-C connector, or a 12 V power supply - a network adapter that you saw. You can immediately notice that the 12 V adapter has noticeably better noise characteristics than USB, and in the future I stuck to measurements with this 12 V adapter. Here you can see the different types of chokes that I used with the converters. 12 V converter number one is used everywhere. Well, let's say, this is direct - this is a direct connection. That is, without chokes. This is 4.7 mH - radial ones, they look like resistors. These are round coils wound on ferrite rings, 2 mH
you see. You see, these are a little inferior to 12 V 2.2 millihenries - the same radial as resistors, that is, here are four, let's say, here are two, and here is converter number 1, with those black ferrites that they have, well, the ones I showed you. Well, let's immediately see what happens if 2.2 mH is compared with 4.7 mH. Let's do it this way: here are 4.7 mH, and here are 2.2 mH, why is 2.2 mH better, given that their inductance is lower and they seem to dampen high-frequency components less? Here, perhaps, the topic of warming up the converter worked, but most likely, after all, this is caused by the reduced resistance of this ferrite 2.2 mH by direct current literally, in contrast to this one, I suppose. Well, in the end, here they are all dumped in one heap. These are all correctors without chokes - 1, 2, 3,
direct connection - direct, you see, I am writing and in comparison with the network, transformer. That is, the network transformer with a stabilizer is clearly better than these converters, but this difference cannot be called fatal. It cannot. That's how things are, my dears... In conclusion, what can be said about this.. That a miracle certainly did not happen, but I will completely and entirely trust these converters of the first and second types at least, which I showed you and use them for similar purposes, delicate purposes, frankly speaking, instead of a bulky network converter, which by the way in height, this transformer, it doesn’t even fit here. There is less space here in Thorens than it seems at first glance. Well, you can make it stick out from below, but this is in the
absence of another solution. I found it, it suits me. I like it completely ... I also must say that of course 400 kHz is spewing there with all its might, that is, without any options there is interference there. They somehow overlap, somehow penetrate to the input and from them these humps actually appear, precisely from them, from those that are far away there. Therefore, in principle, chokes can be installed. It would also be a radical way, of course, to shield the boards, and both the preamp boards and the boards of this converter-converter. What to shield with? Well, it’s hard to say. I think permalloy is needed. People say that any metal saves...
I assembled my first preamplifier, if you remember in the first video about Sukhov's preamplifiers , in a box soldered from double-sided foil-clad textolite. I did this 40 years ago when I worked at Sergei Pavlovich Korolev's association, where we developed the Odyssey amplifiers. I mean, they were developing it in my presence, I didn't develop it. I was a senior technician there. I soldered something there, the robot serviced them, which tested them. But I saw how it was done and picked up some of these,
let's say, clever tricks. In terms of the fact that unshielded equipment cannot be used at all because sooner or later, one way or another, it will manifest itself. By the way, what I had when I first tested these boards. I had two more boards for 15 and 18 V, so I drove them, a number of measurements turned out good, and a number of measurements, the majority, were very bad. That is, so bad that I cursed all this, wrote an angry letter to Sukhov, that this method can only be used for children's toys, trains and gliders. I mean power supply.
Then it turned out that I put a soldering iron here, and it has a regulator. Well, such an elegant, thin soldering iron, for SMD in particular. And a power regulator there. So. When it is turned on - there is interference. Monstrous interference, so you understand ... That is, this was not there and nowhere near ... It was like this, let's say. Well, really bad. In short, I thought somehow I soldered it wrong ... Well, how
can you solder three wires wrong?! That is, I applied it to the input here, measured it with a tester. Here comes 12 V, there is something there, or there comes 5 V from USB, the output voltage is also plus or minus... Everything is fine... I don't have an oscilloscope, but Sukhov's preamplifier, as a device with a huge gain, immediately reacted to all this... That is, this is an absolutely practical thing.
That is, if I built everything like this and approached the player with this soldering iron, well, it doesn't matter what to solder here, then it would whistle there, in the sound... Actually , this is absolutely terrible. And if it was shielded - the amplifier. This would not have happened. Well, and of course, there is no need to approach
delicate audio devices with high-noise devices. These are my last instructions, advice, so to speak... Well, I hope someone will find it interesting and useful. And now, for now, friends...
2025-06-01 04:37
