Class A headphone amplifier

In the basic version of the class D amplifier there are no A / C processors but the output filter is important

michald wrote:

Tell your children about your bedtime stories.

Take part in the blind ABX test, select the amplifier correctly and then write out that someone is telling fairy tales. A CodeBoy colleague relies on scientific research, not pseudo-scientific opinions.

michald wrote:

In class D, the sound is "created" and the quality depends on the A / C processors, power supplies and smoother filter. In class A, analog sound - the whole waveform is amplified simultaneously. You made a mistake with the AB class.

Nonsense. In class A, AB the sound quality is also influenced by "Power supplies and smoothing filters". In class A, AB, the quality also depends on the reinforcement elements. Less shamanic electronics, more reading and understanding of physics, mathematics and electronics textbooks. Then it will be easier to understand the surrounding world, and not to explain it with strange theories, for example: "the entire course is strengthened simultaneously. You have mistaken the AB class". I am not writing this ironically, you have to work from scratch, sometimes go back to fundamental concepts.

"the entire course is strengthened simultaneously." - So what, the whole three-minute song gets stronger at once? No, the signal value is amplified. Value as a function of time dependent. The function is continuous. You can't even describe correctly, express your thoughts.

Mandatory reading, not necessarily in this order:
1. Anatomy and physiology of hearing
2. Sampling theorem, Nyquist-Shannon theorem + additional mathematical issues related to the topic if needed
3. Bolton "Outline of Physics"
4. Horowitz, Hill "The Art of Electronics"
5. Measurement techniques
6. A textbook for logic

In the 20th century, people also claimed that radio waves were evil and burned the masts. Probably in class D, the cows give curdled milk

CodeBoy wrote:

In the 20th century, people also claimed that radio waves were evil and burned the masts. Probably in class D, the cows give curdled milk

It is the result of a lack of knowledge. Primary.

popiol667 wrote:

CodeBoy wrote:

In the 20th century, people also claimed that radio waves were evil and burned the masts. Probably in class D, the cows give curdled milk

It is simply the result of a lack of knowledge. Primary.

https://www.youtube.com/watch?v=MkqmZeCm31Y

CodeBoy wrote:

popiol667 wrote:

CodeBoy wrote:

In the 20th century, people also claimed that radio waves were evil and burned the masts. Probably in class D, the cows give curdled milk

It is simply the result of a lack of knowledge. Primary.

https://www.youtube.com/watch?v=MkqmZeCm31Y

And how, it can't play well, because it's class D and it is obvious that class D can't play well, because it doesn't amplify the whole signal at the same time. In addition, it is a digital amplifier, not an analog one, that is, it dirty the sound and interferes with it. And so on.... ;)

And now seriously - I do not understand the statement from the above film that guitars are not "played with the world, with such commitment" ... I just do not know what it means and I think that the author of the film had to add his 3 cents to the description of the sound, while when there was nothing to add, because the amplifier is completely transparent in sound and exemplary. Well, but if I'm wrong, please correct it. Plus, apart from the impressive engineering, it's just really pretty

Fimek wrote:

"played with the world, with such commitment" ... I just do not know what it means and I think that the author of the film had to add his 3 cents to the description of the sound

The point is that music is dealt with more emotionally than with physical quantities measured in a standardized way. With the sound played electronically, it is impossible to reproduce 100% what we would hear live. What we hear is influenced, among others, by the sound source, amplifier, speakers / headphones, how a given person hears and many, many other things. The fact that the amplifier is completely transparent does not necessarily mean that the sound is pleasant to listen to, it is the individual feelings of the person who is listening to the audio system. Similarly with reviews of other things, imagine that someone describes you the color of the car they like very much, they will not give RGB values, but will use words like: deep, warm, bloody, etc.

As for the discussion, which class is better A, AB or D, according to I shouldn't be told that this one is better than the "because so and now". However, it would be decisive to compare the specific two models of amplifiers from different classes and then describe the differences between them and this does not mean that if class AB performs better than D from these two, then any other D is bad.

mycodename47 wrote:

Similarly with reviews of other things, imagine that someone describes you the color of the car they like very much, they will not give RGB values, but will use words like: deep, warm, bloody, etc.

The point is that if you paint the same paint on the Ferrari and the Toddler, the one on the Ferrari will seem prettier than the one on the Toddler. It will be deeper, bloodier, and so on. You will stop seeing the differences when you compare only a part of the car body without knowing which car it belongs to. Similarly, you will hear colossal differences knowing which amplifier is playing and you will stop hearing without knowing which one is currently in use.

Heh ... There is something to it, human nature will subconsciously choose Ferrari as a prettier color. But coming back to audio, if, on a blind test for a given person, both amplifiers sound identical, of course it is wiser to take the cheaper one (if the appearance and other aspects are negligible). However, I do not agree with the fragment of my colleague's statement:

Fimek wrote:

because the amplifier is completely transparent in sound and exemplary

I suspect that most likely none of the participants in this topic listened to the device from the YT link, I certainly did not. But I know that each amplifier "adds" and "subtracts" something from itself, there is no such thing as complete transparency and modeling. Among other things, this subtraction and addition affects the emotions I wrote about, how we hear instruments, vocals and other things in the track we are listening to. It is not real that any amplifier in the world would not distort the sound played in the entire range of work. It is often even so that people like distorted sound more, an example here are designs built on tubes, i.e. additional even and odd harmonics, or softer, sloppy bass compared to the end of power on transistors, which is caused by a transformer with a fairly high impedance on the secondary winding of the output transformer (that is why the loudspeakers for these amplifiers usually have an impedance greater than 4 or 8 Ohm). If you listen to the same track on any tube amplifier and then on a class amplifier, for example, A, B, AB, D and in a blind test, you do not see a difference, then according to me, the listener has hearing difficulties. And I am sure that there will be supporters of "hard", "clean", less distorted Class D sound, as well as those who prefer tube sound.
Even if we take two different amplifiers of different classes, but not tube amplifiers with very similar parameters (measured in a standardized way), it does not mean that they will sound the same, full stop. Whether the blind test docker will hear the difference is another matter. I know that the differences can often be heard.

mycodename47 wrote:

But I know that each amplifier "adds" and "subtracts" something from itself, there is no such thing as complete transparency and modeling.

OK, how do you know that? From own experience? And wasn't this experience disturbed by auto-suggestion?

mycodename47 wrote:

Fimek wrote:

"played with the world, with such commitment" ... I just do not know what it means and I think that the author of the film had to add his 3 cents to the description of the sound

The point is that music is dealt with more emotionally than with physical quantities measured in a standardized way. With the sound played electronically, it is impossible to reproduce 100% what we would hear live. What we hear is influenced, among others, by the sound source, amplifier, speakers / headphones, how a given person hears and many, many other things.

But I do not negate the influence of the loudspeakers or the psophometric characteristics (I think it is called) of our ear, i.e. different subjective perception of music depending on the volume. But I mean only a power amplifier, which is, in principle, a primitive device (not this one - here's a nice engineering), with very good parameters in relation to our hearing aid, provided that someone has not screwed up something. Putting an amplifier, speakers and a human ear into one sack is a mixture of completely different domains. I just think the power amp "doesn't sound" unlike, for example, columns or a room.

However, not referring to your statement, but referring to the topic in general - class D can actually play poorly, but this is not a matter of the principle of operation, but of implementation. As far as I know, the source of distortion in real systems is an incorrectly taken signal for the feedback, i.e. not from behind the LC filter at the loudspeaker, but directly from the PWM output + RC filter, or when the amplifier works without negative feedback at all, and unfortunately they have cheap amplifiers. Moreover, when the upper and lower key engagement dead zone is large, then something like cross-distortion appears and it sounds bad. But let's not mix high-class amplifiers with 2x20W pimp in the mini-tower and let's not generalize by pouring swill on the poor class D

I design high-end audio devices on a daily basis and I managed to convince myself that the amplifier does sound. You can find out about it in any electro market on inexpensive ONKYO, DENON etc. masks. Approach the connected set and you will hear that despite the low price (2-3 thousand PLN) the sound is really nice (apart from the parameters given by the manufacturer). Then press the button "Pure Audio" and you will hear how "average" this amplifier sounds. On average, in quotation marks, because you will feel that there is less bass and that you have to listen to the treble to hear the details, and in fact, everything that can contribute to the signal is omitted in the audio path. Other manufacturers also use such tricks. In class D, for example, decent DSP often has a positive influence.

As for the second part of your statement, I agree, I even tried to convey the same in previous posts. Some time ago, my employer gave me the power end in class D for several thousand zlotys to play and test. and I listened to it a bit. I think he plays very well. I was just comparing it with the ONKYO apliter (AB class) and I guarantee that in blind tests I would recognize which one is currently playing. For example, after the bass, in the ending in class D, the first impression is that there is less of it, but after a few replacements of the amplifiers, you can clearly hear that in class D it is more detailed, hard, I do not know how to describe it in words, you have to hear it , perhaps this is due to the fact that class D has a very low output impedance, which makes it better control the voice coil of the low-tone speaker. Then, after connecting a DIY tube amp, it turns out that the sound is "nicer", "warmer". In all three options, the trumpet or guitar sounds different and you can really hear it. It's hard for me to say where is the best because I would also need a guy with a trumpet and a guitar.

@ mycodename47
Thank you for your answer. Your honest and calm tone of expression sounds very good, but my doubts remain My equipment is turned off, because ... because it is, and because the purchase of speakers and headphones that I have was preceded by searches and also a few disappointments, and the equipment that I use now satisfies me without assistance. It doesn't mean that they are particularly high-class, it just means that I like their sound. However, it was not, the amplifiers are not related to the principle of operation of the amplifier. On the other hand, regarding the low output impedance ... yes - class D has it by definition, but on the other hand most of the work still does the negative feedback present in modern amplifiers, regardless of the class. Maybe ... it would be worthwhile for you to do such an honest, honest test - a blank test - and let the electrode know what came out?

Regards,
Fimek

My point is that there is no amplifier that multiplies the input signal by GAIN and that's it. There are always distortions, in some devices smaller, in others larger, sometimes pleasant, indifferent or bothersome.

As for the negative feedback, which in addition to increasing the bandwidth of the amplifier, reducing non-linear distortions, lowering the output impedance and other advantages, causes a change in the input signal. When you measure the distortions for e.g. 1khz, 2khz, 7khz, 12khz and you get very good results, so ultra-low distortions, it turns out that when the amplifier receives these signals at one moment, it will turn out that each frequency will shift in phase differently, resulting in the sound, apart from amplification, may be strongly distorted. Manufacturers often boast about such measurements, where they show perfect fringes in the spectrum, sometimes even averaged over time, unfortunately these beautiful charts do not show phase distortions. This is where the different-sounding trumpet and guitar come from.

I recommend interesting articles, supported by measurements:
Link
Link see Figures 3.4.6 and 3.4.7

Sometimes it is even so that the amplifiers start to oscillate outside the acoustic band, which is usually not audible, but it causes strong heating of the semiconductors. The simplest solution is to add a small capacity in the feedback loop, which also affects the phase distortions.

popiol667 wrote:

The point is that if you paint the same paint on the Ferrari and the Toddler, the one on the Ferrari will seem prettier than the one on the Toddler. It will be deeper, bloodier, and so on. You will stop seeing the differences when you compare only a part of the car body without knowing which car it belongs to. Similarly, you will hear colossal differences knowing which amplifier is playing and you will stop hearing without knowing which one is currently in use.

Likewise, if you paint a toddler with Ferrari paint, a toddler D-Class will not fart from the exhaust like a Ferrari. So much for the topic.

Added after 2 [minutes]:

mycodename47 wrote:

In class D, for example, decent DSP often has a positive influence.

Read: "changes the color closer to the desired quality".

Added after 7 [minutes]:

The sound is a bit like modern photos made for sale by professionals. Before they are offered, they are treated with post-burn agents. The skin of the models is edited, contrasts and colors are enhanced, etc. Unfortunately, the same applies to music. All you have to do is compare the concerts recorded by a sound engineer for the studio, radio and TV. The latter is practically not suitable as a source for creating a demo, the characteristics of TV signals simply have different requirements. Going back to class D and the rightly mentioned DSP, nowadays it is cheaper and easier to assemble such a set and digitally simulate the desired colors at the output. Hardly anyone can distinguish it, but still it is possible. And most of all, this sound is not real. It is enough to listen to good recordings from the Chopin Competition or the best jazz tracks. Then you know whether the recording "breathes" or whether it is driving with lofty bands and dynamics.

yogi009 wrote:

Likewise, if you paint a toddler with Ferrari paint, a toddler D-Class will not fart from the exhaust like a Ferrari. So much for the topic.

In fact, until you perform the tests on the dynamometer, you do not know what the real parameters of the engine are and what someone put under the hood ...

yogi009 wrote:

Read: "changes the color closer to the desired quality".

The feedback in the analog amplifier can also be said to "change the timbre closer to the desired quality".
DSP is there to adjust the parameters of the amplifier to the signal, so as to reproduce the signal as faithfully as possible, not to change it.

DSP - why not (of course I do not mean indulgence functions from cheap towers), if it will improve the sound quality. It opens up possibilities unavailable for analog technology, e.g. correction of phase shifts of a loudspeaker, suppression of loudspeaker resonances, etc.

I recommend making the panel on an engraving laminate here is the result of my work:

Fimek wrote:

Hello

I would not say that it is a class A amplifier, if you do not know what is going on inside the op-amps - they generally work in class AB. Case two - what about the damping factor? There is a series resistor in the output, which worsens the damping of oscillations from the headphone diaphragms - this seems to be a workaround for the lack of short-circuit protection, but this solution is a very compromise solution. In my opinion, this type of amplifier is a completely unnecessarily small radiator, while it could be energy-saving and still have cosmic parameters, if only it was based on, for example, OPA-134 or another amplifier dedicated to headphones (read designed to work with a low load impedance) e.g. MAX97220A.

I'm not picking on aesthetics because it's OK and you can see that you've put in a lot of work - this should be praised :)

Greetings,
Fimek

And you made an amplifier on the "OPA-134", that you know it would have "cosmic parameters"? How does the cosmic "OPA-134" with an impedance of 600ohm cope? The short-circuit protection is "double". The first is a series resistor at the output, the second is an integrated circuit that will disconnect the headphones immediately. I included everything in the description.

In my amplifier I designed one PCB with everything - dedicated transformer, stabilization, power stage, loudspeaker protection. It took me quite a long time to design the PCB "out of the box". I did everything from schematic diagrams . This cosmic OPA134 is a single operational amplifier, and in my system I have a double OPAMP. Not only that, my systems are powered with +/- 22V, because from one branch of 22V I supply 24V relays that activate the columns. If you know how to easily convert a PCB for this space OPA, and at the same time convert the power supply for the relays, give me some tips. I have a power supply with voltage regulation - this makes it easier. Alternatively, show how you solved this problem at home.

Why quarrels? Check if you will like any LME series amplifiers from Texas. Or, consider the LME49600 buffer. For this automation on OPA137 and you have the next generation of the amplifier. It was described on the forum, I tried it myself, I even collected most of the "bricks", but there is no time to deal with it all the time.

liseczq wrote:

And you made an amplifier on the "OPA-134", that you know it would have "cosmic parameters"? How does the cosmic "OPA-134" with an impedance of 600ohm cope? The short-circuit protection is "double". The first is a series resistor at the output, the second is an integrated circuit that will disconnect the headphones immediately. I included everything in the description.

You didn't understand my intention. I referred to the diagram, the only one you included in the description. OPA134 is not some unbelievable op amp that I have on the altar to pray to its cosmic technology, it's just a modern opamp whose very, very low nonlinear distortion and low noise related to the input make what he has apart from himself. it adds an order or even orders of magnitude below the threshold of human perception. By the way, there are many more such models of amplifiers - to choose from, depending on the color, depending on what you want to get. They cost about a few dollars apiece (if you can get them today, those are the times). The schematic diagram with their use is simple, which makes the whole device small and paradoxically maybe not very output ... but for me it's not a problem, because I like compact systems. I also like the use of modern elements.

PS. It just so happens that I have the knowledge to design a discrete amplifier with very high open loop gain, low closed loop distortion, high slew rate, low current and voltage noise, full load stability, low output resistance and so on and the like. I do not brag, because there is nothing, many people here can do it, but rather depreciate the share of magic and the participation of a sublime ear in audio However, I don't see any sense in this, because the same can be achieved by using a much smaller number of elements and making it additionally energy-saving. For me, the challenge would be to harness this knowledge of an analogue to design a great-sounding fuzz with full awareness of what you are doing, or some other analog guitar effect, or, for example, to design a well-sounding class D amplifier for headphones (precisely because it is allegedly not possible. da). Only when it comes to this amplifier, it would be art for the sake of art, because I wouldn't hear the difference anyway, only measuring instruments would see it.

I am linking to something that may give you food for thought:
https://www.meizu.com/en/accessory/hifipro.html
https://www.audiosciencereview.com/forum/inde...hifi-pro-usb-c-phone-dongle-dac-review.10423/

greetings

Fimek wrote:

OPA134 is not some unbelievable op amp that I have on the altar to pray to its cosmic technology, it's just a modern opamp whose very, very low nonlinear distortion and low noise related to the input make what he has apart from himself. it adds an order or even orders of magnitude below the threshold of human perception.

Well, Fimek, you take a risk giving here an example of a better opamp And what do you think about capacitors from branded electrolytic capacitor manufacturers (eg Nichicon series FW, FG, Elna Silmic II RFS, etc.)? Should their use improve the sound of an analog audio track? Even such a headphone amplifier?

yogi009 wrote:

Well, Fimek, you take a risk giving here an example of a better opamp

It all depends on what it is for. For example, AD797 or LT1028 have several times lower voltage noise, but higher current noise. Therefore, they can work better with a dynamic microphone or a magnetic phono cartridge - it is a matter of noise matching, similar to RF. And for high levels of control - there are nice opamps dedicated to work as the final stage of ADSL modems. Fast, powerful, low-noise Hence, the mentioned OPA is not the best objectively, there is one of many (well ... maybe not that many ). When choosing an opamp, I would take into account the 1 / f noise, which is sometimes negligible, but also large in terms of acoustic frequencies - depending on the model.

yogi009 wrote:

And what do you think about capacitors from branded electrolytic capacitor manufacturers (...)?

I have no idea What do you think? In my opinion - a better capacitor is probably smaller parasites, which can play a real role in the sound, but it depends on a specific place in the circuit. Personally, I would go for tantalum capacitors, which are sufficiently linear and have well-defined and time-stable parameters without the entire audiophile envelope (also probably financial). Unfortunately, they have a large ESR ... but this can be modeled and included in the design. Ceramic, on the other hand, introduces non-linearities and has an electrostrictive effect - hence they are avoided. However, it's not really magic, is it? You can even long DC chains, why not, and avoid interstage couplings altogether.

greetings

yogi009 wrote:

And what do you think about capacitors from branded electrolytic capacitor manufacturers (eg Nichicon series FW, FG, Elna Silmic II RFS, etc.)? Should their use improve the sound of an analog audio track? For example such a headphone amplifier

Generally a better capacitor - greater durability of the built device. I do not understand why some people put the cheapest Chinese food into their DiY constructions, after all, it is only suitable for throwing away from new products, or it will fall after a short time of work.

As for the improvement of the sound by branded capacitors - a common audiophile myth, after all, for Nichicon, Samwha produces, and some say that Samwha is "average" In my opinion, it would affect the sound more by a Low ESR capacitor than some expensive "audiophile" ;)

Of course, someone may disagree with me, but in my opinion, it's a waste of money for these audio-dedicated capacitors, it's better to buy regular Panasonic, Samwha, Nichicon, and if you can afford it, BC Components or Roederstein.

Greetings,
AND.

Fimek wrote:

I personally would go for tantalum capacitors

Exactly. In the listening tests on the preamplifiers that I made several hundred times, the tantalas fell off, the ceramics caused too sharp high bands, the electrolytic ones were the best. I made a simple amplifier based on the note:

https://serwis.avt.pl/manuals/AVT1746.pdf

I did two versions, on Jamicon electrolytes from a nearby electronics store and another, on Nichicon FW (if I remember correctly). Otherwise, the modules were identical. The second module sounded much more selective, the first one as if all the sounds were thrown into one bag. This, of course, is great for picking up when you have both modules on the table. And then fun with the input 2u2. From ordinary electrolyte, through bipolar, MKT Wima and collapsible Audyn 2u2 / 160V. The differences were drastic. And on the forum, as usual, they give the audio-woodoo patch for such listening sessions

yogi009 wrote:

I did two versions, on Jamicon electrolytes from a nearby electronics store and another, on Nichicon FW (if I remember correctly).

No wonder there was a difference between Jamicon and Nichicon, Jamicon is a low shelf.

Greetings,
AND.

yogi009 wrote:

Exactly. In the listening tests on the preamplifiers that I made several hundred times, the tantalas fell off, the ceramics caused too sharp high bands, the electrolytic ones were the best. I made a simple amplifier based on the note:

I just can't agree with you.

To establish the remark - I assume that we are talking about coupling capacitors. Tantalum capacitors with a capacity as in this design have an ESR of a few ohms. Much for a capacitor and negligible for cooperating resistances (here 22k and 1k). I don't see any way that the linear distortion (i.e. the waviness of the frequency response) can cause any audible effect. The human ear is very insensitive to linear distortions. Tantalas do not like reverse polarity - if you just used tantalas according to the polarity as shown in the diagram, they work in a suboptimal way that can simply destroy them or at least cause non-linear effects. To be able to use tantalum as coupling capacitors, they must be biased. As for the bandwidth - it so happens that I once measured their transmission in the range of up to 1.5 GHz. Imagine my surprise when it turned out that with the impedance of the source and the load of 50 Ohm, they correctly transfer the signal to at least 1 GHz. There is therefore no question of clipping the audio signal.

On the other hand, when it comes to ceramics - where you need 22 uF, you had to use the Y5R, which has a terrible voltage capacitance factor. In turn, the electrostrictive effect places them somewhere between the capacitor and the microphone.

In other words - such literal replacement of electrolytic capacitors with tantalum or ceramics is a system error. But by no means anything that could escape the gauges or logic ;)

I would design the audio path in such a way as to avoid interstage coupling with capacitors - I would rather minimize AC coupling in favor of precise DC circuits, and where capacitors would be necessary - I would ensure the initial polarization of such a capacitor. And of course he used tantala.

greetings

Buddy FIMEK, I still don't understand your intention.

The OPAMP that I have at home are:

LME 49860, THD - 0.00003

Space OPA 134, THD - 0.00008

So the OPA has more distortion. Looking at the new one, the OPA is practically weaker than the LME in everything. It is also less modern than the LME. The LME went into production later.

It turns out that the LME is even more cosmic.

Olkus wrote:

Jamicon is a low shelf.

They had one in a local store, but they said here that I was inventing dyrdymały :-)

Added after 58 [minutes]:

Fimek wrote:

I just can't agree with you ...

They are decouplers, but also those in the audio path. I did tests with both X7R (the vast majority) and Y5R ceramics. As for the results ... I know that the tables and results are mine, and the musical ear. The Stradivarius violin has not been described by any model until today.

yogi009 wrote:

Olkus wrote:

Jamicon is a low shelf.

They had one in the local store, but they said here that I was inventing dyrdymały

They are also in my local store But I do not recommend buying them. Those were good capacitors in the 90's, now Jamicon belongs to Yageo and now the capacitors are useless :(

Greetings,
AND.

liseczq wrote:

Looking at the new one, the OPA is practically weaker than the LME in everything.

Because it is. I did comparisons on the dual OPA and the entire LME series. These are good op amps, but OPA2134 has long been referred to in the West as a good entry-level opamp for audio applications. Now you should probably take e.g. LME49860 (or the neighboring models) and the LME49600 buffer at the output.