Apologies for not including this in the video itself, but the main advantage for class D amps is that they can be made EXTREMELY efficiently, whilst also having extremely low distortion. Class D amps are always either fully 'on' or fully 'off'. Meaning they're either outputting efficiently or using basically no power. This means that class D amps can sometimes be 90-95% efficient, vs the 40-80% efficiency of class AB and the maximum 50% of class A. This makes them really suitable for situations where massive power is needed but an enormous class AB amp would not be feasible. Such as for subwoofers or car audio. You hardly need any heat dissipation and so a very powerful class D amp can be made to be extremely compact. The main disadvantages are that they inherently generate a huge amount of high frequency noise which must be filtered out, and many class D amps are also load dependent, meaning the filter response will change depending on the impedance of the connected load. For speakers this is not really an issue as you know everything will be roughly 8 Ohms. But for headphones where it could be anywhing from ~12 Ohm to 600 Ohm, that's a more difficult challenge. For that reason, and since we don't really 'need' the efficiency for headphone amps anyway which don't use much power, class D is rarely seen in headphone amps.
Any class D based on the Universal class D design and especially the modern variations from Hypex, Purifi and such handily beat 98% of A/AB designs in objective and subjective tests. While being significantly more efficient.
Please, excuse me this probably dumb question. Do integrated DAC/Amp chips commonly found in Bluetooth devices like Qudelix and FiiO's BTR lineup (ES9218/9218), as well as USB dongles (CS43131) employ class D amplifier technology?
@@DJ_Force it's not that different... There will be some variance yes, but in practice your ears will hardly be able to tell much of a difference really. Also there are already some class d implementations that solve that problem.
I loved the explanation. And complementing that, it is very easy to use Power MOSFET transistors to implement class D amplifiers, and most MOSFETs I have at home are only limited to its package limits, meaning a small TO-220 transistor handling 75 amps easily (that would be 7kW at 100V) without significant heat. I don't use them on amplifiers (all audio amps I have are AB), but they would be a good choice to make a power amp for sub.
It seems that everyone commenting in these kind of videos had troubles understanding concepts due to having incompetent teachers. Maybe you were just bad students?
@@UntakenNick Think about school as a culling operation. The failures of the profs are hung on the students. My Eco 1 professor was Indian. Took me 2 months to understand "Good Morning Class". I learned Guns and Butter though.
@@robohofo1 Lol, I'm writing my first mock exam on Signal Processing on monday based on the McClellan and Schafer "Signal Processing First" book and I feel a bit overwhelmed by the density and pace of my course. The book is fantastic though. I've even contacted Professor McClellan via E-Mail to ask him, if there is updated coursework for the book, since the included CD is for windows 2000 and he responded in a coupled of days even though he is already emeritus.
@@UntakenNick Now, now, did you expect them to understand concepts by reading technical papers and conducting experiments? They're geniuses, they don't need that!
No one, and I mean no one has explained this better for a person to understand how sound is manipulated by different class amplification. I think it is very easy to understand once you get to the 3rd time watching the video. :)
Truly superlative explanation. Really glad I found this before spending any serious money in amps. Class A is clearly the most signal-honest design, though less efficient.
Nicely done, in my 50+ years of electronics I've not heard a more accurate easily comprehended explanation! I'd say if more had your knack for instruction the world in which we live would have a much higher rate of proficiency. You've actually made the learning process much less boring! I started my learning process when I was 13 YO, you remind me of the fellow I learned from, he was a 75 YO USNavy EW tech!
need more of these videos. so tired of everyone that just says "but that is a topuc for another video" but never get another one. educational content to educate the consumers have a much better effect than giving opinions in reviews
This is a great explanation of the main amplifier types, but it glosses over _why_ you'd go to the trouble of using a Class D amplifier since it sounds more complicated to implement. The answer, of course, is that they can be vastly more efficient than even a well-tuned Class AB. This is because transistors waste extremely little power when they're either in their fully off or fully on states, while a transistor in its linear region acts somewhat like a variable resistor and dumps the waste energy as heat. Since a Class D amplifier only ever has the transistor fully off (where it wastes no power) or fully on (where it also wastes nearly no power since it's extremely low-resistance) there's very low loss overall. Class A amplifiers are usually 25% efficient or less (somewhat higher is possible with tricks), while AB designs are 40-80% efficient. Class D, on the other hand, is usually 80-90% efficient, with even higher values possible. As such, Class D amplifiers can be absolutely tiny and, if well designed, still produce shockingly good sound. It took a while for the designs and implementations to get good enough to avoid distortions or other issues inherent with trying to keep the complicated harmonic-filled output under control, but nowadays you can get dozens of watts (or more) out of something the size of a deck of playing cards with very few compromises. Purists will insist on linear designs, but a well-designed Class D amplifier can produce excellent audio and allows much more compact designs because the reduced waste power. This is especially important where space is at a premium (such as in small apartments) or in portable applications powered by batteries.
Class D is typically used for subwoofer amps. They run cooler, and can be built smaller. Linear response doesn't matter because they run a LPF crossover.
The real life RMS of most speakers is very low, even if it sounds kinda loud. My receiver (plus PC margin for playing back music (lol so prob 1W over idle)) at the wall pulls maybe +12W for silent vs pretty dang loud. Proper RMS 10W is loud as hell. Doesn't seem like 25% efficiency on a class A is a very big deal when IRL RMS is so low unless you are hurting your hearing. Like, oh no my class A amp pulls 50-100W for an hour or two, lol, who gives a shit 😅.
@@chapstickbomber Lower power is fine, if the circuit is super clean. High signal-to-noise ratio and very low THD. But most amps, especially cheaper ones, are not that clean. It's cheaper to make more power, than to build a really clean circuit. Power can compensate for a dirty circuit because it gets loud before it distorts. If you keep it below this threshold, then you won't get the distortion. When you crank it up, you will hear the distortion. On a super clean circuit, you can open it up and it will still remain tight, with less total power.
Is anyone doing GaN amps yet? because a class D amp + GaN transistors seems like you would be able to get to 95 - 98% efficiency and with significantly more compact designs would want a GaN power supply as well at that point or otherwise your amp may end up being smaller than the power supply, also it would enable the transistors run at a much higher frequency making it easier for a filter to filter out the frequencies you don't want.
@@chapstickbomber Well for mobile application yeh that matters if your going to be running on battery and yes if you want a compact design yeh again that matters, halving the power lost to heat means you can probably reduce the size of your amp by ~30%
This is BY FAR the best explanation of this topic I have seen on youtube. Nice job! I found this video first, so I will be watching the DAC video next.
@2:18 The transistor will only switch on when a certain voltage is reached. It requires a threshold voltage for current to start flowing. (same as forward biasing a diode). This is a significant distinction. The moment there is base current, there will be an amplified collector current. Also, most class A amps used in audio are push-pull class A. They look like class AB, but they are biased so hard that neither transistor ever turns off completely. Generally, when the amp is is in the topology shown in this video as class A, the amp will be noted as "singled-ended" and it is practically always a tube amp.
Not a headphone person and more into speakers (sorry!) but this is the clearest explanation I've seen for amplifiers. While I have already known all this it's great
The last 9.5 minutes of my life were not wasted. Thank you! I spent the late 80s in electronics engineering school and I just learned more and ( this is the kicker) comprehended what was in this video than I did in college. Kids today have it made if they want to learn anything!
Easily the most clear explanation I've ever seen for the differences. Even at the fast pace the content was entirely clear and easy to grasp. Thank you!
I don't know why i've never seen a better and succinct explanation as this. After watching this, I can now explain this subject to someone else. Thank you.
For those of us with an RF background, it's easier to think of Class D as a modulation scheme, with the ultrasonic carrier being modulated by the baseband audio. It also helps to also know about Class C, with greater efficiency but more distortion. Class D extends Class C to its limit, by greatly increasing both efficiency and distortion, but then filtering out the distortion products, leaving the amplified baseband signal.
As someone in the matter for decades I found that intro hilarious. NPN refers to the arrangement of semiconductor layers: N-type, P-type, and N-type. N-type is being doped with extra electron, and P-type lacking electron (has a hole instead electron). It's neither "negative" or "positive" in any way. BJT on the other hand stands for Bipolar Junction Transistor. The rest of the video is spot on regarding audio amplifiers.
The N material (silicon doped with pentavalent elements like P. As or Bi) has a surplus of electrons in the conduction band, so calling it "negative" is not incorrect. Likewise the P material has a deficit of electrons in the lattice (being doped with trivalent elements like Ga, In or B) and therefore a net positive ratio of cations to free electrons. They certainly are negative and positive in terms of net electronic charge.
@@karhukivi You just contradicted yourself. Are pentavalent elements more negative than tetravalent elements like silicon? If yes, what happens when you put them together (similar like bonded PN)? Does the current flow by its own, or you must provide external current source like in ordinary conductor? What's with the forward and breakdown voltages, capacity, etc. Are they just a complex or imaginary properties like impedance?
@@DamirUlovec A pentavalent atom like phosphorous replacing a lattice position of quadrivalent silicon has four bonds with the neighbouring four silicon atoms using four electrons, and the fifth electron is free to move through the lattice. That is why the material is negative and conducts electricity better than pure (undoped) silicon. Lots of information on the web if you look for it!
I understand transistors now thanks to your fantastic graphical demonstrations!!! I came here for an explanation of what the differences are between class A, B, AB, & D amps are, and I'm leaving with more knowledge than I set out to learn. Great job!!!
You explained and visualized multiple topics so well to the point where I've re-learned and finally understood everything on a deeper level. Wow. I can find these types of videos only at 5AM haha. Keep up the amazing content!
Thanks for posting this as it cleared up a lot of confusion I've had about amps. Now I know that all of my 70's receivers have class A-B amplifier circuits in them!
👏 WOW. Didn't understand a word of this, but that's my problem. You sir have got a GREAT batch of compliments and "positive feedback" (notice the electronic terms?). Seriously, you've got a great bunch of fans. And you're a genius. Makes me realize what a schmuck I am, should've learned all this in school.
I'm off to the kitchen to get a frying pan and egg to cook on my Class A amp. In 1963 I learned how valves work. Then came transistors and I learned how they work. I couldn't figure out why electrons travel in a different direction to current. In 1967 I went round a transistor factory and saw how they were made. Then ICs came along and I was totally lost . . . . . till I was aged 56 and passed a technical exam in microprocessor computer systems. As others say this is an excellent presentation of transistor workings.
Lot of people pointing out a lot of details, but wanted to say I already knew and ignored every single one of them personally, but cleared up a few concepts I’ve been struggling with due to the nuances in the details, it’s actually quite helpful to be reminded of what does and doesn’t matter basically. I was over here stressing about nonlinearities and forward and reverse biases for example. And for how class A’s have so much hum and distortion but i really appreciate the design after seeing how simple they are.
This is by far far far the best explanation of all this. And some of the best explanation of any audio/music content I've seen on the whole internet. I am a professional musician and music teacher by the way. Good on you, this is mind-boggling value you're providing.
Well that's by far the best explanation I've seen on this subject. I'm quite new to all this and you've managed to clear things up quite a bit, so thank you!
As an electric guitarist who has vague, laymen-consumer understanding of tube amps, this helped me better understand the A and A/B amps I use, and my bass players amp (class D) which is so light and more powerful.
Better if not the best RUclips videos (straight to the point) Nobody has information without fillers (no TV shows anymore = how its made ...) ____ The visual was clear (could be better for some with an *AUDIO comparison* = actually hearing these limitations)
Wow!!! Fantastic! This must have taken so much thought and work in order to make a complex subject visually simple to follow. It would be great if you could add harmonics to your part two, as I’m trying to get my head around how these all affect what we hear.
That’s never going to happen. You’ll get a ton of trolling from people on audio science review saying you are experiencing placebo. I personally think they need to be told that harmonics do come through better with class a and ab but good luck with that…. They will just ignore you and say the “measurements” can’t prove it so therefore it can’t be true lol
Great video! Well explained. Just want to add, that as a guitarist I use tube-amps, and NOTHING beats Class A amps, even with all the downsides: price, heat etc. But once you've owned one, you'l never look back. :)
As an explanation of the topic... perfect. Clear, no dependent questions left obscure, simple but explanatory graphic, conclusive. Ever thought of making a course for certain high school, college and university professors on how to explain something?
Thank you for a clear and straightforward overview of the amplification classes - very interesting. P.S. I found your narrative a lot nicer to listen to when I slowed the playback by 10%, it sounds far more natural.
In the olden days, when we designed these, the D's were actually called Class-S, for "switched". My first design had a 300KHz 3dB bandwidth and was used as a motor-drive for uranium separtion systems. These motor ran up to 90,000rpm, but had variable speeds. We also used for our Lunar Roving Vehicles, but that motor drive ran at only 17KHz. Some of the frequency limits are imposed not by the electronics, but by the actual load. You cannot run speaker "iron" at 300KHz, since it would melt the local copper windings. Our Induction Heaters are typically down at 20KHz or so (variable depending on Curie temperature), and depending on the losses in the target material.
Some remarks… A single transistor (single ended) must be class A (or D) to have a useful output. A double (push-pull) transistor can be class A, B, AB or D to be useful. The configuration doesn’t make it the class, it’s the bias. Push-pull can still be class A. If you go from class A to class D you roughly double the efficiency, so +3 dB. If you go from a 82 dB to a 102 dB sensitive loudspeaker it’s 100 times more efficient (from 84 to 94 dB 10 times, from 89 to 92 dB 2 times). Using an inefficient driver with an efficient amplifier only produces more efficiently heat into the voice coil which makes the driver even more inefficient! On top of that a class D amplifier requires an output filter between amplifier and loudspeaker in which you lose power. Just like the crossover filters. So what’s the point of using class D? If efficiency is priority you must get rid of passive (high level) crossovers and have 3 or even 4 way (or 3 way and sub) loudspeakers, with efficient drivers. Otherwise most power is consumed in output transistors, voice coils, crossover networks and padding resistors (the whole purpose of a padding resistor is making a driver less efficient!).
I am surprised that no description of the transistor ever mentions resistance. In essence, it is a resistor who's resistance is controlled by the base signal. This is why they get hot, resistors turn current into heat.
Maintaining full digital from capture to just before the speaker simplifies a great many sources of distortion throughout the audio recording, production, transmission and reproduction process, essentially limiting it to the microphone, the ADC, the DAC, and the speakers.
brilliant explanation !, especially impressed by the D class and the flappy bird, perfect ! i have had fabulous amps for decades and NOW i understand how they work !
Wow! The best explanation i ever heard, but i already knew wich we best for me. Always a Class A amp. Running hot ? YES , Always high current ? YES .. best quality ? YES ! 3x YES . 🎉
3:41Amplifiers in class A are usualy hot when are mute. In this case all current heat a transistor and a sink. But when play music, part od the current goes to the voice coil and in effect don't heat silicon as much as in mute work. A Class A amplifier that plays loudly will run cooler than one that is silent. 3:53 Amplifier in class A can also run in push pull configuration. It's just a matter of the quantity of the quiescent current, BIAS point. This is like in class AB, by increasing the current you go from class B to class AB, but if you increase it more you will enter to class A, then both transistors conduct current throughout all the cycle. And of course with less efficiency. The SE with single transistor can also operate in class B and C, even D, but has huge THD and IM.
Yes, the description of Class A in the video is at best simplistic, at worst incorrect, as if it is single ended you need a capacitor to prevent the bias current being applied to the speaker, which is why most are push pull.
@@rcflighttestengineer5636 Thanks. The crossover distortion for class B (mentioned at 5:03) can also be reduced to a minimum or to negligible with good design.
30 years ago i had a Sony class d car audio amplifier that was 60 watts per channel. That was the clearest, cleanest sound I've ever heard and the amp ran cool at full power. Current draw was also very low.
Extremely well explained. I think the only thing that's missing is a qualitative analysis. Like, do class A transistors produce a "better" or somehow more favorable sound than AB? Are they cheaper to make? The video frames AB as a clear superior version to class A (and B), is that the case?
Class B biasing is typical in vacuum tube amplifiers. The tubes don't have the nonlinearity of transistors at low input levels. Two tubes in push-pull configuration, in theory, have lower distortion than two transistors running Class A-B.
You can get BOTH the Triangle Wave generator AND the Comparator in the same tiny IC, the venerable and inexpensive *555 timer/oscillator* chip. You feed the analog signal to be amplified into the Pulse Width Modulation pin. You also need to add 1 external transistor to make the basic oscillator output true square waves, so that at zero modulation the output is symmetrical 50/50 waves. You then amplify that to desired current and voltage with transistors operating with very low losses in _switch mode._ If the PWM function is done with the CMOS version of the 555, you can make the square wave frequency be 2 MHz, which is about 6.64 octaves above 20 kHz. Then it has to be filtered with a low-pass filter, likely an RL filter for the sake of efficiency. To have the output be essentially flat between 20 Hz and 20 kHz, you may want to set the _corner frequency_ of the low-pass filter to be 40 kHz or 5.64 octaves. _That still gives a very respectable 33.8 dB attenuation of the raw 2 MHz signal, when using a simple _*_single-pole low-pass filter._* If used with voice-coil speakers, some additional filtering will be provided by the speakers themselves, which behave approximately as RL filters. The overall effect is that very little high-frequency current will flow in the speakers, and it will not be reproduced because it is ultrasonic. Meanwhile the very desirable audio frequencies will flow and produce very nice audio outputs.
I used to spend an entire semester learning about how amplifiers work in the past and it was definitely not as easy to understand as this series. Fantastic work! If only you released the videos sooner 😂
You did a really good job with this. Now take a crack at explaining the increasingly popular ZOLT amplifier design, as found in Linear Tube Audio’s excellent sounding product line. Not really output transformer-less (OTL) not really but kind of class-D; that one will be a good challenge for you.
Great video, thank you for the upload on this short, but very informative, series. Would be great if you could do one on the audio black art that is valve/ tube amps.. Cheers.
As an audio amplifier designer, I would like to clarify and refine a few points regarding distortion and amplifier classes: Class A Distortion Characteristics: The statement "Class A can be harder to get low distortion" is misleading. By design, Class A amplifiers inherently have lower distortion due to their continuous operation within the linear region of the output devices. The active device(s) conduct for the entire signal cycle, avoiding crossover and switching distortion entirely. Achieving ultra-low distortion levels (e.g., 0.01% or lower) is relatively straightforward with a well-designed Class A amplifier, thanks to these linear operating characteristics. In contrast, Class AB amplifiers require meticulous design to achieve similar distortion levels. Class AB and Crossover Distortion: While Class AB amplifiers significantly reduce zero-crossing (or crossover) distortion compared to Class B designs, they cannot eliminate it entirely. This limitation arises because real-world transistors are not ideal, they have non-linearities, and variations in gain. Even with techniques like bias optimization and feedback, some residual crossover distortion often remains, particularly in less expensive or simpler designs.
Very true. Its just that often class A is the simplest design and also usually the least optimized for distortion, or better yet they run tubes in class A because you're already burning so much power what's a little more? Push-pull class A is proably one of the lowest distortion simple designs you can do, as inefficient as it is, since you're cancelling some of the nonlinearity of the most linear region.
Ab historically have been the most common, I called them push/pull amplifiers. Because of the small overlap where both the NPN/PNP transistors are on at the same time effectively shorting each other out but to such a small degree there is very low idle current/heat. The when the signal is applied they provide a very quiet linear response up to their clipping point. I have a SoundCraftsman PM860 Professional mosfet amp that is amazing It will pump 450W into each channel A and B [900W total] at .05THD
Apologies for not including this in the video itself, but the main advantage for class D amps is that they can be made EXTREMELY efficiently, whilst also having extremely low distortion.
Class D amps are always either fully 'on' or fully 'off'. Meaning they're either outputting efficiently or using basically no power. This means that class D amps can sometimes be 90-95% efficient, vs the 40-80% efficiency of class AB and the maximum 50% of class A.
This makes them really suitable for situations where massive power is needed but an enormous class AB amp would not be feasible. Such as for subwoofers or car audio. You hardly need any heat dissipation and so a very powerful class D amp can be made to be extremely compact.
The main disadvantages are that they inherently generate a huge amount of high frequency noise which must be filtered out, and many class D amps are also load dependent, meaning the filter response will change depending on the impedance of the connected load.
For speakers this is not really an issue as you know everything will be roughly 8 Ohms. But for headphones where it could be anywhing from ~12 Ohm to 600 Ohm, that's a more difficult challenge.
For that reason, and since we don't really 'need' the efficiency for headphone amps anyway which don't use much power, class D is rarely seen in headphone amps.
Any class D based on the Universal class D design and especially the modern variations from Hypex, Purifi and such handily beat 98% of A/AB designs in objective and subjective tests. While being significantly more efficient.
Please, excuse me this probably dumb question.
Do integrated DAC/Amp chips commonly found in Bluetooth devices like Qudelix and FiiO's BTR lineup (ES9218/9218), as well as USB dongles (CS43131) employ class D amplifier technology?
I'm surprised by the statement that "everything will be roughly 8 ohms". An 8-ohm speaker will present a very different load at 20hz than 20khz.
@@DJ_Force it's not that different... There will be some variance yes, but in practice your ears will hardly be able to tell much of a difference really. Also there are already some class d implementations that solve that problem.
I loved the explanation.
And complementing that, it is very easy to use Power MOSFET transistors to implement class D amplifiers, and most MOSFETs I have at home are only limited to its package limits, meaning a small TO-220 transistor handling 75 amps easily (that would be 7kW at 100V) without significant heat. I don't use them on amplifiers (all audio amps I have are AB), but they would be a good choice to make a power amp for sub.
If only my EE professors in college explained things even remotely this clearly. Fabulous presentation
Absolutely! I have a EE degree and PhD in digital signal processing and very much enjoyed the explanation....Flappy Bird!
It seems that everyone commenting in these kind of videos had troubles understanding concepts due to having incompetent teachers. Maybe you were just bad students?
@@UntakenNick Think about school as a culling operation. The failures of the profs are hung on the students. My Eco 1 professor was Indian. Took me 2 months to understand "Good Morning Class". I learned Guns and Butter though.
@@robohofo1 Lol, I'm writing my first mock exam on Signal Processing on monday based on the McClellan and Schafer "Signal Processing First" book and I feel a bit overwhelmed by the density and pace of my course. The book is fantastic though. I've even contacted Professor McClellan via E-Mail to ask him, if there is updated coursework for the book, since the included CD is for windows 2000 and he responded in a coupled of days even though he is already emeritus.
@@UntakenNick Now, now, did you expect them to understand concepts by reading technical papers and conducting experiments? They're geniuses, they don't need that!
No one, and I mean no one has explained this better for a person to understand how sound is manipulated by different class amplification. I think it is very easy to understand once you get to the 3rd time watching the video. :)
Learning do be like that. I have to watch these videos over 4-5 times before I really start absorbing what he’s saying.
First 2 to understand, 3rd one for fun🎉
Truly superlative explanation. Really glad I found this before spending any serious money in amps. Class A is clearly the most signal-honest design, though less efficient.
Wow - this was the best explanation I’ve ever seen for amp types 🤯
Glad it helped!
Are you being sarcastic?
@@Grrrr3FKAGrrrrGrrrrGrrrr Not at all... why?
@@Grrrr3FKAGrrrrGrrrrGrrrr lol
Actual gem of a series this was
Nicely done, in my 50+ years of electronics I've not heard a more accurate easily comprehended explanation! I'd say if more had your knack for instruction the world in which we live would have a much higher rate of proficiency. You've actually made the learning process much less boring! I started my learning process when I was 13 YO, you remind me of the fellow I learned from, he was a 75 YO USNavy EW tech!
As a former electronics engineer, this is the best explanation ever! I wish I had a such explanation during my studies 😀
need more of these videos. so tired of everyone that just says "but that is a topuc for another video" but never get another one. educational content to educate the consumers have a much better effect than giving opinions in reviews
This is a great explanation of the main amplifier types, but it glosses over _why_ you'd go to the trouble of using a Class D amplifier since it sounds more complicated to implement.
The answer, of course, is that they can be vastly more efficient than even a well-tuned Class AB. This is because transistors waste extremely little power when they're either in their fully off or fully on states, while a transistor in its linear region acts somewhat like a variable resistor and dumps the waste energy as heat. Since a Class D amplifier only ever has the transistor fully off (where it wastes no power) or fully on (where it also wastes nearly no power since it's extremely low-resistance) there's very low loss overall.
Class A amplifiers are usually 25% efficient or less (somewhat higher is possible with tricks), while AB designs are 40-80% efficient. Class D, on the other hand, is usually 80-90% efficient, with even higher values possible.
As such, Class D amplifiers can be absolutely tiny and, if well designed, still produce shockingly good sound. It took a while for the designs and implementations to get good enough to avoid distortions or other issues inherent with trying to keep the complicated harmonic-filled output under control, but nowadays you can get dozens of watts (or more) out of something the size of a deck of playing cards with very few compromises.
Purists will insist on linear designs, but a well-designed Class D amplifier can produce excellent audio and allows much more compact designs because the reduced waste power. This is especially important where space is at a premium (such as in small apartments) or in portable applications powered by batteries.
Class D is typically used for subwoofer amps. They run cooler, and can be built smaller. Linear response doesn't matter because they run a LPF crossover.
The real life RMS of most speakers is very low, even if it sounds kinda loud. My receiver (plus PC margin for playing back music (lol so prob 1W over idle)) at the wall pulls maybe +12W for silent vs pretty dang loud. Proper RMS 10W is loud as hell. Doesn't seem like 25% efficiency on a class A is a very big deal when IRL RMS is so low unless you are hurting your hearing. Like, oh no my class A amp pulls 50-100W for an hour or two, lol, who gives a shit 😅.
@@chapstickbomber Lower power is fine, if the circuit is super clean. High signal-to-noise ratio and very low THD. But most amps, especially cheaper ones, are not that clean. It's cheaper to make more power, than to build a really clean circuit. Power can compensate for a dirty circuit because it gets loud before it distorts. If you keep it below this threshold, then you won't get the distortion. When you crank it up, you will hear the distortion. On a super clean circuit, you can open it up and it will still remain tight, with less total power.
Is anyone doing GaN amps yet? because a class D amp + GaN transistors seems like you would be able to get to 95 - 98% efficiency and with significantly more compact designs would want a GaN power supply as well at that point or otherwise your amp may end up being smaller than the power supply, also it would enable the transistors run at a much higher frequency making it easier for a filter to filter out the frequencies you don't want.
@@chapstickbomber Well for mobile application yeh that matters if your going to be running on battery and yes if you want a compact design yeh again that matters, halving the power lost to heat means you can probably reduce the size of your amp by ~30%
This is BY FAR the best explanation of this topic I have seen on youtube. Nice job! I found this video first, so I will be watching the DAC video next.
@2:18 The transistor will only switch on when a certain voltage is reached. It requires a threshold voltage for current to start flowing. (same as forward biasing a diode). This is a significant distinction. The moment there is base current, there will be an amplified collector current.
Also, most class A amps used in audio are push-pull class A. They look like class AB, but they are biased so hard that neither transistor ever turns off completely.
Generally, when the amp is is in the topology shown in this video as class A, the amp will be noted as "singled-ended" and it is practically always a tube amp.
This is one of the best instructional videos I have seen in a LONG time!
Really appreciate how you got to the point quickly and didn’t fill the video with fluff
Not a headphone person and more into speakers (sorry!) but this is the clearest explanation I've seen for amplifiers. While I have already known all this it's great
This is honestly the best explanation I have ever seen on this topic. Very easy to understand yet technically detailed. Thank you so much!
The last 9.5 minutes of my life were not wasted. Thank you! I spent the late 80s in electronics engineering school and I just learned more and ( this is the kicker) comprehended what was in this video than I did in college. Kids today have it made if they want to learn anything!
Easily the most clear explanation I've ever seen for the differences. Even at the fast pace the content was entirely clear and easy to grasp. Thank you!
I don't know why i've never seen a better and succinct explanation as this. After watching this, I can now explain this subject to someone else. Thank you.
For those of us with an RF background, it's easier to think of Class D as a modulation scheme, with the ultrasonic carrier being modulated by the baseband audio. It also helps to also know about Class C, with greater efficiency but more distortion. Class D extends Class C to its limit, by greatly increasing both efficiency and distortion, but then filtering out the distortion products, leaving the amplified baseband signal.
seriously wish I had RUclips videos like this when I was taking my EE courses & labs 30 years ago. just incredible
As someone in the matter for decades I found that intro hilarious. NPN refers to the arrangement of semiconductor layers: N-type, P-type, and N-type. N-type is being doped with extra electron, and P-type lacking electron (has a hole instead electron). It's neither "negative" or "positive" in any way. BJT on the other hand stands for Bipolar Junction Transistor. The rest of the video is spot on regarding audio amplifiers.
not neither?
@@EggBastion Thanks.
The N material (silicon doped with pentavalent elements like P. As or Bi) has a surplus of electrons in the conduction band, so calling it "negative" is not incorrect. Likewise the P material has a deficit of electrons in the lattice (being doped with trivalent elements like Ga, In or B) and therefore a net positive ratio of cations to free electrons. They certainly are negative and positive in terms of net electronic charge.
@@karhukivi You just contradicted yourself. Are pentavalent elements more negative than tetravalent elements like silicon? If yes, what happens when you put them together (similar like bonded PN)? Does the current flow by its own, or you must provide external current source like in ordinary conductor? What's with the forward and breakdown voltages, capacity, etc. Are they just a complex or imaginary properties like impedance?
@@DamirUlovec A pentavalent atom like phosphorous replacing a lattice position of quadrivalent silicon has four bonds with the neighbouring four silicon atoms using four electrons, and the fifth electron is free to move through the lattice. That is why the material is negative and conducts electricity better than pure (undoped) silicon. Lots of information on the web if you look for it!
I understand transistors now thanks to your fantastic graphical demonstrations!!! I came here for an explanation of what the differences are between class A, B, AB, & D amps are, and I'm leaving with more knowledge than I set out to learn. Great job!!!
You explained and visualized multiple topics so well to the point where I've re-learned and finally understood everything on a deeper level.
Wow.
I can find these types of videos only at 5AM haha. Keep up the amazing content!
Thanks for posting this as it cleared up a lot of confusion I've had about amps. Now I know that all of my 70's receivers have class A-B amplifier circuits in them!
👏 WOW. Didn't understand a word of this, but that's my problem. You sir have got a GREAT batch of compliments and "positive feedback" (notice the electronic terms?). Seriously, you've got a great bunch of fans. And you're a genius. Makes me realize what a schmuck I am, should've learned all this in school.
Brilliant! For years I have struggled to understand this stuff. This video nails it in under 10 minutes! Thank you GoldenSound!
I'm off to the kitchen to get a frying pan and egg to cook on my Class A amp. In 1963 I learned how valves work. Then came transistors and I learned how they work. I couldn't figure out why electrons travel in a different direction to current. In 1967 I went round a transistor factory and saw how they were made. Then ICs came along and I was totally lost . . . . . till I was aged 56 and passed a technical exam in microprocessor computer systems. As others say this is an excellent presentation of transistor workings.
Lot of people pointing out a lot of details, but wanted to say I already knew and ignored every single one of them personally, but cleared up a few concepts I’ve been struggling with due to the nuances in the details, it’s actually quite helpful to be reminded of what does and doesn’t matter basically. I was over here stressing about nonlinearities and forward and reverse biases for example. And for how class A’s have so much hum and distortion but i really appreciate the design after seeing how simple they are.
Great video! I appreciate that "as fast as possible" doesn't always have to mean "incomprehensible and simplified to the level of being wrong".
This is by far far far the best explanation of all this. And some of the best explanation of any audio/music content I've seen on the whole internet. I am a professional musician and music teacher by the way. Good on you, this is mind-boggling value you're providing.
Well that's by far the best explanation I've seen on this subject. I'm quite new to all this and you've managed to clear things up quite a bit, so thank you!
Dang it! You explained it VERY CLEARLY! Good job! Loved the video!
This is a video posted for 2 weeks when I watch it. It will be the golden standard on this topic. Thanks.
It's the first time I understand the difference in amplifiers, thank you!
I've never seen amplifier class explanation as good as this.
As an electric guitarist who has vague, laymen-consumer understanding of tube amps, this helped me better understand the A and A/B amps I use, and my bass players amp (class D) which is so light and more powerful.
Better if not the best RUclips videos (straight to the point)
Nobody has information without fillers (no TV shows anymore = how its made ...)
____
The visual was clear (could be better for some with an *AUDIO comparison* = actually hearing these limitations)
Wow!!! Fantastic! This must have taken so much thought and work in order to make a complex subject visually simple to follow. It would be great if you could add harmonics to your part two, as I’m trying to get my head around how these all affect what we hear.
That’s never going to happen. You’ll get a ton of trolling from people on audio science review saying you are experiencing placebo. I personally think they need to be told that harmonics do come through better with class a and ab but good luck with that…. They will just ignore you and say the “measurements” can’t prove it so therefore it can’t be true lol
Best explanation about transistors I had seen till now!
Amazing explanation! I never really understood the application of different classes of amplifiers until now!
Simply the best explanation you'll ever see ... brilliantly done (and I used to be a high school teacher so praise where praise is due!)
Great video! Well explained. Just want to add, that as a guitarist I use tube-amps, and NOTHING beats Class A amps, even with all the downsides: price, heat etc. But once you've owned one, you'l never look back. :)
As an explanation of the topic... perfect.
Clear, no dependent questions left obscure, simple but explanatory graphic, conclusive.
Ever thought of making a course for certain high school, college and university professors on how to explain something?
A true Masterclass, in how to articulate and explain a complex topic in simple form! 👍
This actually made these concepts so clear and easy to understand, wow!
That R2R / class A were playing PCars2 tho. Anyway excellently explanation. One of the best nutshell descriptions about the topic I've ever seen
this explanation was so good that i think i almost understand
Thank you very much!! I study in the field of computers and electronics but this is the first time I actually understood how solid state amps work!!
This is a fascinating video, and cleared up a LOT of questions that I’ve had.
Thank you for a clear and straightforward overview of the amplification classes - very interesting.
P.S. I found your narrative a lot nicer to listen to when I slowed the playback by 10%, it sounds far more natural.
In the olden days, when we designed these, the D's were actually called Class-S, for "switched". My first design had a 300KHz 3dB bandwidth and was used as a motor-drive for uranium separtion systems. These motor ran up to 90,000rpm, but had variable speeds. We also used for our Lunar Roving Vehicles, but that motor drive ran at only 17KHz. Some of the frequency limits are imposed not by the electronics, but by the actual load. You cannot run speaker "iron" at 300KHz, since it would melt the local copper windings. Our Induction Heaters are typically down at 20KHz or so (variable depending on Curie temperature), and depending on the losses in the target material.
I've known how A and AB (push pull) amps work for a long time but was not sure about the new fangled class D. This is a very good explanation.
Great video, thank you for that! I love my Class-A HiFi tube amp. Pure simplicity. Just two tubes and a Transformer for each channel.
Some remarks…
A single transistor (single ended) must be class A (or D) to have a useful output. A double (push-pull) transistor can be class A, B, AB or D to be useful.
The configuration doesn’t make it the class, it’s the bias. Push-pull can still be class A.
If you go from class A to class D you roughly double the efficiency, so +3 dB. If you go from a 82 dB to a 102 dB sensitive loudspeaker it’s 100 times more efficient (from 84 to 94 dB 10 times, from 89 to 92 dB 2 times). Using an inefficient driver with an efficient amplifier only produces more efficiently heat into the voice coil which makes the driver even more inefficient!
On top of that a class D amplifier requires an output filter between amplifier and loudspeaker in which you lose power. Just like the crossover filters.
So what’s the point of using class D?
If efficiency is priority you must get rid of passive (high level) crossovers and have 3 or even 4 way (or 3 way and sub) loudspeakers, with efficient drivers. Otherwise most power is consumed in output transistors, voice coils, crossover networks and padding resistors (the whole purpose of a padding resistor is making a driver less efficient!).
I am surprised that no description of the transistor ever mentions resistance. In essence, it is a resistor who's resistance is controlled by the base signal. This is why they get hot, resistors turn current into heat.
Awesome understand in 9 min, with crystal clear, great work 😊
I'm out of words 🤯 This was a fantastic explanation. Thank you so much!
Fascinating and well explained. I never knew how this worked. Makes sense why class D amps don't sound as good now.
Maintaining full digital from capture to just before the speaker simplifies a great many sources of distortion throughout the audio recording, production, transmission and reproduction process, essentially limiting it to the microphone, the ADC, the DAC, and the speakers.
Greatest explanation ever! if only my engineering professors took half of what was explained here!
I've got an exam next week, This video helps tremendously
Thank you. Please continue to make content like this. This is great.
This video is fantastic! I loved the analogy of the flappy bird having a low pass filter!
Hands down, best explanation video ever.
I'm still daily driving a fully serviced (bought new) 1989 Technics SU-600 'New Class A' Integrated Amp... And... LOVING IT! 👌😏 😎🇬🇧
One of the best simpele explanations I have seen!
👍
Wow.....that flappy bird comp was really insightful. Well done.
brilliant explanation !, especially impressed by the D class and the flappy bird, perfect ! i have had fabulous amps for decades and NOW i understand how they work !
Wow! The best explanation i ever heard, but i already knew wich we best for me. Always a Class A amp. Running hot ? YES , Always high current ? YES .. best quality ? YES ! 3x YES . 🎉
3:41Amplifiers in class A are usualy hot when are mute. In this case all current heat a transistor and a sink. But when play music, part od the current goes to the voice coil and in effect don't heat silicon as much as in mute work. A Class A amplifier that plays loudly will run cooler than one that is silent. 3:53 Amplifier in class A can also run in push pull configuration. It's just a matter of the quantity of the quiescent current, BIAS point. This is like in class AB, by increasing the current you go from class B to class AB, but if you increase it more you will enter to class A, then both transistors conduct current throughout all the cycle. And of course with less efficiency. The SE with single transistor can also operate in class B and C, even D, but has huge THD and IM.
Yes, the description of Class A in the video is at best simplistic, at worst incorrect, as if it is single ended you need a capacitor to prevent the bias current being applied to the speaker, which is why most are push pull.
@@rcflighttestengineer5636 Thanks. The crossover distortion for class B (mentioned at 5:03) can also be reduced to a minimum or to negligible with good design.
30 years ago i had a Sony class d car audio amplifier that was 60 watts per channel. That was the clearest, cleanest sound I've ever heard and the amp ran cool at full power. Current draw was also very low.
ok, now i need to know even more about it, class d is even more complex than i expected
You can also do nagative and positive feedback loops by feeding the emitter into the base
Simple explanation. Superb video 🎉🎉
I am listening to you on a class D amp! Sounds great! Peace!
Extremely well explained. I think the only thing that's missing is a qualitative analysis. Like, do class A transistors produce a "better" or somehow more favorable sound than AB? Are they cheaper to make? The video frames AB as a clear superior version to class A (and B), is that the case?
Well done. Fast, but with quality info. Thx!
Class B biasing is typical in vacuum tube amplifiers. The tubes don't have the nonlinearity of transistors at low input levels. Two tubes in push-pull configuration, in theory, have lower distortion than two transistors running Class A-B.
Subbed. Quick, but outstanding explanation!
Excellent presentation specially on the bias side.
You can get BOTH the Triangle Wave generator AND the Comparator in the same tiny IC, the venerable and inexpensive *555 timer/oscillator* chip. You feed the analog signal to be amplified into the Pulse Width Modulation pin. You also need to add 1 external transistor to make the basic oscillator output true square waves, so that at zero modulation the output is symmetrical 50/50 waves. You then amplify that to desired current and voltage with transistors operating with very low losses in _switch mode._
If the PWM function is done with the CMOS version of the 555, you can make the square wave frequency be 2 MHz, which is about 6.64 octaves above 20 kHz. Then it has to be filtered with a low-pass filter, likely an RL filter for the sake of efficiency. To have the output be essentially flat between 20 Hz and 20 kHz, you may want to set the _corner frequency_ of the low-pass filter to be 40 kHz or 5.64 octaves. _That still gives a very respectable 33.8 dB attenuation of the raw 2 MHz signal, when using a simple _*_single-pole low-pass filter._*
If used with voice-coil speakers, some additional filtering will be provided by the speakers themselves, which behave approximately as RL filters. The overall effect is that very little high-frequency current will flow in the speakers, and it will not be reproduced because it is ultrasonic. Meanwhile the very desirable audio frequencies will flow and produce very nice audio outputs.
5:30 one of the phases should be shifted in that graph
Yep! I missed that!
I used to spend an entire semester learning about how amplifiers work in the past and it was definitely not as easy to understand as this series. Fantastic work!
If only you released the videos sooner 😂
You did a really good job with this. Now take a crack at explaining the increasingly popular ZOLT amplifier design, as found in Linear Tube Audio’s excellent sounding product line. Not really output transformer-less (OTL) not really but kind of class-D; that one will be a good challenge for you.
It's all just Lego. Electricity Lego.
Super! The work of a great communicator.
Great video, thank you for the upload on this short, but very informative, series.
Would be great if you could do one on the audio black art that is valve/ tube amps..
Cheers.
fantastic! love the flappy bird approach
I have a sudden urge to play flappy bird
Because current year, it's coming back soon as a fully monetised, pay-to-play abomination, and Dong Nguyen isn't even getting a cut of it.
Hehe
@@circattleof course he isn't. He didn't make the assets. That would be nintindo
There is also Class H that uses several Input Voltages to reduce the Loss, as a Transistor only gets the Voltage-Drop to the next higher Voltage.
Great episode....!!! Kudos and congratulations - great information...!!!
AH. this explained to me why power-transistor PNP and NPN types are often sold in matched-pairs. I wondered about that.
Explained very well 👏
Thank you for explaining transistors to me👌👌
As an audio amplifier designer, I would like to clarify and refine a few points regarding distortion and amplifier classes:
Class A Distortion Characteristics: The statement "Class A can be harder to get low distortion" is misleading. By design, Class A amplifiers inherently have lower distortion due to their continuous operation within the linear region of the output devices. The active device(s) conduct for the entire signal cycle, avoiding crossover and switching distortion entirely. Achieving ultra-low distortion levels (e.g., 0.01% or lower) is relatively straightforward with a well-designed Class A amplifier, thanks to these linear operating characteristics. In contrast, Class AB amplifiers require meticulous design to achieve similar distortion levels.
Class AB and Crossover Distortion: While Class AB amplifiers significantly reduce zero-crossing (or crossover) distortion compared to Class B designs, they cannot eliminate it entirely. This limitation arises because real-world transistors are not ideal, they have non-linearities, and variations in gain. Even with techniques like bias optimization and feedback, some residual crossover distortion often remains, particularly in less expensive or simpler designs.
Very true. Its just that often class A is the simplest design and also usually the least optimized for distortion, or better yet they run tubes in class A because you're already burning so much power what's a little more? Push-pull class A is proably one of the lowest distortion simple designs you can do, as inefficient as it is, since you're cancelling some of the nonlinearity of the most linear region.
Love those type of videos!
You should do a follow-up on all the whacky and rare amp type, like fortnine did for motorcycle engines.
Finally understood. There is a bird in my amp and when I listen too loud it dies. Gotcha
Great explanation. Thanks.
Excellent educational session. well explained.
Ab historically have been the most common, I called them push/pull amplifiers. Because of the small overlap where both the NPN/PNP transistors are on at the same time effectively shorting each other out but to such a small degree there is very low idle current/heat. The when the signal is applied they provide a very quiet linear response up to their clipping point. I have a SoundCraftsman PM860 Professional mosfet amp that is amazing It will pump 450W into each channel A and B [900W total] at .05THD