It’s not easy to explain principles like this in five minutes. I’m sort of amazed how many people are understanding what you have gone through in this short period of time.
There is some confusion in terms as some people consider a push-pull solid state amplifier to be in single ended mode ("mode" causing the difference in terms here) to differentiate it from bridged mode where the load connects between two outputs. Single ended being where the load connects between output and ground. Of course with tubes, single ended usually means one output device.
You can always substitute transistors for tubes when it comes to comparing the architecture of a schematic. The terms remain the same. Bridged also exists with tubes and single ended, even transformer coupled can also be done with one power transistor.
Who sends an email across the globe and waits days or weeks for a response (which may or may not eventuate), when they can do a few minutes research online to find out the answer? People must love listening to Paul's technical sermons everyday. I try not to miss many.
Awesome explaining push push and single. I really just advanced a notch in my self learning of electronics and circuitry. I'm retired diesel mechanic. 61 years old. Trying to learn a new hobby. And finding out it's getting expensive to stock even a little supplies
As an electrical hobbyist this gives me the perfect level of detail and information to understand these concepts. I think the concepts are the same with tube amps (triodes). You have an input stage and a splitter before the power tubes.
Quasi-complementary (both NPN) is a remnant from the early days of transistors when a PNP device didn't exist. Complementary transistors are well enough matched today that the feedback loop takes care of any non-linearity or mis-match between devices. The input stage (differential amplifier), VAS (voltage amplifier), and drive stage are FAR more important then the output stage.
Paul definitely is good at making his videos appear as he is speaking without much prior rehearsing, and that's all good. Like in this case, Paul could have studied the schematics before making the video, looked at schematics while doing his drawings, or even just showed the actual schematics in the video; and it would all have been without the few mistakes in his drawings, but it wouldn't have been half as good.
The amp that he started drawing, the emitter-follower, actually _does not amplify voltage_ (it has an amplification slightly below 1). It is often used for _buffering_ as it has a high input impedance and a low output impedance. It could function as a class-A output amplifier but, as Paul says, it gets very warm as it is not very efficient.
if we ignore all the disadvantages of heat, power consumption, bulkiness, relatively low power in watts, we have to agree that the human ear functions as a single-ended device, because we humans hear other people without interruption, that is, single-ended, we do not have two languages, one producing positive and the other negative sinusoid, so I think that a single-ended amplifier is more faithful than any other way of amplifying music, of course today it doesn't matter at all, because why would you pay 1000 euros for a 10 W single-ended amplifier when you can buy a 100 W for the same money, but if it plays better, that's another matter a couple of sausages, you have to judge for yourself
Thanks paul, I now know what people mean by a push/pull relationship and how it translates to a guitar amp. I think I'll be going for a tube A/B power amp with a modeler.
Correct me if I'm wrong, but of the three topologies he discussed, 1 and 3 are commonly used in low-watt tube amplifiers? Is option 2 more common in modern (solid state) designs?
The BHK outputs are both n-channel mosfets which is good ... but remember these devices are driven by balanced complementary devices further up the line ...
I'm sure Douglas Self would have something to say about the exaggerated crossover distortion and naming that a class B amplifier :) But overall, in the span of 9 minutes, that was a pretty decent explanation of the main difference between single ended and push pull output stages. Although Paul missed to show the needed coupling capacitor that needs to be at the output (or a transformer) to block the needed dc bias voltage of the single ended output stage. And that capacitor is usually frowned upon.
I'm looking for someone who can elucidate on these two concepts without diagrams. The danger with diagrams is it can either illuminate or further confuse the audience depending on the person explaining. For the most part, it confuses because the diagrams distract the audience from REALLY listening with their brains. So, the best option is still to explain without the aid of anything except just talking. I've been in the investment field for more than 30 years and my job is sales and sales management. For my entire life (because of my hearing problem and hence a handicap at meetings), I've always engaged my important clients over a 1-on-1 coffee and used these sessions to speak to them about new investment strategies that we're currently preparing at the white-paper stage. No paper, no pen, not even serviettes (LOL), just coffee and discussion. It forces the sales persons to distill somewhat complicated or esoteric ideas into their essential parts and that first require the sales person himself to have completely understood the concepts himself, the better to then use his own lay language to explain them to his client. So, Paul, maybe you can do another video just talking us through those concepts (no diagrams, of course)?
It may have been more useful to explain the dual NPN example using a quasi-complementary circuit. You have practically done so in your diagram. Also, both single ended and quasi-complementary circuits need to be decoupled, which some might see as a disadvantage over symmetrical complementary.
You could fluid cool the single ended design with a noise-less passive heat exchanger circuit using the heat generated from operation to circulate a coolant through convection coils the same way as a gas bulb refrigerator.
Sonically, push-pull designs are good for rock n' roll while single ended designs are good for more delicate music genres. Weird how solid-state single ended power amp designs tend to stream prodigious amounts of heat in operation and could double as an electric hibachi cooker.
the other use new-class-A where the single-ended-design is used for the output-stage (giving the excellent sound) but the power for the single-ended output-stage is regulated by a separate push-pull-stage. The result is a fine sound without burning exessive energy.
Class D has made incredible improvements over the decades. I say this as somebody who recently built a Nelson Pass Class A power amp, and the current owner of a Class A/B Cambridge Audio integrated amp. Point being that Class D is very quickly catching up, and is often preferable for many consumers.
Cartridges have a very wide frequency range. Some have frequency range from 1 Hz to 100,000 Hz. Vinyl records contain huge amounts of musical information. When these signals are fed to a well designed amp or receiver, the sound is superb. Tapes are close behind but somehow sound more realistic ,punchy bass and warm but you need to know how to use tapes and tape gear.
Wouldn't it be great if that were true? Catridge definitely don't have this type of frequency range. Life for vinyl folk would be pretty good if that were true.
@@Paulmcgowanpsaudio I beg to disagree. Pls check out the specs on some good cartridges. My old vintage Grace F9 frequency response go from 10HZ to over 45,000 Hz and my Supex 900 from 10 to 50,000 for a start. Others have even larger ranges
I'll show my age here. I remember when PNP transistors that would handle much current were very hard to get and most amps just used NPN transistors for outputs.
I ripped some germanium transistors out of an old broken organ, if they don’t work in a fuzz pedal I’m gonna use them for a simple line driver for my guitar.
Yes, more power and higher efficiency. Single ended designs, whether traditional transformer coupled tubes or solid state (ala Nelson Pass), require high efficiency speakers, north of 90db/watt. Usually variations of horn designs or resonant tubes. Very expensive, physically can be quite large (although back loaded folded horns can be as small as conventional floor-standing speakers).
Hi Paul! Why simplify this? The question was directed to you, the master engineer of PS audio, so why not go all in on your explanation? Simple answers are available all over the internet...
If Paul were an egotist and wanted to show off in front of his audience, I expect that's just what he would do. However, fortunately, Paul is a friendly and personable fellow who has set up these videos in the format of an informal chat among friends. Paul is aware that his audience is not made up of an assemblage of Electrical Engineers conducting a Peer Review of each video (although you might think that from the comments section sometimes), but instead, is made up of a wide range of people from those just getting into the hobby, to the mildly curious, to the person who is wanting to understand the fundamentals of how their system works, to the seasoned audiophile with multiple systems and a tube collection who still may not have a degree in Electrical Engineering. Since he keeps his answers simple, straightforward, and concise, he is able to give the most meaningful information to the largest number of people in his audience. A very wise, and friendly, course of action. Let us hope that, when we are Seventy years old, you and I will share the same degree of wisdom that Paul has.
@@Hare_deLune My comment was not meant as criticism at all, on the contrary I admire John who put together all these videos. But you are not all right, many of the videos actually go a bit deeper than this one. Have a look at the lunch videos with his design engineers of speakers and amps. They discuss issues with negative feedback, cone materials etc. That's what I am talking about.
@Lloyd Stout I know. Not sure why you bother answer though. My original question was directed to Paul. I don't really need answers from you or Haradeluna...
@@donpayne1040 You'd have the bottom rail at a negative voltage, top npn conducts a positive current, bottom npn conducts the negative voltage relative to ground.
@@donpayne1040 Well Paul has simplified down to the very basic elements of a push pull design. For a npn transistor the current flow has to be from the collector down to the emitter as per what is shown by the little arrow, therefore the voltage at the collector has to be higher than the emitter for it to conduct current. An npn transistor cannot conduct in the other direction, hence the need for a pnp transistor. So the little circuit Paul sketched out with two npn transistors - it should have had the bottom rrail at a negative voltage and the speaker tapping off point showing the speaker being connected to ground. When the top transistor conducts the positive voltage from the collector to emitter the current flows through the speaker down to ground. When the bottom npn conducts current flows in the opposite direction between a lower voltage and ground, in other words the ground connection is at a higher potential compared to a negative voltage being applied to the emitter for the bottom transistor. In essence what Paul showed was the very basics of a top transistor handling the positive portion of a wave and the bottom transistor handling the negative portion of a wave, this is what moves the voice coil of a speaker in and out, positive voltage pushes the speaker out, negative pulls it in, no output / 0 volts leaves the speaker in the middle, neither in or out - that's what makes the sound you hear from your speaker.
I have an intergated stereo amp with dual mono amp config (of course with their own seperate power supplies) and it's own dedicated "processor" input so it also can be used as a power amp i.e together with a surround reciver. That runs Class A on lower outputs and then switches to Class A/B on higher outputs. It helps keeping the room warm together with my Class A/B surround reciver :P
1) Your overhead fluorescent lighting was reflecting directly onto your whiteboard! 2) Whoever was working your camera doesn’t have a clue how to operate a camera! you need to zoom in on the subject matter watch the video Paul and you realize that your camera person didn’t do that you can’t even see what you’re writing on the whiteboard with your blue felt marker because it’s too far away!
It’s not easy to explain principles like this in five minutes. I’m sort of amazed how many people are understanding what you have gone through in this short period of time.
As Germans say: "Alles hat ein Ende, nur die Wurst hat zwei" - Everything has an end, only the sausage has two.
We have the same saying in Swedish.
@My Man seriously? Huh?
@My Man oh ok.
There is some confusion in terms as some people consider a push-pull solid state amplifier to be in single ended mode ("mode" causing the difference in terms here) to differentiate it from bridged mode where the load connects between two outputs. Single ended being where the load connects between output and ground. Of course with tubes, single ended usually means one output device.
Hello
You can always substitute transistors for tubes when it comes to comparing the architecture of a schematic. The terms remain the same. Bridged also exists with tubes and single ended, even transformer coupled can also be done with one power transistor.
We've had multiple videos covering those topics. It is nice to see something fresh.
@@adamsonlinearrayspeakersha87 NOBODY CARES ABOUT YOUR INANE DRIVEL. Reported.
Thanks for this explanation, Paul. The best I've seen on how class AB amps work.
Who sends an email across the globe and waits days or weeks for a response (which may or may not eventuate), when they can do a few minutes research online to find out the answer?
People must love listening to Paul's technical sermons everyday.
I try not to miss many.
Awesome explaining push push and single. I really just advanced a notch in my self learning of electronics and circuitry. I'm retired diesel mechanic. 61 years old. Trying to learn a new hobby. And finding out it's getting expensive to stock even a little supplies
Awesome, Keith. What an inspiration you are. Thank you.
Very interesting..."
But whats actually the Difference in SOUND...?
Thanks 😆👌🙏
I’ve owned a Pass Labs Aleph 3 (35 watt single-ended class A) amp for many years. I’m thinking of having it buried with me!
I have Aleph J that I built by myself. I think your idea is brilliant!
@@decemberek That’s great. I’ve been thinking of building one of Nelson’s designs myself.
Well, it keeps you warm...
I'm currently a student but the first thing I do when I graduate is building some first watt gear!
As an electrical hobbyist this gives me the perfect level of detail and information to understand these concepts.
I think the concepts are the same with tube amps (triodes). You have an input stage and a splitter before the power tubes.
It was the splitter (or omission thereof when talking about 2x NPN Push-Pull) was where Paul momentary lost the plot.
@@hugobloemers4425 Agreed. Still a good overview for semi techies
Thank you Paul for another fantastic explanation/lesson. You are the best and you've enriched the minds and lives of so many!
my 300B single ended triode tube amp is very hot, puts out only 8w+8w and is pure bliss
Which model are you using? I’m considering building my own based on someone else’s schematics.
Quasi-complementary (both NPN) is a remnant from the early days of transistors when a PNP device didn't exist. Complementary transistors are well enough matched today that the feedback loop takes care of any non-linearity or mis-match between devices. The input stage (differential amplifier), VAS (voltage amplifier), and drive stage are FAR more important then the output stage.
Great video and explanation. Thank you Paul.
When I can't get a push pull, I settle for single ended.
Paul definitely is good at making his videos appear as he is speaking without much prior rehearsing, and that's all good. Like in this case, Paul could have studied the schematics before making the video, looked at schematics while doing his drawings, or even just showed the actual schematics in the video; and it would all have been without the few mistakes in his drawings, but it wouldn't have been half as good.
Here here!
Great explanation, thanks!
Thanks!
The amp that he started drawing, the emitter-follower, actually _does not amplify voltage_ (it has an amplification slightly below 1). It is often used for _buffering_ as it has a high input impedance and a low output impedance.
It could function as a class-A output amplifier but, as Paul says, it gets very warm as it is not very efficient.
That is correct. A buffer or emitter follower amplifies current and passes voltage.
02:27 If that was how my transistor was designed ! No wonder why I get distortion all the time - LMAO
if we ignore all the disadvantages of heat, power consumption, bulkiness, relatively low power in watts, we have to agree that the human ear functions as a single-ended device, because we humans hear other people without interruption, that is, single-ended, we do not have two languages, one producing positive and the other negative sinusoid, so I think that a single-ended amplifier is more faithful than any other way of amplifying music, of course today it doesn't matter at all, because why would you pay 1000 euros for a 10 W single-ended amplifier when you can buy a 100 W for the same money, but if it plays better, that's another matter a couple of sausages, you have to judge for yourself
Thanks paul, I now know what people mean by a push/pull relationship and how it translates to a guitar amp. I think I'll be going for a tube A/B power amp with a modeler.
Correct me if I'm wrong, but of the three topologies he discussed, 1 and 3 are commonly used in low-watt tube amplifiers? Is option 2 more common in modern (solid state) designs?
The BHK outputs are both n-channel mosfets which is good ... but remember these devices are driven by balanced complementary devices further up the line ...
I'm sure Douglas Self would have something to say about the exaggerated crossover distortion and naming that a class B amplifier :)
But overall, in the span of 9 minutes, that was a pretty decent explanation of the main difference between single ended and push pull output stages. Although Paul missed to show the needed coupling capacitor that needs to be at the output (or a transformer) to block the needed dc bias voltage of the single ended output stage. And that capacitor is usually frowned upon.
I'm looking for someone who can elucidate on these two concepts without diagrams.
The danger with diagrams is it can either illuminate or further confuse the audience depending on the person explaining. For the most part, it confuses because the diagrams distract the audience from REALLY listening with their brains.
So, the best option is still to explain without the aid of anything except just talking.
I've been in the investment field for more than 30 years and my job is sales and sales management. For my entire life (because of my hearing problem and hence a handicap at meetings), I've always engaged my important clients over a 1-on-1 coffee and used these sessions to speak to them about new investment strategies that we're currently preparing at the white-paper stage. No paper, no pen, not even serviettes (LOL), just coffee and discussion. It forces the sales persons to distill somewhat complicated or esoteric ideas into their essential parts and that first require the sales person himself to have completely understood the concepts himself, the better to then use his own lay language to explain them to his client.
So, Paul, maybe you can do another video just talking us through those concepts (no diagrams, of course)?
Knoweldge +1 . Thanks for great explanation!
This was a good episode. Even though I know all of that you made it entertaining
We need a do-over on this one Paul.
That helped me better understand my Class A push/pull stereo guitar amp.
It may have been more useful to explain the dual NPN example using a quasi-complementary circuit. You have practically done so in your diagram. Also, both single ended and quasi-complementary circuits need to be decoupled, which some might see as a disadvantage over symmetrical complementary.
You could fluid cool the single ended design with a noise-less passive heat exchanger circuit using the heat generated from operation to circulate a coolant through convection coils the same way as a gas bulb refrigerator.
Can i play a single ended 10watt amp on my borea 3 90db speakers??
Sonically, push-pull designs are good for rock n' roll while single ended designs are good for more delicate music genres. Weird how solid-state single ended power amp designs tend to stream prodigious amounts of heat in operation and could double as an electric hibachi cooker.
Hello
Most amplifiers are either class AB or D today. and both those designs are good for all kinds of music. Also the delicate stuff.
agreed. 300b set tube at 8w is the best for violin, singers, jazz
@@oysteinsoreide4323 Good, not great.
@@mitchellfox3130 What are you meaning? You really wasn't very generous with your wording my friend.
Mr Paul, Please you describe about Class A push-pull amplifier in next episode.
I must be confused: aren’t Class A amplifiers always single-ended?
@@EliasTheHunter I'm confused too.
As 7 of 9 would say, "The design works, but it is inefficient."!
the other use new-class-A where the single-ended-design is used for the output-stage (giving the excellent sound) but the power for the single-ended output-stage is regulated by a separate push-pull-stage. The result is a fine sound without burning exessive energy.
Yes , Nelson Pass amps run hot (class A) and they sound incredible. He hasn't succumbed to the cheap ICE module trend (class D).
Class D has made incredible improvements over the decades. I say this as somebody who recently built a Nelson Pass Class A power amp, and the current owner of a Class A/B Cambridge Audio integrated amp.
Point being that Class D is very quickly catching up, and is often preferable for many consumers.
When going full complementary, try to get matched pairs of transistors.
Good luck with that! Latency will still be an issue.
John pleitz has the patent on all n channel output design
Cartridges have a very wide frequency range. Some have frequency range from 1 Hz to 100,000 Hz. Vinyl records contain huge amounts of musical information. When these signals are fed to a well designed amp or receiver, the sound is superb. Tapes are close behind but somehow sound more realistic ,punchy bass and warm but you need to know how to use tapes and tape gear.
Wouldn't it be great if that were true? Catridge definitely don't have this type of frequency range. Life for vinyl folk would be pretty good if that were true.
@@Paulmcgowanpsaudio I beg to disagree. Pls check out the specs on some good cartridges. My old vintage Grace F9 frequency response go from 10HZ to over 45,000 Hz and my Supex 900 from 10 to 50,000 for a start. Others have even larger ranges
I'll show my age here. I remember when PNP transistors that would handle much current were very hard to get and most amps just used NPN transistors for outputs.
Hi there :)
crossover distortion might be the term you were trying to think of
I ripped some germanium transistors out of an old broken organ, if they don’t work in a fuzz pedal I’m gonna use them for a simple line driver for my guitar.
The holy grail of Ge NPN transistors for guitar use is the AC187 (The PNP complementary is the AC188)
Not easy to get at this point in time.
Thanks Sir for the info
so push pull will result in more power? or its just efficient
Yes, more power and higher efficiency.
Single ended designs, whether traditional transformer coupled tubes or solid state (ala Nelson Pass), require high efficiency speakers, north of 90db/watt. Usually variations of horn designs or resonant tubes. Very expensive, physically can be quite large (although back loaded folded horns can be as small as conventional floor-standing speakers).
the goo oozing out of my ears thanks you
Thank goodness you didn't have to dive into Quasi-complementary.
class A amp ? and A/B class ?
What about class D compared to class AB?
Hi Paul! Why simplify this? The question was directed to you, the master engineer of PS audio, so why not go all in on your explanation? Simple answers are available all over the internet...
If Paul were an egotist and wanted to show off in front of his audience, I expect that's just what he would do.
However, fortunately, Paul is a friendly and personable fellow who has set up these videos in the format of an informal chat among friends.
Paul is aware that his audience is not made up of an assemblage of Electrical Engineers conducting a Peer Review of each video (although you might think that from the comments section sometimes), but instead, is made up of a wide range of people from those just getting into the hobby, to the mildly curious, to the person who is wanting to understand the fundamentals of how their system works, to the seasoned audiophile with multiple systems and a tube collection who still may not have a degree in Electrical Engineering.
Since he keeps his answers simple, straightforward, and concise, he is able to give the most meaningful information to the largest number of people in his audience.
A very wise, and friendly, course of action.
Let us hope that, when we are Seventy years old, you and I will share the same degree of wisdom that Paul has.
@@Hare_deLune My comment was not meant as criticism at all, on the contrary I admire John who put together all these videos. But you are not all right, many of the videos actually go a bit deeper than this one. Have a look at the lunch videos with his design engineers of speakers and amps. They discuss issues with negative feedback, cone materials etc. That's what I am talking about.
@Lloyd Stout I know. Not sure why you bother answer though. My original question was directed to Paul. I don't really need answers from you or Haradeluna...
Single ended circuit is great for Preamp
Thank you!
Thanks dude!!
Got to love 84% efficiency tho
wow. If there was one question, ever, it "would have" been this. Great. I wonder how two npn's (in the same direction) can both push and pull, though?
It's the audio source that decides the push or pull if I understand this right.
@@Projacked1 ...but then they'd be firing at the same time. Ill check a bascom king on YT.
@@donpayne1040 You'd have the bottom rail at a negative voltage, top npn conducts a positive current, bottom npn conducts the negative voltage relative to ground.
@@wazuo8354 Yeeeh...nope. Too much. Too much of the unknown for me.
@@donpayne1040 Well Paul has simplified down to the very basic elements of a push pull design. For a npn transistor the current flow has to be from the collector down to the emitter as per what is shown by the little arrow, therefore the voltage at the collector has to be higher than the emitter for it to conduct current. An npn transistor cannot conduct in the other direction, hence the need for a pnp transistor. So the little circuit Paul sketched out with two npn transistors - it should have had the bottom rrail at a negative voltage and the speaker tapping off point showing the speaker being connected to ground.
When the top transistor conducts the positive voltage from the collector to emitter the current flows through the speaker down to ground. When the bottom npn conducts current flows in the opposite direction between a lower voltage and ground, in other words the ground connection is at a higher potential compared to a negative voltage being applied to the emitter for the bottom transistor.
In essence what Paul showed was the very basics of a top transistor handling the positive portion of a wave and the bottom transistor handling the negative portion of a wave, this is what moves the voice coil of a speaker in and out, positive voltage pushes the speaker out, negative pulls it in, no output / 0 volts leaves the speaker in the middle, neither in or out - that's what makes the sound you hear from your speaker.
Thanks!
So now we know it: PS Audio amplifiers sound better because their output FETs identify them self as N-channel ;)
i still work in triodes old bean
how does this guy wear the same outfit every day
You lost me at "We'll try and simplify it down".
Thought you were going to play a guitar through your bad ass amp, oh well.
Super good
Question is why is an NPN better than PNP can't see any reason why
No homework today😋
I have an intergated stereo amp with dual mono amp config (of course with their own seperate power supplies) and it's own dedicated "processor" input so it also can be used as a power amp i.e together with a surround reciver.
That runs Class A on lower outputs and then switches to Class A/B on higher outputs. It helps keeping the room warm together with my Class A/B surround reciver :P
push pull tubes!
Bono again!
1) Your overhead fluorescent lighting was reflecting directly onto your whiteboard!
2) Whoever was working your camera doesn’t have a clue how to operate a camera! you need to zoom in on the subject matter watch the video Paul and you realize that your camera person didn’t do that you can’t even see what you’re writing on the whiteboard with your blue felt marker because it’s too far away!
But other than that, you thought this was absolutely great, right ??????????
And, turn off auto white-balance, PLEASE!!
Myself and others could see his drawings. Why can't you?
Paul's 'camera person' is clueless because it's an iPhone clamped into a portable stand!
Lol! 😂
1) Who watches these for production value?
2) I have no problem seeing the diagrams.
Understanding them is a different story. LOL
I understanded nothing