I'd swap the current sourcing resistor for a choke. Should help with efficiency a lot. Also will improve the output power (it will double the available voltage swing).
The old car radios used that technique. Paired with a 6 Ohm 6X9 speaker, those radios were loud! They used PNP germanium transistors, so the coil went to ground, and the amplifier was positive ground referenced. That's why they ran hot, because the bias current had to be high enough, for the impedance of the speaker, and at 0 signal, the transistor had nearly the full 12~13 volts across it at ~2.5 amps. That gave a full 24 volt swing(+,-12v) available to the speaker. The large round base (TO36 case style)transistors were mounted on a very large heatsink, which usually got hot. The heat sink had to be big enough, for the transistor dissipating ~30 watts, to stay below 80°C, on a hot day. Germanium transistors have a much lower junction temperature rating,(85°C) than modern Silicon transistors.(usually 150°C) The germanium transistors were in bigger packages, mounted on bigger heatsinks, because they had less temperature swing, than silicon. Those early Class "A" output stages were used in the first radios, to replace tube radios, back in the late '60's. They indeed used a choke, instead of a load resistor, for just the reasons you mention. One transistor could produce the same amplitude, as 4 transistors in a push-pull H-bridge configuration, which replaced the class "A" design, to save power, and allow a much smaller heatsink. A stereo class "A" would draw more power than the headlights! The output stages would put about 12 watts RMS into a 6 Ohm speaker, with the engine running.(about 13.5 volts across the transistor)
@@vincentrobinette1507 Interesting tidbit of history :-) I was inspired by the typical tube class A, but because a transistor can have such low impedance we dont need to use a transformer (or in other words two tightly coupled inductors), we can just use an inductor and couple capacitively ;-)
@@akkudakkupl By replacing R1 in this schematic with a heavy inductor coil, the DC resistance of the inductor coil was low enough, that the speaker can be connected in parallel with the inductor, no output capacitor necessary. The
At the risk of adding complexity, I think the John Linsley Hood class A amp is worth trying where an active load is used in place of the resistor. Nelson Pass made some very favourable comments on this and added some interesting variations to that "1969" design.
Many "hi end" audio brands sell PURE CLASS A amplifiers. Yes, they afre like stoves, pretty inneficient, but they will never suffer from crossover distortion or unmatched transistors on each half wave. Old tube amp in single ende config using a transformer as a load (as well as DC polarization) were pure class A.
This is a lab amp and is a good example of how simple a power amp is but you do need a strong output to handle that much current and not over heat it would have been good the tactics to have you schematic turn brown shaped like the heat sink
Thanku very much, actually in RUclips there are two extremes very high level practical or very high level theory thankyou for the balance for biginers keppem coming sir
Great tutorial. I've been driving myself nuts trying to work out how to make a small practice guitar amp with this circuit. I can plug an iPhone or some other headphone-level source into it and it works, but with a guitar (and a few different amp designs) I either get nothing, low signal, or crazy distortion. Is there a simple buffer or pre-amp circuit that can bolt onto the front of a design like this?
You should make a 2.1 version of this amplifier but with more power, more transistors a heat sink bigger and a cooler fan. I have a power supply of 15V/6A.
I'm curious, have you tried to validate your spice transistor model with the characteristics of the actual transistor in your circuit? With such a simple circuit it could be instructive to confront simulation and actual measurements.
Interesting as usual. But I could not understand how R6 works and the reason why it's there. Should be happy if you'd spend a few words on that. Thanks for the good work.
@John, Thanks for the great video. It is always fun to follow your channel. I have tried the circuit on LTSpice based on the model from On-Semi and managed to get similar result, of course, after many tries and errors to use the model correctly. And I have tried the actual circuit on breadboard too, it works!
Could you do a video on Line level and if there's any standards for that. Maybe some comparisons using preamp music device as an example. Love you videos especially when you do your own amps. Also maybe you could explain how record players connect and why they need ground and how to build that into you amps or preamp
JohnAudioTech sir john you have 1 cap that is not in the schematic can you tell me what value it is? And where is conected?in the schematic are 2 caps but on bredboard 3 thank you
Thanks for sharing the video. Interesting and nice how it sounds. It sounds quite well, although maybe it is not very efficient, but it sounds great. Would it sound good with a guitar? I wonder how it would perform. Maybe another transistor is needed to rise the guitar signal level to line level...
Hi John, can you please share the BDX33B.lib file, probably via Google drive or other means, I want to try the simulation but couldn't find the transistor model, appreciate for your help. I have tried the model from OnSemi, but the result is totally different. In addition, the BF=123.448 is much smaller than the datasheet which claims hFE=2500, I'm confused with those values.
I can confirm that with On's model Vce is at 14v. And the output is neglible. It's a long shot from what is measured on the bench. Either reality is broken or there is something really wrong in the model. And the bdx33b is not on Cordell's model page.
@@copernicofelinis Yeah, my simulation showed 14.135V at Vc and 106.6mA for Ic Another thing I found weird was that current drawn in the video showed 1.25A, if multiplied with the 5W 8 ohm, it would dissipate 10W and I'm not sure if the resistor can withstand it for minutes...
@@copernicofelinis My bad, last night I tried this again and managed to get the simulation worked. The simulation and real circuit have slightly difference in voltages at various points but good enough to come out 1 watt clean output as demonstrated by John. Great work from John, I have to say this again.
@@apang9999 did you use a different model? Because with the model supplied by On and with the values used by John, it doesn't output anything relevant. If I change BF to 2400 and almost all resistors' values I can get more than 1W with about 6Vpp, but distortion is 5.7%. The point is that I have to change everything in the circuit.
Dear John , i constructed the amp . It has a very good sound and it is playing even without a pre-amp. I connected a small tube pre amplifier ,that its more a buffer , and i am very pleased with the result. Although the R1 resistor is heating up very high . I would be gratefull if you give me an advise about that. For the time being the amp is provided with power from a bench power suply. I am thinking of using a ATX (pc power suply) in 12 volts, unfortunately because i dont have a 15 volts transformer. What do you think about that? I am thanking you for your uploads and i am waiting for your reply . Greetings from Greece Athens
Edited to avoid confusion. Yes, with some very minor alteration. In fact it gets even simpler with FETs and they can sound quite amazing too. If you want a very simple super hifi amplifier use a double triode for voltage gain (and stereo) and choose a simple FET output stage (either push pull or single ended like this one) and depending on the power supply voltages chosen and your output topography you can get as many watts as you want with excellent quality and very low component count, maybe ten or so parts per channel plus power supply - in fact shared components make a stereo version have less components per channel count. Edit No, not a stupid question at all.
@@martinda7446 You need feedback for a mosfet, as mosfets are voltage controlled current sources. The output would be heavily distorted as the voltage doesn't correlate linearly with the current.
John thanks but practically this has a lot of power loss and heat generation. What’s the solution ? Which class power amplifier works more efficient? Does an op-amp helps? How?
Hi! I want to make a small powered speaker for my C64 PC and have an old LM 384 5W IC amp available... can you do a video on a good circuit for that? Thanks!
hi John. How was the ajustment of R3 afterall? I am planning to build this amp and use it with a small tube preamp i m having .My basic problem is that i nave not an oscilloscope .I am amateur as you see. waiting for your reply
Well, it's sad that our 1 year journey with the other amplifier just ended in a pile of boards. I was expecting you would make a nice case, put it in and then on your desk where you would enjoy using it.
Cool, but until you showed the preamp feeding the signal to the base of transistor I was wondering how the heck you were creating so much sound from the speakers. I like playing with and experimenting with 1 transistor amps and have not done so for a while. I found in general expecting much more than 1/2 to 1 watt was pretty much wishful thinking and I was wondering if you have or could dream up a simple amp that really pushed the limit. If so I'd sure like to see the video.
Got a question here, the circuit is very much similar to a standard common emitter circuit, however, I found the R1 is in different location, isn't it that R4 should be connected to Vcc?
Jack, R4 is connected that way to stabilize the bias. It's a sort of feedback. Iirc it's called self-bias and it is not that much different from the feedback realized by emitter degeneration.
DC in a speaker will cause the cone to shift away from it's mid-position, which can introduce distortion. Too much DC voltage will cause the voice coil to burn up.
Otherwise you will blow the coil or if you dont reach this point you will heat the coil for no reason putting stress into the speaker. DC is like NO-NO for a speaker
I’m having so much trouble with amplifiers. My resistor values are never right, it’s always super quiet. I don’t understand what I’m doing wrong. Asking for a friend.
Personal experience with many years of SPICE modeling - always gets you “in the ballpark”, never gets you a “grand slam homer”. You still need to prototype, and tweak. But, SPICE gives you a good starting point. Regarding your comment about harmonic distortion, here is a good guide: 0.1% - no human can detect this level in music or speech. 1.0 % - a professional musician might hear it, if they are familiar with the piece being played. 3.0 % - people with “golden ears” will start to hear it. 5.0 % - most people will notice it. 10% - sounds like crap. Clipping, regardless of THD, hurts your ears to hear it.
Bob Pease used to say the soldering iron was his simulator, yeah. But here the simulation shows 100 mA while the measurement screams 1.2 amps. That's not in the same ballpark at all. It's in a different stadium in another state. Something is wrong with On's model. That's why it would be nice to see John's model.
@@davekazoroski6548 yes, but John is using a different model for the BJT. I believe it would be instructive to confront it with the model downloadable from the manufacturer's website. With the latter model Vce is 14 V and the current is 110 mA. Edit: did I just happen to have read your comment the second you've posted it?
In theory there is no difference between theory and practice. In practice there is.
I made this with TIP120 today. It sounds really nice and was fun to build on a breadboard. Thanks!
That was such a chill tune. 🎼
I'm gonna have that goofy song in my head all day.
That song sounds like something from Brazil's music!! It's definitely something from South America lol
I'd swap the current sourcing resistor for a choke. Should help with efficiency a lot. Also will improve the output power (it will double the available voltage swing).
Yes, I did that in an earlier video. I wanted a rudimentary circuit that worked.
The old car radios used that technique. Paired with a 6 Ohm 6X9 speaker, those radios were loud! They used PNP germanium transistors, so the coil went to ground, and the amplifier was positive ground referenced. That's why they ran hot, because the bias current had to be high enough, for the impedance of the speaker, and at 0 signal, the transistor had nearly the full 12~13 volts across it at ~2.5 amps. That gave a full 24 volt swing(+,-12v) available to the speaker. The large round base (TO36 case style)transistors were mounted on a very large heatsink, which usually got hot. The heat sink had to be big enough, for the transistor dissipating ~30 watts, to stay below 80°C, on a hot day. Germanium transistors have a much lower junction temperature rating,(85°C) than modern Silicon transistors.(usually 150°C) The germanium transistors were in bigger packages, mounted on bigger heatsinks, because they had less temperature swing, than silicon.
Those early Class "A" output stages were used in the first radios, to replace tube radios, back in the late '60's. They indeed used a choke, instead of a load resistor, for just the reasons you mention. One transistor could produce the same amplitude, as 4 transistors in a push-pull H-bridge configuration, which replaced the class "A" design, to save power, and allow a much smaller heatsink. A stereo class "A" would draw more power than the headlights! The output stages would put about 12 watts RMS into a 6 Ohm speaker, with the engine running.(about 13.5 volts across the transistor)
@@vincentrobinette1507 Interesting tidbit of history :-) I was inspired by the typical tube class A, but because a transistor can have such low impedance we dont need to use a transformer (or in other words two tightly coupled inductors), we can just use an inductor and couple capacitively ;-)
@@akkudakkupl By replacing R1 in this schematic with a heavy inductor coil, the DC resistance of the inductor coil was low enough, that the speaker can be connected in parallel with the inductor, no output capacitor necessary. The
At the risk of adding complexity, I think the John Linsley Hood class A amp is worth trying where an active load is used in place of the resistor. Nelson Pass made some very favourable comments on this and added some interesting variations to that "1969" design.
I didn't know class A amplifiers could be used as power amplifiers until now. Thanks for posting this video.
Many "hi end" audio brands sell PURE CLASS A amplifiers. Yes, they afre like stoves, pretty inneficient, but they will never suffer from crossover distortion or unmatched transistors on each half wave. Old tube amp in single ende config using a transformer as a load (as well as DC polarization) were pure class A.
Snickers? Great stuff! - Thanks for sharing.
This is a lab amp and is a good example of how simple a power amp is but you do need a strong output to handle that much current and not over heat it would have been good the tactics to have you schematic turn brown shaped like the heat sink
Thanks for your Insights!
It would have been helpful for us newbs to give a brief overview of the amp operation and follow the signal path.
I would like to see you follow this up with a tube based design.
sounds hi-fi
Like a Minang song. Minang is most part of West Sumatera province.
Cheers from Indonesia
try it with a Sziklai and a whopping inductor in place of the collector resistor
I've built a few amps like this but also with another preamp transistor. It sounded ok but always got less output than predicted.
Thanku very much, actually in RUclips there are two extremes very high level practical or very high level theory thankyou for the balance for biginers keppem coming sir
soon, i'll make my own class A, with my filters... hope it worth the effort XD
Sounds great!
Great tutorial. I've been driving myself nuts trying to work out how to make a small practice guitar amp with this circuit. I can plug an iPhone or some other headphone-level source into it and it works, but with a guitar (and a few different amp designs) I either get nothing, low signal, or crazy distortion. Is there a simple buffer or pre-amp circuit that can bolt onto the front of a design like this?
You should make a 2.1 version of this amplifier but with more power, more transistors a heat sink bigger and a cooler fan. I have a power supply of 15V/6A.
I can feel the heat here in NZ....or is it just called summer?
Na. It's the bushfire from Sydney!
I'm curious, have you tried to validate your spice transistor model with the characteristics of the actual transistor in your circuit?
With such a simple circuit it could be instructive to confront simulation and actual measurements.
Hello good video Please tell me what size of twisted wire you are using
Hello good video Please tell me what size gauge of twisted wire you are using
You didn't mention your final bias resistance, although I suspect it will be a little different from circuit to circuit
Good work
Cool amp. Great explanation for newbies. This can double as a toaster while listening to music. LOL.
Back in the days when transistors cost an arm and a leg this would have been state of the art for hobbyists :-)
Interesting as usual. But I could not understand how R6 works and the reason why it's there. Should be happy if you'd spend a few words on that. Thanks for the good work.
For an easy life try a LM317 current source ? 1 ohm = 1.25 amps . It can be bypassed if required with a capacitor. LN317 is fast enough for class A.
Thank you....!!!
I want to make a amp for my ham radio. It has output for an external speaker of about 2 to 4 watts at 8 ohm would like 15 to 20 watt at 4 ohm
Is there 5 watt version of this?? I like it so simple but sound is tube like :)
I like it good starting point. Gain twenty vs two twenty?
@John, Thanks for the great video. It is always fun to follow your channel. I have tried the circuit on LTSpice based on the model from On-Semi and managed to get similar result, of course, after many tries and errors to use the model correctly. And I have tried the actual circuit on breadboard too, it works!
i will try this. thanks
Could you do a video on Line level and if there's any standards for that. Maybe some comparisons using preamp music device as an example. Love you videos especially when you do your own amps. Also maybe you could explain how record players connect and why they need ground and how to build that into you amps or preamp
sounds good, i'm begging to think you can design amps in your sleep :)
I was messing with a design similar to this for a homebrew shortwave hybrid superhet all i can say it's bleeding LOUD.
Could you use a light bulb as a replacement for the 8ohm/5W resistor?
Yes, Nelson Pass does it with one of his amps.
JohnAudioTech sir john you have 1 cap that is not in the schematic can you tell me what value it is? And where is conected?in the schematic are 2 caps but on bredboard 3 thank you
@@azizzable It is across the supply rails. Good practice to have one even though this circuit is probably stable without it.
Sounds great i like it :))
Thanks for sharing the video. Interesting and nice how it sounds. It sounds quite well, although maybe it is not very efficient, but it sounds great. Would it sound good with a guitar? I wonder how it would perform. Maybe another transistor is needed to rise the guitar signal level to line level...
Hi John, can you please share the BDX33B.lib file, probably via Google drive or other means, I want to try the simulation but couldn't find the transistor model, appreciate for your help. I have tried the model from OnSemi, but the result is totally different. In addition, the BF=123.448 is much smaller than the datasheet which claims hFE=2500, I'm confused with those values.
I can confirm that with On's model Vce is at 14v. And the output is neglible. It's a long shot from what is measured on the bench. Either reality is broken or there is something really wrong in the model. And the bdx33b is not on Cordell's model page.
@@copernicofelinis Yeah, my simulation showed 14.135V at Vc and 106.6mA for Ic Another thing I found weird was that current drawn in the video showed 1.25A, if multiplied with the 5W 8 ohm, it would dissipate 10W and I'm not sure if the resistor can withstand it for minutes...
@@copernicofelinis My bad, last night I tried this again and managed to get the simulation worked. The simulation and real circuit have slightly difference in voltages at various points but good enough to come out 1 watt clean output as demonstrated by John. Great work from John, I have to say this again.
@@apang9999 did you use a different model? Because with the model supplied by On and with the values used by John, it doesn't output anything relevant. If I change BF to 2400 and almost all resistors' values I can get more than 1W with about 6Vpp, but distortion is 5.7%. The point is that I have to change everything in the circuit.
Dear John , i constructed the amp . It has a very good sound and it is playing even without a pre-amp. I connected a small tube pre amplifier ,that its more a buffer , and i am very pleased with the result. Although the R1 resistor is heating up very high . I would be gratefull if you give me an advise about that. For the time being the amp is provided with power from a bench power suply. I am thinking of using a ATX (pc power suply) in 12 volts, unfortunately because i dont have a 15 volts transformer. What do you think about that? I am thanking you for your uploads and i am waiting for your reply . Greetings from Greece Athens
Would a mosfet also work as the semiconductor in this circuit?
P.S.: Stupid question I know.
Edited to avoid confusion. Yes, with some very minor alteration.
In fact it gets even simpler with FETs and they can sound quite amazing too. If you want a very simple super hifi amplifier use a double triode for voltage gain (and stereo) and choose a simple FET output stage (either push pull or single ended like this one) and depending on the power supply voltages chosen and your output topography you can get as many watts as you want with excellent quality and very low component count, maybe ten or so parts per channel plus power supply - in fact shared components make a stereo version have less components per channel count.
Edit No, not a stupid question at all.
@@martinda7446 You need feedback for a mosfet, as mosfets are voltage controlled current sources. The output would be heavily distorted as the voltage doesn't correlate linearly with the current.
I would like to know that too. I've been trying to make small guitar amplifiers, and it would be great to have a full FET amplifier.
@@martinda7446 Is there any example schematic for us to visualize that? Thanks for your answer :)
Sure would!
Hi John, have you done a class B (Push pull) with FET's?
THIS WOULD NOT WORK VERY WELL BECAUSE OF CURRENT LIMITATIONS IN FETS.YOU NEED LARGE AMOUNTS OF CURRENT IN A PUSH PULL CONFIGURATIONS.
Push Pull MOSFET to be specific
Would love to see a quasi design and explanation. 🙏
Also maybe a 10 watt light bulb as a current source
@@bixy9347 Actually the light bulb as a current source isn't a bad idea. You get to see fluctuations in bulb brightness as the music plays top
John. I think I missed it, how/what do you set that 10k pot to? Great channel .Cheers.
Can I use this amp to make a guitar amp?
John thanks but practically this has a lot of power loss and heat generation. What’s the solution ? Which class power amplifier works more efficient? Does an op-amp helps? How?
Class AB or B amplifier.
Hi! I want to make a small powered speaker for my C64 PC and have an old LM 384 5W IC amp available... can you do a video on a good circuit for that? Thanks!
hi John. How was the ajustment of R3 afterall? I am planning to build this amp and use it with a small tube preamp i m having .My basic problem is that i nave not an oscilloscope .I am amateur as you see. waiting for your reply
How to make a transformer coupled amplifier
excuse me sir i dont have an osciloscope cand you tell me what output voltage should i have .on output to be clean like yours?
Do you have a number on the gain the circuit came out to?
I need to make one for my IEMs ! What you got ?
Well, it's sad that our 1 year journey with the other amplifier just ended in a pile of boards. I was expecting you would make a nice case, put it in and then on your desk where you would enjoy using it.
I'm certainly going to make myself an amplifier out of it once boards are professionally made. The design stage what is finished.
Cool, but until you showed the preamp feeding the signal to the base of transistor I was wondering how the heck you were creating so much sound from the speakers.
I like playing with and experimenting with 1 transistor amps and have not done so for a while. I found in general expecting much more than 1/2 to 1 watt was pretty much wishful thinking and I was wondering if you have or could dream up a simple amp that really pushed the limit. If so I'd sure like to see the video.
Ive gotten 8.5RMS out of 1 2sc5200 in class a configuration several times
what are the ways to reduce class A distortion ?
Isolate the transistor's nonlinearities from the circuit as much as possible but that means more complexity. There is no free lunch.
@@JohnAudioTech i see, thanks.
Got a question here, the circuit is very much similar to a standard common emitter circuit, however, I found the R1 is in different location, isn't it that R4 should be connected to Vcc?
Jack, R4 is connected that way to stabilize the bias. It's a sort of feedback. Iirc it's called self-bias and it is not that much different from the feedback realized by emitter degeneration.
in simple it is combination of voltage divider and collector fb bias
Could you remove the emitter resistor?
It is necessary in order to set the bias current and operating point.
What brand are those speakers you are using?..
What about the 5W class A amp based on 2x BD131 transistors and 1x BC212L transistors just Google for 5W class A amp.
What is the need to block DC voltage from getting into speakers?
DC in a speaker will cause the cone to shift away from it's mid-position, which can introduce distortion. Too much DC voltage will cause the voice coil to burn up.
Otherwise you will blow the coil or if you dont reach this point you will heat the coil for no reason putting stress into the speaker. DC is like NO-NO for a speaker
I’m having so much trouble with amplifiers. My resistor values are never right, it’s always super quiet. I don’t understand what I’m doing wrong. Asking for a friend.
I used low ohm big resistors and i can get 9 watts rms out of 1 2sc5200
How much Watts???
Personal experience with many years of SPICE modeling - always gets you “in the ballpark”, never gets you a “grand slam homer”. You still need to prototype, and tweak. But, SPICE gives you a good starting point.
Regarding your comment about harmonic distortion, here is a good guide:
0.1% - no human can detect this level in music or speech.
1.0 % - a professional musician might hear it, if they are familiar with the piece being played.
3.0 % - people with “golden ears” will start to hear it.
5.0 % - most people will notice it.
10% - sounds like crap.
Clipping, regardless of THD, hurts your ears to hear it.
Bob Pease used to say the soldering iron was his simulator, yeah. But here the simulation shows 100 mA while the measurement screams 1.2 amps. That's not in the same ballpark at all. It's in a different stadium in another state. Something is wrong with On's model. That's why it would be nice to see John's model.
I thought John said simulation predicted 1 amp, and he was reading 1.2 amps?
If so, he is in the ballpark, his tweaking got the grand slam.
@@davekazoroski6548 yes, but John is using a different model for the BJT. I believe it would be instructive to confront it with the model downloadable from the manufacturer's website. With the latter model Vce is 14 V and the current is 110 mA.
Edit: did I just happen to have read your comment the second you've posted it?
Needs more cat :)
@Charles B. He was sleeping in the front room when I made the video. I figured I give him time off for this one.
Snickers heard the music and beat it.
Never wake up a sleeping cat. He will remember it, usually at 5am.
Sounds like How To Basic
Why build an obsolete design??
obsolete = absolute