I really love this video. You take a simple circuit, show how it performs with "errors" and slowly remove those errors to come up with a pretty good amplifier. Thanks John!
Pioneer sold 2 small amps in the late 70s that used an LM378 as a power opamp to driver unbiased output transistors (TIP31/32). One used a 32V supply for 20Wpch (model GM-20?) and the smaller used a +/_12V split supply for 16Wpch. They had a 22ohm base-to-emitter resister. They were fairly small and worked fine. LM377/378/379s were quite popular and RS sold the 377 for experimenters. The 379 was used in several pinball machines, too. Very hard to find now.
I've done this with a 2N3904/2N3906 (TO92) pair for a small op-amp circuit and needed something to drive a pair of 32 ohm headphones. Worked well, doesn't take up a lot of space or draw much current.
This puristic design is very inspiring. I have allready simulated your idea with LTSpice in order to optimize the design and to test it on breadboard. First of all I added a snubber at the output of the amplifier and a low pass filter (1 kOhm, 3.3 nF) at it input. Instead of the MJE3055 / MJE2955 I will use 2N6488 / 2N6491 because they have a current gain of 250 at 1 A and higher transit frequencies. The base-emitter bypass resistor seems to work better in case of 270 or 330 Ohm. I will test it on a breadboard as soon as possible.
I'd be interested in your results. One mistake I made was not using a lower "feed forward" resistor (as I call it). Since the op amp is providing the current in the +/- .6v range, the op amp has some headroom for a lower value resistor. It would be interesting to model a complex speaker load on the output and see how the amp deals with that.
@@JohnAudioTech the "feed forward resistor" doesn't really feed forward. My simulations show, that the feedback loop of the op amp mainly provides a very fast jump of the base emitter voltage across the resistor in order to open the respective emitter follower very quickly. So it is not really the current of the op amp that drives the speaker. The current is provided by the emitter followers. It means, a sufficiently high slew rate of the op amp is really important, especially at higher audio frequencies and high output voltages.
@@FlyingByDave if you use an op amp with a slew rate of 10V/µs it crosses 1.4 V within 140 ns. The limiting factor is the GBP of the output stage transistors. If you use the combination D44VH10 / D45VH10 you can achieve about 1 µs cross over notch. This is not audible.
Great song off a great album. Mountain was much more than "Mississippi Queen" (as most people know them). They had ties to Cream (one of my favorite bands) and you can hear it in some of their music.
@@JohnAudioTech oh yeah Felix pappalardi Leslie West I think mountain is one of the most underrated bands ever thanks for the reply John always good to talk to you I don't know if you remember me but Im formerly the blown capacitor. Sorry to hear about snickers.
It remembers me as I was 18 and was messing with a bread board creating from scratch and discrete components a class a transistor amp... Class b was too complicated for me at that time and I had no youtube and no teacher... I figured out how to bias at 50% stauration my transistor and feed it with audio and hook the output thru resistors and put a cap to the speaker at the output pin of the transistor.. well, I think it was something like that... about 16 years ago... I was surprised how you could achieve such good quality with unbiased setup. I would eventually like to get a breadboard crash course on class H amp, those used by large PA manufacturer these days to achieve high output and low power dissipation by keeping the transistors at 90% saturation all the time varying the supply voltage just above the full saturation line of the output sinewave... Well, that's how I understand it... Not sure that could fit on a bread board, you tell me... I've got no clue. But to me, power effiency is quite an important thing... and those class H amp can give a lot of output with little power at the input, even for low power application such as an 50 watt rms per chanel computer speaker amp... I've always been into integrating a decent audio power amp to my computer tower so I can connect my speakers directly to it without using a discrete amp... and such class H amp would be suitable to be powered by the internal 12v rail of the power supply... Thanks again for this crash course. Take care!
I tried the same with TDA7294 and it works well. But to be honest TDA7294 does not need a booster transistor to provide extra current, the chip itself is very capable. The only thing pulling back my tda7294 was the power supply. With additional transistors TDA 7294 runs cooler, that definitely an advantage 😀. Thank you for the video John 👍.
Excellent Action Mang.! I look forward to the most brilliant J.A.T.- Amp, And to be able to purchase one board, This Mang is all the way across the Pond in New Zealand, and am willing to scratch the funds together to get one posted when your amp boards meet your final version of J.A.Tech excellence, thanks Mang Sincere regards B-b, Stay Awesome.!
cool! I'm fiddling with this topology myself, but I found easier to use an inverting configuration for single supply operation (less big caps :) and resistors). Don't know if it makes any difference otherwise
I've made an amplifier using this concept. I first tried this about 20 years ago, having been frustrated by that double-diode biasing technique. Also, as I wanted to significantly reduce the amount of current used, just to bias the output components.
I am really looking forward to the discrete amplifier. , I will make a 4ch version for my home.. although i am not sure if using tda7294 would be better or not , I will also make a separate board with 4 transistors in parallel with balancing resistors ofc to drive a subwoofer.. (all of my speakers are 4ohm)
Hi John, have you made any component value changes to the schematic ? I'm looking forward to seeing the next build, when you get further with your testing would you be able to look at my layout design to see if there are any glaring 'no no's that you see before I have any boards made , and if I do have them made I would very much like to send you one also its probally a bit early to ask what price you kit is being pitched at, want to thank you again for the tutorials associated with the various stages of JAT501 design, my board uses some SMD components at the input stage I'm quite pleased with my design aesthetically too and would very much like your opinion ..cheers.
The latest schematic and BOM was posted on Patreon last Spring. I changed the supply bypass caps to 330uf low ESR, and the coil in the output should be in the .7 to 1 uH range but nothing else. I was very happy with initial tests of the prototype that I doubt there will be any other changes.
Hi John , The output caps on Class B amplifiers are not only to block the DC , they are also there to provide the speaker current for the negative swing . One speaker terminal is connected to ground the other is connected to the top of the PNP transistor, the bottom of the PNP transistor is connected to ground and so, without a capacitor, the current path is from Ground To Ground with no potential difference. There could could not be a current flow in that path without the charging and discharging of a capacitor.
@@JohnAudioTech The setting of it at half the supply voltage alone would not cause current to flow through the speaker on the negative swing. It is a necessary presence of the capacitor with, specifically, its charging and discharging, that allows current to flow through the speaker when the PNP transistor conducts.
This amplifier is in class B + C or AB + C. The opamp is class B or AB and a class C transistor. If you connect a resist to the output of the opamp to ground, you will make it work in class A. Great presentation. Fast opamp provides good performance. but it did not limit the bandwidth, which raises concerns about the stability under the capacitive load. typical test is 8ohm in parallel 0.47uF or 4ohm + 1uF and 1khz rectangle.
Gréât job John I realy love your lésions/demo When I see this vidéo I have some questions... Why do more? Why it’s not Hi-Fi How to increase power by adding extra transistors? Why sometime I see 0.48 Ohms résister bêtement NPN & PNP? Regards Davy
I consider HiFi to have lower distortion. Under .1 %. There's no standard to what HiFi is though. This can be a 10 watt amp with more supply voltage. The emitter resistors are for thermal stability and linearity. There is no real need for them in the this circuit.
Hi John I have an Phillips subwoofer and it's amplifier is gone so I want to make an amplifier for it,the ratings of subwoofer speaker are - 16ohm 80watt so do you have any idea what kind of amplifier I make for it?and what number or Voltage and amp transformer for amplifier do I use.
Hi John, thanks for the cool videos. Would you consider making a amp design using the LM3886 ic. The JAT amp series you did was great and i think if you do the same with the LM3886 ic, it should be great as well. All the best.
Hello. I made TDA 2005 amplifier in bridge mode with bootstrap (36dB gain). Circuit was from PDF of this produkt. First chip did't work and the second blows out and now has a big hole on front of them. I didn't do any mistake. I check it about ten times. Do you know the better circuit for TDA2005 in bridge mode?
Hi John, thanks for the video! 👍 Would there be any benefit to using a reverse-biased (snubber/flyback/etc.) diode across the speaker output? Or does it not matter at this power level?
our critters are with Jesus waiting on us in heaven well those of us that make it i understand the love of ones cat i got a few my self family and best friend mush love a tear reading the community page but hes happy now. that may sound silly to some but all i can say is a deep dive and one will find it true Prayers sir
I really love this video. You take a simple circuit, show how it performs with "errors" and slowly remove those errors to come up with a pretty good amplifier. Thanks John!
Pioneer sold 2 small amps in the late 70s that used an LM378 as a power opamp to driver unbiased output transistors (TIP31/32). One used a 32V supply for 20Wpch (model GM-20?) and the smaller used a +/_12V split supply for 16Wpch. They had a 22ohm base-to-emitter resister. They were fairly small and worked fine. LM377/378/379s were quite popular and RS sold the 377 for experimenters. The 379 was used in several pinball machines, too. Very hard to find now.
And I thought I saw and knew everything about amp design..but this is totally new to me...good job creating interest where it died long ago...
These are the kind of things I like to watch. Thanks Jhon.
I'm impressed. It's all about the op-amp secret weapon.
John rocks!
I've done this with a 2N3904/2N3906 (TO92) pair for a small op-amp circuit and needed something to drive a pair of 32 ohm headphones. Worked well, doesn't take up a lot of space or draw much current.
A upload right in time :) got vaccinated yesterday and ended up getting sick, so now I have something nice to watch while laying in bed!
This puristic design is very inspiring. I have allready simulated your idea with LTSpice in order to optimize the design and to test it on breadboard. First of all I added a snubber at the output of the amplifier and a low pass filter (1 kOhm, 3.3 nF) at it input. Instead of the MJE3055 / MJE2955 I will use 2N6488 / 2N6491 because they have a current gain of 250 at 1 A and higher transit frequencies. The base-emitter bypass resistor seems to work better in case of 270 or 330 Ohm. I will test it on a breadboard as soon as possible.
I'd be interested in your results. One mistake I made was not using a lower "feed forward" resistor (as I call it). Since the op amp is providing the current in the +/- .6v range, the op amp has some headroom for a lower value resistor. It would be interesting to model a complex speaker load on the output and see how the amp deals with that.
@@JohnAudioTech the "feed forward resistor" doesn't really feed forward. My simulations show, that the feedback loop of the op amp mainly provides a very fast jump of the base emitter voltage across the resistor in order to open the respective emitter follower very quickly. So it is not really the current of the op amp that drives the speaker. The current is provided by the emitter followers. It means, a sufficiently high slew rate of the op amp is really important, especially at higher audio frequencies and high output voltages.
Hi Gunter.
The resistor does do a useful job, actually. The more current that goes through it, the lower the required op-amp slew rate.
@@FlyingByDave if you use an op amp with a slew rate of 10V/µs it crosses 1.4 V within 140 ns. The limiting factor is the GBP of the output stage transistors. If you use the combination D44VH10 / D45VH10 you can achieve about 1 µs cross over notch. This is not audible.
I love your videos, thanks.
Just putting this out there I would love to see a class H amplifier made
Theme for an imaginary western!!!... great video john!
Great song off a great album. Mountain was much more than "Mississippi Queen" (as most people know them). They had ties to Cream (one of my favorite bands) and you can hear it in some of their music.
@@JohnAudioTech oh yeah Felix pappalardi Leslie West I think mountain is one of the most underrated bands ever thanks for the reply John always good to talk to you I don't know if you remember me but Im formerly the blown capacitor. Sorry to hear about snickers.
Kind of surprised I haven't seen manufacturers do this.
Actually an amplifier called the Quad 405 used this concept back in the 70s. Of course, they went several steps further for a high power amp.
HI, really appreciate these videos.
so awesome ! need to try this !
It remembers me as I was 18 and was messing with a bread board creating from scratch and discrete components a class a transistor amp... Class b was too complicated for me at that time and I had no youtube and no teacher... I figured out how to bias at 50% stauration my transistor and feed it with audio and hook the output thru resistors and put a cap to the speaker at the output pin of the transistor.. well, I think it was something like that... about 16 years ago...
I was surprised how you could achieve such good quality with unbiased setup. I would eventually like to get a breadboard crash course on class H amp, those used by large PA manufacturer these days to achieve high output and low power dissipation by keeping the transistors at 90% saturation all the time varying the supply voltage just above the full saturation line of the output sinewave... Well, that's how I understand it...
Not sure that could fit on a bread board, you tell me... I've got no clue. But to me, power effiency is quite an important thing... and those class H amp can give a lot of output with little power at the input, even for low power application such as an 50 watt rms per chanel computer speaker amp... I've always been into integrating a decent audio power amp to my computer tower so I can connect my speakers directly to it without using a discrete amp... and such class H amp would be suitable to be powered by the internal 12v rail of the power supply...
Thanks again for this crash course. Take care!
Great video.
Nice to hear a quick burst of Mountain !
John, how about doing a video on current feedback, or motional feedback speakers/amplifiers?
Top notch. Thank you.
I tried the same with TDA7294 and it works well. But to be honest TDA7294 does not need a booster transistor to provide extra current, the chip itself is very capable. The only thing pulling back my tda7294 was the power supply. With additional transistors TDA 7294 runs cooler, that definitely an advantage 😀. Thank you for the video John 👍.
🙋🏾♂️
really great & interesting video to watch, thanks John
Hi John Great stuff.. what would the low pass filter be comprised of?
A simple series resistor and shunt capacitor.
Excellent Action Mang.! I look forward to the most brilliant J.A.T.- Amp, And to be able to purchase one board, This Mang is all the way across the Pond in New Zealand, and am willing to scratch the funds together to get one posted when your amp boards meet your final version of J.A.Tech excellence, thanks Mang Sincere regards B-b, Stay Awesome.!
Cool stuff!
I would love to see your schematic with the dual supply
cool! I'm fiddling with this topology myself, but I found easier to use an inverting configuration for single supply operation (less big caps :) and resistors). Don't know if it makes any difference otherwise
I've made an amplifier using this concept. I first tried this about 20 years ago, having been frustrated by that double-diode biasing technique. Also, as I wanted to significantly reduce the amount of current used, just to bias the output components.
Very interesting !!! I'm curious about how bad is the performance with a crappy slow and noisy op amp like LM741.
I am really looking forward to the discrete amplifier. , I will make a 4ch version for my home.. although i am not sure if using tda7294 would be better or not , I will also make a separate board with 4 transistors in parallel with balancing resistors ofc to drive a subwoofer.. (all of my speakers are 4ohm)
the circuit looks very like the QUAD 405 curent dumping amp.
3rd comment Sir, waiting for your every new videos . Really appreciate your hard work. Thank you 😊
Did you mention that the outputs need heatsinks if it is run at higher levels for any time?
Tanks so much sir
Hi John, have you made any component value changes to the schematic ? I'm looking forward to seeing the next build, when you get further with your testing would you be able to look at my layout design to see if there are any glaring 'no no's that you see before I have any boards made , and if I do have them made I would very much like to send you one also its probally a bit early to ask what price you kit is being pitched at, want to thank you again for the tutorials associated with the various stages of JAT501 design, my board uses some SMD components at the input stage I'm quite pleased with my design aesthetically too and would very much like your opinion ..cheers.
The latest schematic and BOM was posted on Patreon last Spring. I changed the supply bypass caps to 330uf low ESR, and the coil in the output should be in the .7 to 1 uH range but nothing else. I was very happy with initial tests of the prototype that I doubt there will be any other changes.
Love to see a similar workout of a 'circlotron' amp ... obviously solid state and not toobes ....
Hi John ,
The output caps on Class B amplifiers are not only to block the DC , they are also there to provide the speaker current for the negative swing . One speaker terminal is connected to ground the other is connected to the top of the PNP transistor, the bottom of the PNP transistor is connected to ground and so, without a capacitor, the current path is from Ground To Ground with no potential difference. There could could not be a current flow in that path without the charging and discharging of a capacitor.
Yes I'm well aware of that. That is why it has to be set at 1/2 the supply voltage.
@@JohnAudioTech The setting of it at half the supply voltage alone would not cause current to flow through the speaker on the negative swing. It is a necessary presence of the capacitor with, specifically, its charging and discharging, that allows current to flow through the speaker when the PNP transistor conducts.
@@woodcoast5026 I know how it works. I didn't want to get into a long explanation of something we both understand.
Shouldn't audio amplifiers be powered by a split-rail power supply? Allows for a simpler design.
@@FlyingByDave Split rail is the best but as I said in the video, some people want single supply for ease in battery use.
3055 2955 pair, will these never die?
This amplifier is in class B + C or AB + C. The opamp is class B or AB and a class C transistor. If you connect a resist to the output of the opamp to ground, you will make it work in class A. Great presentation. Fast opamp provides good performance. but it did not limit the bandwidth, which raises concerns about the stability under the capacitive load. typical test is 8ohm in parallel 0.47uF or 4ohm + 1uF and 1khz rectangle.
nice sharing
Gréât job John
I realy love your lésions/demo
When I see this vidéo I have some questions...
Why do more?
Why it’s not Hi-Fi
How to increase power by adding extra transistors?
Why sometime I see 0.48 Ohms résister bêtement NPN & PNP?
Regards
Davy
I consider HiFi to have lower distortion. Under .1 %. There's no standard to what HiFi is though. This can be a 10 watt amp with more supply voltage. The emitter resistors are for thermal stability and linearity. There is no real need for them in the this circuit.
Pretty cool.
Hi John I have an Phillips subwoofer and it's amplifier is gone so I want to make an amplifier for it,the ratings of subwoofer speaker are - 16ohm 80watt so do you have any idea what kind of amplifier I make for it?and what number or Voltage and amp transformer for amplifier do I use.
Hi John, thanks for the cool videos. Would you consider making a amp design using the LM3886 ic. The JAT amp series you did was great and i think if you do the same with the LM3886 ic, it should be great as well. All the best.
Can u use class ab amplifier ic instead of opamp and increase the watts?
Hello. I made TDA 2005 amplifier in bridge mode with bootstrap (36dB gain). Circuit was from PDF of this produkt. First chip did't work and the second blows out and now has a big hole on front of them. I didn't do any mistake. I check it about ten times. Do you know the better circuit for TDA2005 in bridge mode?
Great, what's the maximum you get from it?
Hi John, thanks for the video! 👍 Would there be any benefit to using a reverse-biased (snubber/flyback/etc.) diode across the speaker output? Or does it not matter at this power level?
I may not be following you. Connecting a diode across the speaker will short half of the AC signal.
@@JohnAudioTech I'm afraid you got me spot on. D'oh! Thanks though!
put me down for a couple boards when you sell them
Is there a substitute tor the LM 4562 Op amp,, I can't find one?
Don't know if John replied, but almost any dual op-amp chip will work OK. Try LM072, LM082, NE5532, LM833, any 45?? chip (4558, 4560....)
1. Did you measure THD at higher frequencies, say 5 kHz or 10 kHz? 2. Could you simplify the layout by using the op amp in the inverting mode?
'It sounds really great... if your ears are screwed up that is. '
Low noise, low THD, low IMD, so why it is not a Hi-Fi amp?
John,where is snicker?
He gone.
Check his community page.
our critters are with Jesus waiting on us in heaven well those of us that make it i understand the love of ones cat i got a few my self family and best friend mush love a tear reading the community page but hes happy now. that may sound silly to some but all i can say is a deep dive and one will find it true Prayers sir