I once made an subwoofer amplifier using OPA452 and a set of BDX66/67, with a heavy +/- 35V power supply. It must have delivered around 100 watt into my 4 ohm 15 inch speaker.
I made almost the exact same amplifier when I was in college, but I used a '741 op-amp, and I had some resistors to give a bit of bias to the outputs. The biasing resistors helped with the crossover distortion, but decreased the output power a bit. Interesting that great minds think alike though! :-)
I recently had great fun designing and testing a "current dumping" amplifier myself. They sound surprisingly good. Though i did succumb to the urge of trying to increase power to my 4 ohm load. It seemed as though the high voltage type op amps cant deliver the currents required to fulfill the load within the dead band (through the feed-forward resistor). Some can't even source/sink enough current to drive the output transistors outside of the dead band even with Betas of 120. It would be great to see you implement a voltage gain stage in another video.
Hi John. I have had good luck using a tl084 quad op amp with 1 amp as a gain stage feeding another in the current amp stage that drives two complimentary darlingtons (tip120/125) that have a beta of 1000 so the small current of the opamp was enough to feed the transistors. I did not use the resister as you do because I saw no ZC notch. My problem is that to keep the opamp and transistors in their linear range means sacrificing about 6 volts which limits output power.
Hi John, i'm in the process of building this great little dual supply more power amplifier and i have a question. The single supply version previous to this had a voltage divider network to drop the voltage to the 4562 op amp to half the supply rail voltage. I'm wondering if i need to add that voltage divider circuit to this circuit as well ?
If the speakers are not hugely efficient, this amplifier makes perfect sense, basically a class a amplifier+ class B helping it give more power. Great video John. By the way with 2 diodes and 2 resistors at the output transistors will it get rid of that notch?
It's a pure Class B amp. Adding biasing will remove the notch (turning it into a Class AB design), but at the expense of more complexity. This was showing just how simple you could make an amplifier.
Hey John, have you heard of using the TL431 adjustable zener as an amplifier? There's even an example in the datasheet using a transformer like in a tube amp! I wonder how good that'd work.
I've often wondered how this would perform with germanium transistors, as there would be less dead zone. I don't know if anyone even still makes those.
It doesnt matter, because of the feedback loop. But you have to exceed the forward voltage of the LED to light it up. This is 1.8 V and more, up to 2.5 V. Dependent on the efficiency of your speakers this could be allready very loudly.
Hi John! You could use that circuit, which supposingly can never be REALLY good, for a special kind of testing transistors: Plug various types in and watch, which type is coping best with the bad circumstances. This way I found out that the well beloved sanken 1943/5200 are not that good ...
There are several circuit schematics around for the TDA2030...TDA2050, LM1875. Usually they use common emitter booster stages with control resistors in the power leads. Because of the circuit design you can also use higher voltages for the booster stage than for the IC by use of additional zener diodes in the power leads. So you can operate easily with 100 Wrms or more.
@@gkdresden yeah, all the data sheets for small chip amps cover said circuits. I was just saying in reference to this design it would be a simple way of getting a small IC to drive a low impedance load with a very basic design.
The advantage of these "class B" amplifiers (not really class B because the op amp has a quiescent current) is their inherent thermal stability. There are no means required to keep them thermally stable.
@@gkdresden worth noting the basic Class B definition is only used on describing classes for newbies to avoid confusion. The actual Class B definition accepts the use of a small idle current to allow the removal off crossover distortion. It only becomes a Class AB if there's enough quiescent current to run in class A at low powers for a significant amount of the audio signal. A power opamp driving a current dumping output definetley falls into class B. With the additional output in class C.
The point of biasing is to remove the dead zone where neither output transistor is conducting (causes crossover distortion). This circuit relies on negative feedback to move through that zone as quickly as possible, but it's still there.
I bot some tda 2050 amp board from China just to kill my curiosity. The amp use single ps with input and output cap. Using 0-12v as p supply, the bias is almost zero. The sound is as good as the ps cud provide, with better ps caps and chokes the sweeter it sound. Of course 12v cudnt gv good power, but it's a surprise class b cud sound so good. This circuit here is an eye opening also, i hv to try it as small heat sink sud be ok. Then will use small resistant to hv tiny bit of bias to hv 0.2-0.4v between Base emitter and see if the sound wud improve... And of course using a pantode in place of the op amp. In case some one tried it, pls advise.
Crossover distortion - I think I already hear it. Both output transistors remain off near .6 VDC = dead zone. Needs idle current where both Qs remain biased "on". Higher output supply voltage makes that .6V less significant, but Darlington Qs just increase the dead zone, so higher voltage PS ends up forcing a conventional design to drive 3055 transistors. Your amp design is simple but not for clean hi-fi. Listen to upper octave piano notes/chords at low volume level to quickly expose resulting harmonic distortion. No distortion meter is required for that bad news.
More gain, but you increase the dead zone where neither are conducting by another 1.2v and you reduce the output swing, so lower power delivered to speaker.
@@Centar1964 That would likely work pretty well, but it's more complexity and from John's tests it doesn't seem like more gain will improve the performance significantly. Another way of approaching this could be a faster op-amp with higher output current. The higher the slew rate, the smaller the notch, though oscillation might start to be a problem.
Hello. I build this amp with a pair of BD139/140 driving two 2SA1943 and the OP is OPA2604. So far so good, i am getting an output of 30watt in 4 ohms load, with 42vdc single power supply. The amp is working just fine. The question is if the power amp in my configuration can support a 2 ohms load because i want to bridge the 2 channels and drive a 4 ohms speaker and i do not want to blow up my transistors. This is the amplifier: ua-cam.com/users/shortsl-9JVflm6hk
Hi John, just learnt the sad news about Snickers. I'm very sorry for your loss. I'm proud to say that I almost cried.
Love it, you read my mind by upping the power. I have ideas for this circuit. Thanks, one of my favorite channels
Thanks john ,I enjoy your knowledge and videos, super cool stuff, as a novice circuit builder finding your videos a great help and also fun to watch!
Thanks for another great video, John! Would love to see this soldered up.
Hi John,, Thank you for the Low pass filter on your Schematic.
I once made an subwoofer amplifier using OPA452 and a set of BDX66/67, with a heavy +/- 35V power supply. It must have delivered around 100 watt into my 4 ohm 15 inch speaker.
Sounds good to me. I like how simple it is too.
hi john! greetings from chile, i learn a lot with your vids
Really interesting John, thankyou!
I made almost the exact same amplifier when I was in college, but I used a '741 op-amp, and I had some resistors to give a bit of bias to the outputs. The biasing resistors helped with the crossover distortion, but decreased the output power a bit. Interesting that great minds think alike though! :-)
Nice video, I used something similar before as an output buffer for analog outputs.
I recently had great fun designing and testing a "current dumping" amplifier myself. They sound surprisingly good. Though i did succumb to the urge of trying to increase power to my 4 ohm load. It seemed as though the high voltage type op amps cant deliver the currents required to fulfill the load within the dead band (through the feed-forward resistor). Some can't even source/sink enough current to drive the output transistors outside of the dead band even with Betas of 120. It would be great to see you implement a voltage gain stage in another video.
Hi John. I have had good luck using a tl084 quad op amp with 1 amp as a gain stage feeding another in the current amp stage that drives two complimentary darlingtons (tip120/125) that have a beta of 1000 so the small current of the opamp was enough to feed the transistors. I did not use the resister as you do because I saw no ZC notch. My problem is that to keep the opamp and transistors in their linear range means sacrificing about 6 volts which limits output power.
Hi John, i'm in the process of building this great little dual supply more power amplifier and i have a question. The single supply version previous to this had a voltage divider network to drop the voltage to the 4562 op amp to half the supply rail voltage. I'm wondering if i need to add that voltage divider circuit to this circuit as well ?
If it uses a dual supply, it doesn't need the divider network on the input to set the midpoint reference, if I understand what you are asking.
Thanks John. Your the best!
If the speakers are not hugely efficient, this amplifier makes perfect sense, basically a class a amplifier+ class B helping it give more power. Great video John. By the way with 2 diodes and 2 resistors at the output transistors will it get rid of that notch?
It's a pure Class B amp. Adding biasing will remove the notch (turning it into a Class AB design), but at the expense of more complexity. This was showing just how simple you could make an amplifier.
I am planning to build this amp.
I have a question about the schematic, isn't the 100uf capacitor symbol reversed on the negative voltage rail ?
Hey John, have you heard of using the TL431 adjustable zener as an amplifier? There's even an example in the datasheet using a transformer like in a tube amp! I wonder how good that'd work.
Thanks!
Thank You!
nice sharing
Im so curious, what if we change the feed forward R with C? maybe 1 microfarad or whatever value op amp can handle
I've often wondered how this would perform with germanium transistors, as there would be less dead zone. I don't know if anyone even still makes those.
Nope
hi John, very intresting video. I'd like to build such simple amp. Do you think thermal runaway can be a problem?
There's no bias so no thermal runaway.
What filter caps are used here ?
I often put LED with series resistor at the output of an amplifier.
Does it make the output signal goes bad?
It doesnt matter, because of the feedback loop. But you have to exceed the forward voltage of the LED to light it up. This is 1.8 V and more, up to 2.5 V. Dependent on the efficiency of your speakers this could be allready very loudly.
Hi John! You could use that circuit, which supposingly can never be REALLY good, for a special kind of testing transistors: Plug various types in and watch, which type is coping best with the bad circumstances. This way I found out that the well beloved sanken 1943/5200 are not that good ...
Very interesting!
You need one resistor from opamp output to ground , then you will get current feedback too .
Are the capacitor units in uF? I.e. the .1 capacitor on the top rail? Is that 0.1uF?
Thanks
You right. 0.1uF or 100 nanofarad.
Thanks, since the circuit is so simple, maybe it can be very tiny if making a PCB and with SMT components, very portable.
I would really like to build this Higher power amplifier but I cannot find the FJP 5200/1943. Does anyone know who still has these?
The MJE3055/2955 will work okay. Get the ON Semi ones with high gain. The FJP5200 is out of production.
@@JohnAudioTech Thanks John...
I would love to see if you can review a 8 pin audio amplifier called la4525 it can be used in mono and stereo
Sir make a video on distortion figure of quasi complementary configuration.
If you swapped to a power IC like a TDA2050, and used a 1 ohm resistor you'd have a nice current dumping output stage based amplifier.
There are several circuit schematics around for the TDA2030...TDA2050, LM1875. Usually they use common emitter booster stages with control resistors in the power leads. Because of the circuit design you can also use higher voltages for the booster stage than for the IC by use of additional zener diodes in the power leads. So you can operate easily with 100 Wrms or more.
@@gkdresden yeah, all the data sheets for small chip amps cover said circuits. I was just saying in reference to this design it would be a simple way of getting a small IC to drive a low impedance load with a very basic design.
The advantage of these "class B" amplifiers (not really class B because the op amp has a quiescent current) is their inherent thermal stability. There are no means required to keep them thermally stable.
@@gkdresden worth noting the basic Class B definition is only used on describing classes for newbies to avoid confusion. The actual Class B definition accepts the use of a small idle current to allow the removal off crossover distortion. It only becomes a Class AB if there's enough quiescent current to run in class A at low powers for a significant amount of the audio signal.
A power opamp driving a current dumping output definetley falls into class B. With the additional output in class C.
Wait what is the purpose of biasing the amp
The point of biasing is to remove the dead zone where neither output transistor is conducting (causes crossover distortion). This circuit relies on negative feedback to move through that zone as quickly as possible, but it's still there.
fery cool
I bot some tda 2050 amp board from China just to kill my curiosity. The amp use single ps with input and output cap. Using 0-12v as p supply, the bias is almost zero.
The sound is as good as the ps cud provide, with better ps caps and chokes the sweeter it sound.
Of course 12v cudnt gv good power, but it's a surprise class b cud sound so good.
This circuit here is an eye opening also, i hv to try it as small heat sink sud be ok. Then will use small resistant to hv tiny bit of bias to hv 0.2-0.4v between Base emitter and see if the sound wud improve...
And of course using a pantode in place of the op amp.
In case some one tried it, pls advise.
Crossover distortion - I think I already hear it. Both output transistors remain off near .6 VDC = dead zone. Needs idle current where both Qs remain biased "on". Higher output supply voltage makes that .6V less significant, but Darlington Qs just increase the dead zone, so higher voltage PS ends up forcing a conventional design to drive 3055 transistors. Your amp design is simple but not for clean hi-fi.
Listen to upper octave piano notes/chords at low volume level to quickly expose resulting harmonic distortion. No distortion meter is required for that bad news.
Wouldn't darlingtons give ya the gain ya need?
More gain, but you increase the dead zone where neither are conducting by another 1.2v and you reduce the output swing, so lower power delivered to speaker.
@@ferrumignis True...how bout a Sziklai pair then...
@@Centar1964 That would likely work pretty well, but it's more complexity and from John's tests it doesn't seem like more gain will improve the performance significantly.
Another way of approaching this could be a faster op-amp with higher output current. The higher the slew rate, the smaller the notch, though oscillation might start to be a problem.
You mean a Class-B amplifier. And, they've always been bad with crossover distortion.
A Op amp has by definition a negative and a positive input you show neither it can’t work at all
Full schematic is at the end of the video.
Idont think bias makes no more power at all makes sense if you have Voltage and current to put with your amp can handle
Bias is used to reduce distortion, not to make more power.
Hello. I build this amp with a pair of BD139/140 driving two 2SA1943 and the OP is OPA2604. So far so good, i am getting an output of 30watt in 4 ohms load, with 42vdc single power supply. The amp is working just fine. The question is if the power amp in my configuration can support a 2 ohms load because i want to bridge the 2 channels and drive a 4 ohms speaker and i do not want to blow up my transistors. This is the amplifier: ua-cam.com/users/shortsl-9JVflm6hk