To simplify it for people on why open loop gain is so high, sometimes it needs to be in order to actually pull the voltage high enough for the negative feedback to equal the input voltage. If that gain was limited to 100x, it would be unable to overcome losses/distortion to bring the voltage of the negative feedback in line with the input. It's open loop gain decides how high it's output voltage can be compared to it's input. So the higher the gain, the lower the amplitude of distortion it cannot correct.
The "very high" open-loop gain is at DC, as the input frequency increases the gain drops, reaching unity somewhere in the low MHz range. Something to pay attention to if you need to have input frequencies much above the audio range, especially if your closed-loop gain is pretty high.
Nice demonstration. I suggest that the reason for the small discrepancy between your calculated output voltage before you moved the 100ohm resistor (847mV) and the measured output voltage (824mV) is likely to be caused by the non-zero output impedance of the op-amp. That would add to the 100ohms, increasing the voltage at the voltage divider junction, just as you observed.
I have a few computer speaker amplifier board that have op amps on them. These last few videos have helped me figure out why they're hooked up the way they are, and what inverting, and non inverting means. But having tried to trace/look at what the connections are, makes a little more sence, I'm still not ready to try and make a simple buffer/boost amp, unless I get lucky enough to find a diagram for the chipset I have. (I haven't looked around yet) Thanks for covering the op amp topic.
A while back i had taken a vintage dictation microphone and a cheap 15w guitar amp and fiddled with a single op amp until it sounded great. It was fun. Highly recommend you get one on a breadboard and dive in.
I think the easiest demonstration would be with FFT or Spectrum Analyzer and reactive load. We could see the changes in harmonic spectrum or IMD with changing amount of NF and different load.
Hi john. I like your skills very much. I have tried some circuits from your tutorials and all worked 100%. Can you please make a video on input and output Buffers, regarding active crossovers in relation to audio. I have seen your video on 18db 3rd order crossover and it worked but the buffer thing, balanced input, balanced output etc is confusing me. Please consider my request.
He used the 100 ohm to demonstrate the ability negative feedback has on correcting for effects such as finite/nonzero output impedance. In a realistic case this would be a much more complex circuit, such as a push pull output stage that gets part of its output fed back to the differential pair or op amp. The feedback reduces the output impedance of the honkin' BJTs or Mosfets and corrects for voltage drop across them much like John's example of correcting for the voltage drop across the 100 ohm resistor.
That's tough to answer while keeping it brief. Feedback is a necessary evil, and not all feedback is created equally so in an effort to not throw the baby out with the bathwater let me take a stab at it (no, not a stab at the baby). Not enough feedback can result in excessive distortion and too low of damping factor, whereas excessive feedback can result in oscillations and overshoot of transients. There's a sweet spot. There's more than one kind of feedback. Typically called local and global. Local is often emitter degeneration or base-collector feedback (or similar for FETs). Global is most often, and it is convenient in that it accounts for many effects in the signal path, however it is usually the culprit for causing oscillations. Global is easy to apply, does not take much design skill. Local is much more involved. Negative feedback was a mess when it first came out because of oscillations. Check out the Williamson tube amp topology. In 1954 David Hafler and another designer came up with the Ultralinear Mode of tube operation that blended triode and pentode modes via screen current taps to obtain very linear amp performance without negative feedback. Unfortunately transistors don't have anything comparable to a screen grid or we'd probably see more of ultralinear. But to your point, it does kinda suck the way Stereophile mag and Stereo Review mag etc use the "low negative feedback" as an advertising gimmick. Go figure, buyer beware.
Hi john, wouldn't it be cool to remove the switcher and install a little quiet linear or in the Feeltech? I've seen teardown where the device has plenty of room to do so. Would be cheap and maybe worthwhile?!
I am loving your channel ! Vox AC 30 guitar amps have that sound because they removed the negative feedback loop and that created a pleasant distortion/color . Does it make any sense to remove the negative feedback loop in a solid state discrete amp to emulate the VOX Ac 30 sound ? Does it make any sense to use an op amp with no negative feedback loop to emulate that VoX Ac 30 sound ?
Most solid state amplifiers have so much open loop gain that they would become unusable. For guitar, distortion is better done in a preamp stage designed for that.
A thumbs up is positive feedback!
lol
Nice one !
Very good explanation of how op amps work. Thank you John.
To simplify it for people on why open loop gain is so high, sometimes it needs to be in order to actually pull the voltage high enough for the negative feedback to equal the input voltage. If that gain was limited to 100x, it would be unable to overcome losses/distortion to bring the voltage of the negative feedback in line with the input. It's open loop gain decides how high it's output voltage can be compared to it's input. So the higher the gain, the lower the amplitude of distortion it cannot correct.
Fantastic real scenario demonstration !
At 1:27...
Me: So, why don't you just make ten the loudest?
JohnAudioTech: This one goes to eleven.
As soon as John said "The gain of this amplifier is 11" I was right back to that scene in Spinal Tap.
Great video John - thanks!
Love your work. Will be here to learn more!
Very clear and easy to digest !...cheeres.
1 extra thumb for Snickers, the cat :)
The "very high" open-loop gain is at DC, as the input frequency increases the gain drops, reaching unity somewhere in the low MHz range. Something to pay attention to if you need to have input frequencies much above the audio range, especially if your closed-loop gain is pretty high.
Indeed. I covered that in the other video I mentioned but should have said something here.
I remember reading about this a long time ago. It's one of the properties of Op Amps that makes them so useful
Simple yet effective explanation, Thank you 🙏. Waiting for more... 👍
Nice demonstration. I suggest that the reason for the small discrepancy between your calculated output voltage before you moved the 100ohm resistor (847mV) and the measured output voltage (824mV) is likely to be caused by the non-zero output impedance of the op-amp. That would add to the 100ohms, increasing the voltage at the voltage divider junction, just as you observed.
Thanks, John. Very informative. Snickers said, "meeoo-yawnnnnn."
I was looking for this yesterday, thanks for this awesome video!
I have a few computer speaker amplifier board that have op amps on them. These last few videos have helped me figure out why they're hooked up the way they are, and what inverting, and non inverting means.
But having tried to trace/look at what the connections are, makes a little more sence, I'm still not ready to try and make a simple buffer/boost amp, unless I get lucky enough to find a diagram for the chipset I have. (I haven't looked around yet)
Thanks for covering the op amp topic.
Douglas Blake ,, thanks. I do need to take some time to search up some diagrams for the few op amps I have, and make use of them.
A while back i had taken a vintage dictation microphone and a cheap 15w guitar amp and fiddled with a single op amp until it sounded great. It was fun. Highly recommend you get one on a breadboard and dive in.
From the difference voltage between both inputs one can calculate the open loop gain.
Your feedback is what we're after!
Negative feedback also lowers the output impedance of an amplifier as well as reduces distortion.
John is BACK!
I think the easiest demonstration would be with FFT or Spectrum Analyzer and reactive load. We could see the changes in harmonic spectrum or IMD with changing amount of NF and different load.
Snicker's negative feedback lands in the litter box.
Hadto hook up and op amp and try it myself!
@JohnAudioTech do you have a book with all of the formula for calculating the values in your circuits ???
Hi john. I like your skills very much. I have tried some circuits from your tutorials and all worked 100%. Can you please make a video on input and output Buffers, regarding active crossovers in relation to audio. I have seen your video on 18db 3rd order crossover and it worked but the buffer thing, balanced input, balanced output etc is confusing me. Please consider my request.
Is the ripple you mentioned at 4:20 a result of DDS (direct digital synthesis) from the function generator?
I think it is noise from its power supply.
Maybe a dumb question: so why bother to have 100ohm resistor, anyway?
He used the 100 ohm to demonstrate the ability negative feedback has on correcting for effects such as finite/nonzero output impedance. In a realistic case this would be a much more complex circuit, such as a push pull output stage that gets part of its output fed back to the differential pair or op amp. The feedback reduces the output impedance of the honkin' BJTs or Mosfets and corrects for voltage drop across them much like John's example of correcting for the voltage drop across the 100 ohm resistor.
Why do audiophiles hate negative feedback?
That's tough to answer while keeping it brief. Feedback is a necessary evil, and not all feedback is created equally so in an effort to not throw the baby out with the bathwater let me take a stab at it (no, not a stab at the baby).
Not enough feedback can result in excessive distortion and too low of damping factor, whereas excessive feedback can result in oscillations and overshoot of transients. There's a sweet spot.
There's more than one kind of feedback. Typically called local and global. Local is often emitter degeneration or base-collector feedback (or similar for FETs). Global is most often, and it is convenient in that it accounts for many effects in the signal path, however it is usually the culprit for causing oscillations. Global is easy to apply, does not take much design skill. Local is much more involved.
Negative feedback was a mess when it first came out because of oscillations. Check out the Williamson tube amp topology. In 1954 David Hafler and another designer came up with the Ultralinear Mode of tube operation that blended triode and pentode modes via screen current taps to obtain very linear amp performance without negative feedback. Unfortunately transistors don't have anything comparable to a screen grid or we'd probably see more of ultralinear.
But to your point, it does kinda suck the way Stereophile mag and Stereo Review mag etc use the "low negative feedback" as an advertising gimmick. Go figure, buyer beware.
Hi john, wouldn't it be cool to remove the switcher and install a little quiet linear or in the Feeltech? I've seen teardown where the device has plenty of room to do so. Would be cheap and maybe worthwhile?!
Huh?
I am loving your channel !
Vox AC 30 guitar amps have that sound because they removed the negative feedback loop and that created a pleasant distortion/color . Does it make any sense to remove the negative feedback loop in a solid state discrete amp to emulate the VOX Ac 30 sound ? Does it make any sense to use an op amp with no negative feedback loop to emulate that VoX Ac 30 sound ?
Most solid state amplifiers have so much open loop gain that they would become unusable. For guitar, distortion is better done in a preamp stage designed for that.
@@JohnAudioTech thank you four time
This is why I like zero feedback amps
John please reply.
Do you think single supply 12v is sufficient for TL072 in Baxandall Tone Control?
This would leave you with around 3Vrms of max output. That should be passable with most line level applications.
Off topic but do you spell... Boujero cell?
boucherot
@@JohnAudioTech thank you!
I think Snickers said it all ... ( BORING ! ) ...
Sorry John, but I only give negative feedback. 😉 (@EEVBlog) 👍
My cat is just as uninterested as yours is, I just can't get him to understand ohms law. (or Newtonian mechanics)
But he can handle the effects of gravity better than you can.