Hi Kley, BIG BIG BIG THANK YOU for this video. At 11:00 the swap trick ended me 6 years trying to fix my 76' orange amp. I bought it broken and changed a bunch of components but. It was still huming a big 100hz after half on the matser. I asked for help to many people on the net and more, most of the time they didn't even answer. Today I completed the restoration. Now I can bust up my ears with a smile. Thank you !
Nice summary and tests. Feedback is a complicated subject, especially in tube amplifiers where you are enclosing reactive elements like transformers and capacitors inside the loop. Basically the volume drop you're talking about is only occurring in the frequency range where the NFB is "working". Outside of the frequency response of the open loop amplifier, the NFB is low or even zero, so the amp's gain at the low and high frequencies will be very large. This causes low and high frequency oscillations, and IM distortion. In a guitar amp, who cares I guess, but in hi-fi NFB tube amps can cause listener fatigue and smear transients, among other things. Simpler is usually better, and I always prefer to design tube amps without global NFB loops.
Folks asking about solid state amps might take into consideration that many of the preamp circuits do have negative feedback within themselves but not a simple one like we are seeing in pure tube circuits… and they are pretty much just optimizing the gain of that stage… one of the original reasons for neg feedback was for noise…and distortion control.
I just got a 68 bandmaster in a trade. It still has the ground switch hooked up. I was wondering what your thoughts were on using the ground switch for my negative feedback loop
hi,i made experiment on oscilloscope with my amp tube so i got a single end audio and i putted a potentiometer 10k to test but i dont see no différence in the shape of the square waveform the only thing i see is the decrease gain ,so is it normal to see nothing on the shape of the waveform on the scope ,Is the différence occurs only with ears! thanks
Thanks a lot Kley. Very clear and very useful. When you said "a single pole double throw switch" (= SPDT) around 9m30s did you maybe mean a SPST switch? Wouldn't SPST be enough in that application, that is, engaging the 820 ohm resistor or esclude it totally from the circuit ( = no NFB at all)? Or did misunderstand the use of that switch?
I tried adding a NFB to a Fender 5E3 taking a wire from the 8ohm speaker out tip connection to pin 3 of V2 with a 56k resistor in series. I got a loud hum which wasn't there before. What have I done wrong? Its not a loud squeal, so I haven't tried swapping the output transformer leads to the power valves. The large resistor is recommended on many amp forums and by Rob Robinette and Trinity Amps, but I have seen much lower resistances in Princetons etc.
@@KleyDeJong Thanks for the reply. I wasn't sure as everything I've read says loud squeal, and I heard that on my Princeton build and swapped the OT leads to cure it. This is a loud hum so it threw me. I will give swapping the OT leads a try, as I really want to tame the gain so I can use the volume over 2. Currently I have put an alternative lower value cathode bypass cap in the switch circuit which helps a bit.
@@KleyDeJong Swapping the OT leads gave me the squeal I expected so I swapped them back. I found the source of the hum. I had used Cliff jack sockets that are isolated from the chassis instead of Switchcraft which earth themselves. The Cliff jacks work fine until you attach a NFB loop and then they must be earthed along with the OT ground. Earthing the OT ground along with the output jacks cured the hum, got the NFB doing what it should, and from what I can gather safer too. Nice silent operation again and a 'clean' amp setting - well, cleaner.
@@landonespeche7871 No need to take it to a tech, thanks. I have worked it out for myself. There was no wire to ground from the speaker output jacks, which works fine without the neg feedback. But when the neg feedback is connected to the speaker output jack, a ground wire must be connected to the jack also. Attaching the ground wire to the speaker jack circuit has completely cured my problem. I now have a three position neg feedback switch.
I did just that. It sounds great, but also keep in mind that to maintain some kind of damping factor with the speaker, I had to change the 6V6 to triode mode. Output is still around 4 watts. It's so warm and touch-sensitive, I'd never go back to the stock 5F2A circuit. Using a new Jensen P10R.
Good question. I think NFB would be desirable. Because the frequency response is more flattened or hifi then the nature of the pedal can shine through more.
@@KleyDeJong thank you. What I meant was more related to dynamics and headroom. See I've got this twin reverb clone with switchable NFB. I can select various degrees of NFB. And I suspect that on higher NFB settings, the amp adds more compression - my drive pedals don't rise perceived volume as much as on the same pedal settings into lower NFB. Does this make sense?
Hello Kley! Thanks for the video. I have a 1965 Vibrolux and it doesn't break up at all almost. can I double the size of the 820ohm resistor to make the amp break up earlier? should I adjust the bias and the nfb according to specs even when using a higher resistor value?
Is it a non reverb amp? Those circuits are very clean. The cathode resistor value change will be subtle. I would experiment with putting the nfb on a switch to defeat it. I would also check the bias of the power tubes. Good luck!
what are your thoughts about the following options: Keep the feedback loop and install a .01 µF in series with the NBF resistor. This cap will prevent the lower and mid frequencies to go through the NFB loop. Depending on the cap value treble frequencies will be fed back to the phase inverter which cancels out the treble in the main path. The tone gets less harsh and you still have a noticeable effect. You should experiment with different cap values. Start with 0.01µF. Increase the feedback loop resistance value. A good starting point is twice the original resistor value. This reduces the effect of the feedback loop, making the amp break up more when the NFB is engaged. If you make the NBF switchable there will be less volume difference if you use a high resistor value. (Using a .01 µF cap is also recommended to minimize the difference a little bit between NBF on and off).
@@Matan2222222 Interesting. I have not tried the cap in series. I wonder if that would function similarly to a presence control. I really like adjustable NFB. My favorite is to take the stock value and cut it in half and then put a pot in series with that new resistor that is like 10x the value. So if you have 820 ohms stock, then use a 470 ohm resistor and a 5k or 10k pot.
First time someone teaches Negative Feedback in plain rice & beans ! ...and can this same principle be applied to regular Solid State mini guitar amps and improve the distorted (ugly) harshness that these have from factory ?? Thank You so much ! ps. I will watch all your tube video tutorials 100% for sure !
Hi Kley, BIG BIG BIG THANK YOU for this video.
At 11:00 the swap trick ended me 6 years trying to fix my 76' orange amp.
I bought it broken and changed a bunch of components but.
It was still huming a big 100hz after half on the matser.
I asked for help to many people on the net and more, most of the time they didn't even answer.
Today I completed the restoration.
Now I can bust up my ears with a smile.
Thank you !
Hi Kley, the last 5 minutes of this video is a black screen until there is a caption about feedback settings. Just FYI
Max Neg. Feedback: 11:38; 11:41
Min. Neg. Feedback: 11:48; 11:52
Nice summary and tests. Feedback is a complicated subject, especially in tube amplifiers where you are enclosing reactive elements like transformers and capacitors inside the loop.
Basically the volume drop you're talking about is only occurring in the frequency range where the NFB is "working". Outside of the frequency response of the open loop amplifier, the NFB is low or even zero, so the amp's gain at the low and high frequencies will be very large. This causes low and high frequency oscillations, and IM distortion.
In a guitar amp, who cares I guess, but in hi-fi NFB tube amps can cause listener fatigue and smear transients, among other things.
Simpler is usually better, and I always prefer to design tube amps without global NFB loops.
Just what I was looking for. Thank you for posting!
What a great video! This is exactly the information I was looking for. Thanks so much man!
Folks asking about solid state amps might take into consideration that many of the preamp circuits do have negative feedback within themselves but not a simple one like we are seeing in pure tube circuits… and they are pretty much just optimizing the gain of that stage… one of the original reasons for neg feedback was for noise…and distortion control.
Thank you very much. Excellent info.
I just got a 68 bandmaster in a trade. It still has the ground switch hooked up. I was wondering what your thoughts were on using the ground switch for my negative feedback loop
@@bobbielynn808 Excellent idea!
hi,i made experiment on oscilloscope with my amp tube so i got a single end audio and i putted a potentiometer 10k to test but i dont see no différence in the shape of the square waveform the only thing i see is the decrease gain ,so is it normal to see nothing on the shape of the waveform on the scope ,Is the différence occurs only with ears! thanks
Thanks a lot Kley. Very clear and very useful. When you said "a single pole double throw switch" (= SPDT) around 9m30s did you maybe mean a SPST switch? Wouldn't SPST be enough in that application, that is, engaging the 820 ohm resistor or esclude it totally from the circuit ( = no NFB at all)? Or did misunderstand the use of that switch?
Yes that could work fine!
I tried adding a NFB to a Fender 5E3 taking a wire from the 8ohm speaker out tip connection to pin 3 of V2 with a 56k resistor in series. I got a loud hum which wasn't there before. What have I done wrong? Its not a loud squeal, so I haven't tried swapping the output transformer leads to the power valves.
The large resistor is recommended on many amp forums and by Rob Robinette and Trinity Amps, but I have seen much lower resistances in Princetons etc.
Positive feedback can cause different problems than a squeal. I'd try swapping the ot leads.
@@KleyDeJong Thanks for the reply. I wasn't sure as everything I've read says loud squeal, and I heard that on my Princeton build and swapped the OT leads to cure it.
This is a loud hum so it threw me. I will give swapping the OT leads a try, as I really want to tame the gain so I can use the volume over 2.
Currently I have put an alternative lower value cathode bypass cap in the switch circuit which helps a bit.
@@KleyDeJong Swapping the OT leads gave me the squeal I expected so I swapped them back. I found the source of the hum. I had used Cliff jack sockets that are isolated from the chassis instead of Switchcraft which earth themselves. The Cliff jacks work fine until you attach a NFB loop and then they must be earthed along with the OT ground. Earthing the OT ground along with the output jacks cured the hum, got the NFB doing what it should, and from what I can gather safer too. Nice silent operation again and a 'clean' amp setting - well, cleaner.
take it to a tech.
@@landonespeche7871 No need to take it to a tech, thanks. I have worked it out for myself. There was no wire to ground from the speaker output jacks, which works fine without the neg feedback. But when the neg feedback is connected to the speaker output jack, a ground wire must be connected to the jack also. Attaching the ground wire to the speaker jack circuit has completely cured my problem. I now have a three position neg feedback switch.
Thank you for this. How do you think a tweed Princeton w a 10" alnico would sound with no NFB?
I did just that. It sounds great, but also keep in mind that to maintain some kind of damping factor with the speaker, I had to change the 6V6 to triode mode. Output is still around 4 watts.
It's so warm and touch-sensitive, I'd never go back to the stock 5F2A circuit.
Using a new Jensen P10R.
how would a amp with negative feedback respond to drive pedals?
Good question. I think NFB would be desirable. Because the frequency response is more flattened or hifi then the nature of the pedal can shine through more.
@@KleyDeJong thank you. What I meant was more related to dynamics and headroom. See I've got this twin reverb clone with switchable NFB. I can select various degrees of NFB. And I suspect that on higher NFB settings, the amp adds more compression - my drive pedals don't rise perceived volume as much as on the same pedal settings into lower NFB. Does this make sense?
@@filippfekete9420 nfb does change the onset of distortion. It suppresses gain. So yeah that makes sense to me.
Good info, thanks
Key is top!
Top job!!!
Hello Kley! Thanks for the video. I have a 1965 Vibrolux and it doesn't break up at all almost. can I double the size of the 820ohm resistor to make the amp break up earlier? should I adjust the bias and the nfb according to specs even when using a higher resistor value?
Is it a non reverb amp? Those circuits are very clean. The cathode resistor value change will be subtle. I would experiment with putting the nfb on a switch to defeat it. I would also check the bias of the power tubes. Good luck!
@@KleyDeJong its a reverb version. so changing the resistor to a much larger value will not have a huge effect on breakup?
what are your thoughts about the following options:
Keep the feedback loop and install a .01 µF in series with the NBF resistor. This cap will prevent the lower and mid frequencies to go through the NFB loop. Depending on the cap value treble frequencies will be fed back to the phase inverter which cancels out the treble in the main path. The tone gets less harsh and you still have a noticeable effect. You should experiment with different cap values. Start with 0.01µF.
Increase the feedback loop resistance value. A good starting point is twice the original resistor value. This reduces the effect of the feedback loop, making the amp break up more when the NFB is engaged. If you make the NBF switchable there will be less volume difference if you use a high resistor value. (Using a .01 µF cap is also recommended to minimize the difference a little bit between NBF on and off).
@@Matan2222222 Interesting. I have not tried the cap in series. I wonder if that would function similarly to a presence control.
I really like adjustable NFB. My favorite is to take the stock value and cut it in half and then put a pot in series with that new resistor that is like 10x the value. So if you have 820 ohms stock, then use a 470 ohm resistor and a 5k or 10k pot.
@@KleyDeJong thanks again Kley! Appreciate your help!
First time someone teaches Negative Feedback in plain rice & beans !
...and can this same principle be applied to regular Solid State mini guitar amps
and improve the distorted (ugly) harshness that these have from factory ??
Thank You so much !
ps. I will watch all your tube video tutorials 100% for sure !
Should take a look at "uncle doug" channel...
When you are talking about the amp having less gain you are speaking of the power amp section, right?
It applies to all the circuitry within the nfb loop.
Would you be willing to break down a Roland Jazz Chorus schematic?
I'll see if I can find one. I believe it is solid state, so I'm not as familiar.
Without NFB sound comes a bit dry to me.