There are other ways to amplify *after the concertina. It does possibly pose an issue as it will be another stage wrapped in the NFB loop.
example: https://audiokarma.org/forums/index.php?threads/a-schematic-i-just-stumbled-across-again.975233/
I suspect it can do with some value adjustment, especially on the voltage divider/grid reference for the top triode sections. I have a spice model for it, and have to play with it.
Douglas
Last edited:
Could we say the driver is a PP SRPP lol, I don't understand everything there.
Especially why the 10uf only and why the 20H chokes everywhere, does it mean something?
Thanks for the schematic
Especially why the 10uf only and why the 20H chokes everywhere, does it mean something?
Thanks for the schematic
After consolidating all the data and digesting everything I obtained from multiple THD and Microphone/THD/Noises tests and Impulse response and listening tests....
It appears that the 6BL7 PP amp which clips at around 8 Watts (1.3%THD) has a Bass NOISE profile similar to the best SS amplifier that I could lay my hands onto.
It has probably a limited amperage capacity of around 3A instantaneous VS 8A for the KT88 which has a massive OT, and 15A to 30A for the SS amp, which translate to a slightly lower impulse which is barely visible in the Graphs.
I HAVE NO CLUE WHY the noise in the bass is lower, it is a TRIODE, the KT88 is working in PENTODE,
It has a bit less feedback, the SS has more, the Circuit is simpler : 6sl7 gain in triode (fix bias) + 6sl7 concertina --)
The kt88 has a paraphrase then Cathode followers which could be partly responsible for adding noise? they are 6sn7, gain paraphrase is 6sl7.
Maybe the bigger tubes have more noise?
I could use EL84 and get more power, however they lose detail and are very sweet, the 6bl7 retain much detail without sounding like a triode.
What will happen if I double the number of 6bl7?
What happen to noise if I use a bigger tube kt120? I already measured it and presumably and objectively it is better than the noise of the KT88 (it also does sound better than all the EL and KT tubes). BUT it doesn't have the lower noise of the SS or the 6BL7
To really make an ultimate amp it would be a deceit that the noise floor on real loudspeaker is actually higher than a the best SS and the 6BL7.
The 6BL7 would require an automatic bias to bias each sections instead of biasing an whole tube, it is 8 bias per mono block.
4X6bl7 per mono could reach into 10 watts territory. The big caveat is the waste of heater at 6 Amps per side for a low power & lack of availability of good quality new tubes & low life expectancy with tube failures.
The ultimate addition would be a specialized output transformer which would hopefully enable the tube to reach into the definition territory.
I think I would also abandon tube rectification and use the highest quality low THD NOS 6sl7 that I have here.
In all my amps the sound of the 6BL7 is more musical, and the bass is the best, richest and pleasant.
It appears that the 6BL7 PP amp which clips at around 8 Watts (1.3%THD) has a Bass NOISE profile similar to the best SS amplifier that I could lay my hands onto.
It has probably a limited amperage capacity of around 3A instantaneous VS 8A for the KT88 which has a massive OT, and 15A to 30A for the SS amp, which translate to a slightly lower impulse which is barely visible in the Graphs.
I HAVE NO CLUE WHY the noise in the bass is lower, it is a TRIODE, the KT88 is working in PENTODE,
It has a bit less feedback, the SS has more, the Circuit is simpler : 6sl7 gain in triode (fix bias) + 6sl7 concertina --)
The kt88 has a paraphrase then Cathode followers which could be partly responsible for adding noise? they are 6sn7, gain paraphrase is 6sl7.
Maybe the bigger tubes have more noise?
I could use EL84 and get more power, however they lose detail and are very sweet, the 6bl7 retain much detail without sounding like a triode.
What will happen if I double the number of 6bl7?
What happen to noise if I use a bigger tube kt120? I already measured it and presumably and objectively it is better than the noise of the KT88 (it also does sound better than all the EL and KT tubes). BUT it doesn't have the lower noise of the SS or the 6BL7
To really make an ultimate amp it would be a deceit that the noise floor on real loudspeaker is actually higher than a the best SS and the 6BL7.
The 6BL7 would require an automatic bias to bias each sections instead of biasing an whole tube, it is 8 bias per mono block.
4X6bl7 per mono could reach into 10 watts territory. The big caveat is the waste of heater at 6 Amps per side for a low power & lack of availability of good quality new tubes & low life expectancy with tube failures.
The ultimate addition would be a specialized output transformer which would hopefully enable the tube to reach into the definition territory.
I think I would also abandon tube rectification and use the highest quality low THD NOS 6sl7 that I have here.
In all my amps the sound of the 6BL7 is more musical, and the bass is the best, richest and pleasant.
mmmm After more cogitation...
It is time to upgrade the capacitor section of the power modules in kt88... to 560uf each channel on the Transformer, replacing the 220uf caps.
On top of this I reinstalled the power chokes... I hope this can all lower the bass NOISE.
If it does nothing it is inherent to the Pentode mode.
Then I don't see the point to upscale to KT120, even if I like more the sound of it, which is way superior to KT88..
Does anyone knows if the noise floor in the 40-500hz region is higher for KT120 than for the KT88... I believe it is LOWER a little bit, no idea WHY, that is what I measured...
Now lets see what will happen with the recent modifications... to the kt88 specification of noise !
It is time to upgrade the capacitor section of the power modules in kt88... to 560uf each channel on the Transformer, replacing the 220uf caps.
On top of this I reinstalled the power chokes... I hope this can all lower the bass NOISE.
If it does nothing it is inherent to the Pentode mode.
Then I don't see the point to upscale to KT120, even if I like more the sound of it, which is way superior to KT88..
Does anyone knows if the noise floor in the 40-500hz region is higher for KT120 than for the KT88... I believe it is LOWER a little bit, no idea WHY, that is what I measured...
Now lets see what will happen with the recent modifications... to the kt88 specification of noise !
It did barely nothing to double the power supply and better install the chokes.
The specifications at 8Watts into 8 ohms are as follow:
20hz -0.4dbv THD+N(unweighted) 0.022%
1khz 0dbv THD+N 0.053%
20khz -1.13dbv THD+N 0.028%
OT is rated 80 watts into 30hz -3db ,
It has very little feedback and has kt88 in Tetrode mode, grid bias. bias 55mA, V=470V
GNF around 4db. + 5db feedback through Cathode Feedback dedicated windings, bias resistors 1ohm.
unfortunately the screen voltages are shared through a resistor and a capacitor, which is not ideal for channel separation which is quite good at low frequencies,
but degrades very fast at higher one, at 1khz it is only -45db separation, but I did some listening tests and it is just acceptable (using resistor load in playing channel, max volume music playing, the other channel a loudspeaker I could barely hear a faint noise my ear glued to it)
The specifications at 8Watts into 8 ohms are as follow:
20hz -0.4dbv THD+N(unweighted) 0.022%
1khz 0dbv THD+N 0.053%
20khz -1.13dbv THD+N 0.028%
OT is rated 80 watts into 30hz -3db ,
It has very little feedback and has kt88 in Tetrode mode, grid bias. bias 55mA, V=470V
GNF around 4db. + 5db feedback through Cathode Feedback dedicated windings, bias resistors 1ohm.
unfortunately the screen voltages are shared through a resistor and a capacitor, which is not ideal for channel separation which is quite good at low frequencies,
but degrades very fast at higher one, at 1khz it is only -45db separation, but I did some listening tests and it is just acceptable (using resistor load in playing channel, max volume music playing, the other channel a loudspeaker I could barely hear a faint noise my ear glued to it)
Last edited: