In the case of negatively biased valve driven in positive field the driver solution is even simpler: a normal resistance loaded cathode or source follower.
My 45 push-pull amplifier is biased and loaded as a nominal 5W class A1 (THD less than 1%). However being directly driven by a 6SN7 cathode follower will not clip until over 15W output power (THD=5%) Of course there is no feedback.
My 45 push-pull amplifier is biased and loaded as a nominal 5W class A1 (THD less than 1%). However being directly driven by a 6SN7 cathode follower will not clip until over 15W output power (THD=5%) Of course there is no feedback.
It depends. My speakers are efficient enough not to worry about it.
I don't push it to that level except (probably, I haven't measured it) occasionally in presence of very short peaks with symphonic orchestra but in that case I strongly doubt you really can tell you can hear it because of the very short transient, the extremely soft clipping (which you will loose if apply feedback), and the sustained SPL. Above 92-93 db SPL I don't think you can hear 5% THD (mostly 3rd and 2nd harmonic in my case). It's physiological to the human hearing system.
Distortion figures of the amplifier at lower level are: 0.03% @0.1W, 0.15% @1W, 0.7% @5W. IMD is even much lower......It just sounds better without feedback.
No you don't see. Feedback is ineffective when it starts to clip. Didn't you know? Then I should re-tweak the speakers which are made for the non-zero Zout of the amplifier etc.
There are a lot of things to consider and THD is certainly the minor issue of all....
because fb creates new higher order d.
if majority of value is 2nd hd, i would not bother much. ears are sensitive to odd order.
In fact mine doesn't clip. My speakers are TRUE 94 dB/1w/1m. Average continuous power required (by me) is typically in the range of 30-40 milliwatt. It's louder than you might think! So at least 24-25 dB headroom left (i.e. 10W output) without any hint of clipping. Then you might want to listen to the difference between a hard clipping pentode amp with feedback and one ultra-soft clipping amplifier without. Just in case I also have other amplifiers up to 200W Pout. This is really elementary stuff. I don't need advice, sorry.
Last edited:
as I said, I connected to the cathode g1
(in this way I can get 6W and over)
http://oi67.tinypic.com/sb7p7s.jpg
Last edited:
Dude, it does not.
See post 11; http://www.diyaudio.com/forums/tubes-valves/191161-distortion-spectrum-vs-feedback.html#post2612427
Last edited by a moderator:
True, no feedback sounds better... To me anyways. What sounds "better" is very subjective.
The problem with feedback is that is adds "compression" to the sound. I don't know if that is the correct technical term, but feedback definitely compresses the sound.
So the theory is to build an amp that doesn't need feedback to make it more linear. IE. Design the amp with super low distortion from the get go or build a stage that introduces feedback that will "cancel out" the whole amp's distortion characteristic.
And I'm totally off topic. Sorry.
It's not the feedback causing compression but it does indirectly. Applying feedback will make the Zout lower, increasing the damping factor will introduce compression (i.e. distortion). Less clarity, less details especially in the midrange. This is one of the main arguments which would make a current drive (i.e. constant current source) desirable. A zero feedback triode amp is midway, it's a constant power source, it doesn't have the major issues coming with current drive (which is very well possible but not easy to implement) and it doesn't suffer (or it only suffers very little) the problems coming with low Zout at the same time. "In medio stat virtus".....😉
Running a pentode in this mode is not new at all, was used in the 1930's to get more audio for public address systems. But never run SE, too much distortion.
The distortion in SE is mainly even order, 2H, 4H, 6H & so on. There were no FETs in the 30s so the stage was coupled to a medium or low mu SE or PP triode(s) by an Interstage transformer to deliver the required audio power, in some cases a few watts in a 100W amp.🙂
In this cct D% is the same as for a pentode, made worse by a complex load like a loudspeaker. Without NFB the DF is the same a pentode connected. the speaker goes where it wants. But some people like that, so all is OK.
This mode is Class A2, the tubes screen grid is being used as the control grid.
In PP as used by the ancients they ran these ccts as zero bias, Class B. Very common at that time.😀
Refer to the plate curves on the RCA 46 attached.
The distortion in SE is mainly even order, 2H, 4H, 6H & so on. There were no FETs in the 30s so the stage was coupled to a medium or low mu SE or PP triode(s) by an Interstage transformer to deliver the required audio power, in some cases a few watts in a 100W amp.🙂
In this cct D% is the same as for a pentode, made worse by a complex load like a loudspeaker. Without NFB the DF is the same a pentode connected. the speaker goes where it wants. But some people like that, so all is OK.
This mode is Class A2, the tubes screen grid is being used as the control grid.
In PP as used by the ancients they ran these ccts as zero bias, Class B. Very common at that time.😀
Refer to the plate curves on the RCA 46 attached.