triodes have internal V feedback, that's what sets a low mu and gives low plate Z
cascode connection can interrupt the plate V feedback in a triode
a properly biased/fully bypassed screen grid pentode is a tube circuit with little/no feedback when used common cathode - the screen grid can be used in ultralinear mode to tune the amount of feedback
its basically impossible to design useful amplifier circuits without feedback - so its better to recognize the faulty "no feedback" heuristic for the nonsense it is, learn to design with feedback
Last edited:
Cite please. I keep hearing this quoted on the web but it isn't mentioned in any academic text or paper.
I'd go so far as to claim that the two-port behaviour of triode is "feedback" as a major redefinition of the word.
So (for starters) Lynn Olson has it all wrong at 0.3%[url]?
One advantage of se amps for diy is the arts count. They can be really easy to build. Once you've got a 2.5-5k decent opt and a preferably choke input 3-500v ps the world's your oyster, lots of fun to be had trying out different tubes/ topologys.
S
S
You are missing the point of the question question: with the same valves, what will be the difference between PP and SE, beyond extra watts?
PP should remove most of the second, a/ presuming you manage to get it perfectly balanced and b/ presuming your driver stage can cope with the additional loads (as it needs to swing twice as hard).
However, there's a few flies in the ointment
a/ the behaviour of the OPT (Shoog I've still to check out on your comments on Piltron - thanks for the lead as it's counter to everything I've seen and read before)
b/ you now have to add another (distorting) element to do phase splitting - which is adding MORE odd harmonics.
Finally you've used the word "all" - which I've seen other use to imply that a PP pentode amplifier will be "more linear" than a SE 45. Which ain't true
http://www.diyaudio.com/forums/tube...e-internal-negative-feedback.html#post2390464
several replies link references supporting internal feedback being a major design feature of tubz, limiting triode gain, the subject of circuit and tube design innovation to get more gain per tube from the beginning of the vacuum tube era
seems you missed the point altogether - yes CC triodes can have low distortion - fully due to the internal plate V feedback - not as counter evidence
something has to modify the I =k*Vgc^(3/2) Langmuir Child gm relation which would without any other effect result in considerably more distortion in CC amps
Last edited:
Thank you for Stockman 1953 which I'll follow through with interest.
It's also just a model, and not particularly useful one at that. (I've found one cite so far).
For starters, it is completely at odds with what we know (see Bruno Putzeys "the F word" for the latest discussion ) about the effect of feedback on distortion spectra.
So, I repeat - not only is saying that "triodes have internal feedback" is a complete redefinition of the word "feedback" as defined and taught but, as is happening here and elsewhere, to then extrapolate to say that "every practical circuit has feed back in it" remains best described as steaming (internet) horse puckey.
You could dip your toe into the experimental water much more cheaply by doing a mosFET SE and avoid all the transformers except the Mains.
Pass does lots of these.
Listen to some and then decide whether you want to do the expensive experiment.
Pass does lots of these.
Listen to some and then decide whether you want to do the expensive experiment.
No, for the "Burger King lunch" budget it's a waste of time , especially if you don't own speakers which are known good match with this type of amps.
If you really want to know , just buy used SET amp which you can sell without loss .(Preferably an ambitious $5k DIY project which is typically worth $500 on the market😉
Rrgrds
If you really want to know , just buy used SET amp which you can sell without loss .(Preferably an ambitious $5k DIY project which is typically worth $500 on the market😉
Rrgrds
All one has to do is vary the voltage on the plate of a triode to see that it does vary the cathode current. Clearly FEEDBACK. It's the no feedback people who have gone to great lengths to make up their silly religion to hide that. Someone has been reading too much horse puckey on those "just believe" forums.
Though I respect and admire Lynn Olsens work, on this feedback issue I think he is clearly wrong and I am with smoking-amps (and almost all engineers who have thought about this issue) on this one.
Its a form of local feedback. Local feedback tends to have the least effect on the overall character of the final harmonic profile. Increasing the cathode resistor of a triode to introduce local degenerative feedback has almost the same effect as the intrinsic internal feedback of a triode when it comes to harmonics.
Shoog
Its a form of local feedback. Local feedback tends to have the least effect on the overall character of the final harmonic profile. Increasing the cathode resistor of a triode to introduce local degenerative feedback has almost the same effect as the intrinsic internal feedback of a triode when it comes to harmonics.
Shoog
Last edited:
I think the more relevant argument about triode N Fdbk comes up when one talks about whether it is "linear" feedback.
The Child-Langmuir relation shows it (plate V) has a 3/2 power feedback effect on cathode current. Since the input grid also has a 3/2 power effect, the result is "linear" voltage gain when current is held constant. Not so linear when current is allowed to vary. In practice, one tries to keep the current variation small.
When one considers proximity grid wire effects, the grid no longer follows the strict 3/2 power law, so one talks about the mis-tracking as Mu variation. The grid is typically closer to a square power law, like a Mosfet, over parts of its transfer curve in practical devices. Although a frame grid style tube with relaxed (increased) grid to cathode spacing should get one back to 3/2 power law for the grid. Usual Frame G tubes have the spacing as small as practicable though for maximum gm.
The Child-Langmuir relation shows it (plate V) has a 3/2 power feedback effect on cathode current. Since the input grid also has a 3/2 power effect, the result is "linear" voltage gain when current is held constant. Not so linear when current is allowed to vary. In practice, one tries to keep the current variation small.
When one considers proximity grid wire effects, the grid no longer follows the strict 3/2 power law, so one talks about the mis-tracking as Mu variation. The grid is typically closer to a square power law, like a Mosfet, over parts of its transfer curve in practical devices. Although a frame grid style tube with relaxed (increased) grid to cathode spacing should get one back to 3/2 power law for the grid. Usual Frame G tubes have the spacing as small as practicable though for maximum gm.
Last edited:
As they say, a picture is worth thousand words. Are there any FFT measurements to be cited?
Am using leak tl12 plus mono blocks with matching pre amp . All its completly stripped snd rebuilt. The pre amp now has outboard power supply and extra phono for left channel. The speakers are the original sandwich completly overhauled and bananna connectors fitted. The sound quality is such that i am more than happy. A single ended is obviously better, but how much, i wonder? The amps seem to drive anything and system images superbly.
Sent from my iPhone using Tapatalk
Sent from my iPhone using Tapatalk
It really is a boring area to discuss at this stage - since its been worried to death a thousand times before.
Its all not very informative for the person who asked the original question anyway.
If you want to reinvent the wheel then can I suggest you go over to this thread and restart it:
http://www.diyaudio.com/forums/tube...te-resistance-internal-negative-feedback.html
Or read what a real engineer says:
http://ken-gilbert.com/images/pdf/Inherent_FB_inTriodes.pdf
Shoog
Its all not very informative for the person who asked the original question anyway.
If you want to reinvent the wheel then can I suggest you go over to this thread and restart it:
http://www.diyaudio.com/forums/tube...te-resistance-internal-negative-feedback.html
Or read what a real engineer says:
http://ken-gilbert.com/images/pdf/Inherent_FB_inTriodes.pdf
Shoog
Last edited:
Is single ended worth it?
well only if you like music..however PP is nice as well but its not the same..because it has two tubes..😀
Some people say its esoteric nonsense, perhaps they are right..but I have no wish to go back to PP..I wonder why?
My first experience with SE was not good it sounded so warm..2nd harmonic..it was a shock..but as always its how its implemented..
There is good and not so good SE and PP..
I remember years ago pulling out one side of a PPP amp and cringing as I switched on waiting for the mush and guess what at low volume it sounded the same..???? what..yep it sounded the same..I would never advise someone to do this it was a moment of madness.. Curiosity got the best of me..just for a few moments.
Then I thought why am I burning 4 tubes when I am only running at low power.. I needed a gapped PP/SE output Tx..😀
Regards
M. Gregg
well only if you like music..however PP is nice as well but its not the same..because it has two tubes..😀Some people say its esoteric nonsense, perhaps they are right..but I have no wish to go back to PP..I wonder why?
My first experience with SE was not good it sounded so warm..2nd harmonic..it was a shock..but as always its how its implemented..
There is good and not so good SE and PP..
I remember years ago pulling out one side of a PPP amp and cringing as I switched on waiting for the mush and guess what at low volume it sounded the same..???? what..yep it sounded the same..I would never advise someone to do this it was a moment of madness.. Curiosity got the best of me..just for a few moments.
Then I thought why am I burning 4 tubes when I am only running at low power.. I needed a gapped PP/SE output Tx..😀
Regards
M. Gregg
Last edited:
re: feedback.. Why are you all arguing about this.. It's obvious the posters who say they build "no feedback SET's" mean they havent added their own external feedback loop. Everyone who's read this forum or been involved w/ tubes for any lenght of time knows the technical aspect of the inherent feedback of a triode.. But for the sake of conversation.. That's not what they are talking about!
And about the problems w/ no 'added' feedback SE amps, I still go back to magz's statement that it's intoxicating.
Sure, after you drink a glass or 3 of wine, you're not performing at your peak. But there are a lot of people who feel this state is at times "better" than the peak performance sober alternative.
The O.P. didn't ask if a SE amp was in every way technically superior to PP. he asked if it was worth building. And like drining a fine wine, yes it is.
And about the problems w/ no 'added' feedback SE amps, I still go back to magz's statement that it's intoxicating.
Sure, after you drink a glass or 3 of wine, you're not performing at your peak. But there are a lot of people who feel this state is at times "better" than the peak performance sober alternative.
The O.P. didn't ask if a SE amp was in every way technically superior to PP. he asked if it was worth building. And like drining a fine wine, yes it is.
Ask and ye shall receive (attached). What makes triodes unique among active devices is the strong dependence of plate current on Vpk. Just look at the plate characteristics of any triode.
When the signal swings the grid more negative, the diminishing drop across the plate load increases Vpk. This increasing voltage tries to make Ip increase as well. The opposite happens when the signal takes the grid less negative. Since the grid and plate are pulling in opposite directions, that's NFB by definition. As with a NFB system, you sacrifice open loop gain, and decrease the Zo -- just what you see with triodes with their low u-factors and plate resistances.
You can make a triode act more like a pentode if you neutralize the inherent NFB with PFB, and this was done frequently for RF circuits where you wanted the gain, and the increased selectivity.
I agree with Olson about a lot of things, but not this time. There is no escaping NFB. None at all. Every device has some finite rk, e, s. Just because you don't see that resistor on the PCB doesn't mean it's not there. It is, and it causes the same sort of degeneration as an actual, physical, resistor. There is also that internal triode feedback as well. When Olson makes those "no feedback" claims, he's just being sloppy with his terminology. (Then, again, how many people our there are gonna know that?)
It's like your inside my head 😉
I feel somewhat guilty for inadvertently starting this topic. But yes I was talking about a type of feedback loop that I add in the design (global) and not feedback intrinsically already found in Triodes. I figured this would be obvious but I guess it wasn't so I apologize.
- Status
- Not open for further replies.