dhaen said:Ricky,
I respectfully suggest that you go away and do some more research. As has been explained in previous posts, what you are contemplating is well beyond a first valve project.
Here are a few thoughts:
Single-ended designs produce a most specific sound signature. Try listening to some before jumping in. (I like it but some don't)
Single-ended amps are more complex that push-pull: Sounds paradoxical? No! For a reasonably performing SE amp, the power supply is more complex than the amp.
Paralleling output valves in SE amps is controversial; Some don't like it as it might cause loss of performance. You need better matched valves than push-pull.
To produce 40 watts SE with a single valve is difficult and expensive. I've done it, using frightening voltages and lots of dosh. I wanted it to drive my existing speakers. It would have been much more sensible to build a 5 or 10 watt amp and replace my speakers. I think I just did it because I could
With respect, I never said that this was my first valve project. I was originally of the opinion that triode-strapped pentodes and tetrodes did not sound like a triode. You all seem to have different opinions on the subject. I was asking which, in your collective opinion, would be the best sonically.
poynton said:
Hi Andy. I wound my own mains Tx but got stock transformers from Danbury Electronics for Opt.
Ricky,
Trying to avoid leaving you here with all the voices intermingling and you have to come to a conclusion .......
My little bit:
All comments have merit, but - phrases such as "is better" or "sounds better" are qualitive. What do figures show? With respect to other members, I am never sure what "optimal use" means. There is no such thing as an optimal triode for audio. Triodes have certain characteristics; and optimality is achieved by the correct application for whatever one wants to achieve. But one "optimality" may not sound the same as another. I am being general I fear, but one often gains with a certain parameter at the cost of another.
I respect previous posts, but I cannot see what the problem is with triode-connected pentodes (and I come from as far back as 1950s). Such an arrangement has a certain set of characteristics, which I cannot find inferior (again, define the term!) to so-called optimum triodes. As one member said, triode development more or less stopped with the advent of pentodes, but I have never seen a commited triode circuit perform significantly better than a good triode-strapped pentode one. (I suspect that some of this notion is intuitive; I stand to be convinced.) As was said by several members, unless each design is optimised for the application, it is meaningless to compare. (E.g. a straight triode circuit will predictably sound better than a pentode at low output - the latter has a high output impedance; inferior loudspeaker damping etc. But pentodes have higher maximum output, and so on. The extrapolation "triodes are better than pentodes" is an unmerited conclusion.)
I myself prefer UL; the "in-betweenness" is far closer to triode operation than pentode except for almost pentode output power. (Studies will make this clear - e.g. fig. 1-1 for a KT88 in the book "An Introduction to Audio Frequency Amplifier Design", from GEC, reprint 1994 by Audio Amateur Press in the UK. Other examples, analyses abound.)
Enough for now - I wish you success! Only do reserve some scepticism.
Trying to avoid leaving you here with all the voices intermingling and you have to come to a conclusion .......
My little bit:
All comments have merit, but - phrases such as "is better" or "sounds better" are qualitive. What do figures show? With respect to other members, I am never sure what "optimal use" means. There is no such thing as an optimal triode for audio. Triodes have certain characteristics; and optimality is achieved by the correct application for whatever one wants to achieve. But one "optimality" may not sound the same as another. I am being general I fear, but one often gains with a certain parameter at the cost of another.
I respect previous posts, but I cannot see what the problem is with triode-connected pentodes (and I come from as far back as 1950s). Such an arrangement has a certain set of characteristics, which I cannot find inferior (again, define the term!) to so-called optimum triodes. As one member said, triode development more or less stopped with the advent of pentodes, but I have never seen a commited triode circuit perform significantly better than a good triode-strapped pentode one. (I suspect that some of this notion is intuitive; I stand to be convinced.) As was said by several members, unless each design is optimised for the application, it is meaningless to compare. (E.g. a straight triode circuit will predictably sound better than a pentode at low output - the latter has a high output impedance; inferior loudspeaker damping etc. But pentodes have higher maximum output, and so on. The extrapolation "triodes are better than pentodes" is an unmerited conclusion.)
I myself prefer UL; the "in-betweenness" is far closer to triode operation than pentode except for almost pentode output power. (Studies will make this clear - e.g. fig. 1-1 for a KT88 in the book "An Introduction to Audio Frequency Amplifier Design", from GEC, reprint 1994 by Audio Amateur Press in the UK. Other examples, analyses abound.)
Enough for now - I wish you success! Only do reserve some scepticism.
I thank everyone for theit input on this thread.
Firstly, this isn't my first outing with valves. There have been pentode push-pull designs before this. All of your comments have, in fact, re-iterated what my friends in the high end HiFi retail sector have said and that is the fact that it comes down to personal taste.
I have a number of friends in this field that point out to me that although they prefer triodes, the Quad 22's were tetrode ( KT66) push-pull ultra-linear and, driven properly, give outstanding results although I believe that it was in fact optimised to drive the ESL57 loudspeaker but, nevertheless, Peter Walker was a very clever man.
I really do welcome the debate so far. I, and many others can learn from all of the valuable input you have all given.
Thank you very very much indeed.
Ricky.
Firstly, this isn't my first outing with valves. There have been pentode push-pull designs before this. All of your comments have, in fact, re-iterated what my friends in the high end HiFi retail sector have said and that is the fact that it comes down to personal taste.
I have a number of friends in this field that point out to me that although they prefer triodes, the Quad 22's were tetrode ( KT66) push-pull ultra-linear and, driven properly, give outstanding results although I believe that it was in fact optimised to drive the ESL57 loudspeaker but, nevertheless, Peter Walker was a very clever man.
I really do welcome the debate so far. I, and many others can learn from all of the valuable input you have all given.
Thank you very very much indeed.
Ricky.
Ricky,
I have had my say, but just regarding the Quad: The Quad 22 was actually the stereo pre-amplifier. I believe you are referring to the Quad II power amplifier.
If so, this was indeed an innovative design. Just one thing to note: The output stage was basically UL, but the OPT primary portion between G2 and B+ was transferred to the cathode side of each KT66, thus there was in addition some local feedback. That is what made this design better than other output stages. In that sense one cannot compare with normal UL or anything else. It did require a special output transformer. I have understood that lately Lundahl makes something similar, otherwise it was not available commercially.
I have had my say, but just regarding the Quad: The Quad 22 was actually the stereo pre-amplifier. I believe you are referring to the Quad II power amplifier.
If so, this was indeed an innovative design. Just one thing to note: The output stage was basically UL, but the OPT primary portion between G2 and B+ was transferred to the cathode side of each KT66, thus there was in addition some local feedback. That is what made this design better than other output stages. In that sense one cannot compare with normal UL or anything else. It did require a special output transformer. I have understood that lately Lundahl makes something similar, otherwise it was not available commercially.
Let me just re-iterate a few things here.
Pentodes strapped as triodes sound like triodes, because essentially in that mode they are triodes. But, they sound like DIFFERENT triodes, so comparing them with any given true triode is like comparing two different triodes.
That being said, the supply of true audio triodes has pretty much come down to two types, whereas there are many more audio pentodes pentodes that are usable triode strapped. But, these were of course designed to work as pentodes, so shat you get when they are triode strapped, may not be a particulairly good triode, even if it, by extension, comes with the 'audio' moniker, inherited from it's intended application in pentode mode. On the other hand, you can get very good triodes by triode strapping pentodes decidedly not intended for audio use...
Finally, triode and pentode amplifiers really have quite different characteristics, one of which is that triode amps look more like a voltage source, which is what the grand majority of speakers expect to be driven with. Naked pentode mode amps need speakers intended for current mode drive, these are anything but the norm. In order to enable a pentode amp to look more like a voltage source, some form of feedback is needed, which immediately makes it a whole different beast compared to most triode amplifiers which do not necesairly need feedback - so we are back to comparing apples with oranges.
Pentodes strapped as triodes sound like triodes, because essentially in that mode they are triodes. But, they sound like DIFFERENT triodes, so comparing them with any given true triode is like comparing two different triodes.
That being said, the supply of true audio triodes has pretty much come down to two types, whereas there are many more audio pentodes pentodes that are usable triode strapped. But, these were of course designed to work as pentodes, so shat you get when they are triode strapped, may not be a particulairly good triode, even if it, by extension, comes with the 'audio' moniker, inherited from it's intended application in pentode mode. On the other hand, you can get very good triodes by triode strapping pentodes decidedly not intended for audio use...
Finally, triode and pentode amplifiers really have quite different characteristics, one of which is that triode amps look more like a voltage source, which is what the grand majority of speakers expect to be driven with. Naked pentode mode amps need speakers intended for current mode drive, these are anything but the norm. In order to enable a pentode amp to look more like a voltage source, some form of feedback is needed, which immediately makes it a whole different beast compared to most triode amplifiers which do not necesairly need feedback - so we are back to comparing apples with oranges.
ilimzn said:
Ilimzn,
As usual I have to think carefully when replying to a post by you! (compliment).
I hear (OK, read then
) what you are saying. But "intended for audio use" and "audio triodes" are relative terms from a theoretical point of view. Indeed, you are correct by mentioning pentodes "not intended for audio use" making good triodes, supporting in a way my point. And some triodes used for good audio these days were originally designed as general tubes within their capabilities. (As far as I know - that was long ago!)Also yes, triode-strapped pentodes are different triodes, but not necessarily less suitable for audio use than original triodes. (I get the impression that what is intimated is that "true" triodes will - must? - sound better. That is not apparent from the transfer characteristics.) Further, pentodes designed for audio ... that again is a general term. They have often been used for r.f. work, e.g. a good example is the 6L6, the first audio beam tetrode, but it found admirable application in r.f. as the well-known 807 (exactly a 6L6 only adapted for higher voltage).
On the triode side I find a number of up-market designs using the 6AS7 dual triode, thus apparently considered suitable. But the 6AS7 was originally designed for power supply regulator use, not audio, should one want to bring the "designed for" phrase in.
Thus I am saying that a triode-strapped pentode can make a perfectly good audio triode. If a straight triode were to have been designed having the same transfer graphs, it would have qualified as an "audio triode" just as suitable as any other "audio triode".
So, I have some difficulty to picture what characteristics an "audio triode" might have as distinct from any other power triode. Obviously there are unsuitable tubes for audio, both triodes and pentodes; we are not talking about such. Naturally characteristics must be investigated and a well-motivated choice made.
Regards
Interesting Thread!!
Ricky:
If you are contemplating building a (relatively) high power SE triode amp, remember, due to the logarithmic nature of the relationship between sound level and power---a 40 Watt amp will sound only a "tiny bit" louder than a 20Watt amp. It will not be "twice as loud", rather, the difference in output will be hard to detect, for most.
Johan, or anyone that cares to answer:
A related question that I've had for some time is: Why are pentodes always strapped for triode operation? What would be the disadvantage of letting the screen unconnected? Are they strapped only to pick up the slight additional dissipation, or are there other reasons?
Ricky:
If you are contemplating building a (relatively) high power SE triode amp, remember, due to the logarithmic nature of the relationship between sound level and power---a 40 Watt amp will sound only a "tiny bit" louder than a 20Watt amp. It will not be "twice as loud", rather, the difference in output will be hard to detect, for most.
Johan, or anyone that cares to answer:
A related question that I've had for some time is: Why are pentodes always strapped for triode operation? What would be the disadvantage of letting the screen unconnected? Are they strapped only to pick up the slight additional dissipation, or are there other reasons?
dhaen said:To produce 40 watts SE with a single valve is difficult and expensive. I've done it, using frightening voltages and lots of dosh. I wanted it to drive my existing speakers. It would have been much more sensible to build a 5 or 10 watt amp and replace my speakers. I think I just did it because I could
Tell us more...
dave
Johan Potgieter said:
It's not difficult at all. What's good in an audio power triode is good linearity, and the low rp (for good woofer damping) that goes with a low u-factor. That's what you see with types like the 45, 50, 300B, 845 power triodes. Some of the series pass triodes have similar characteristics since pulling substantial plate currents at low Vpk's is also desireable.
What's desireable in an RF power triode is maximum power sensitivity. That means high u-factors, sometimes so high that you can operate with very small Q-Point currents without bias (the "zero bias" tubes, such as the 811). Here, you're not concerned with linearity since you lose your harmonic distortion since you're opertaing into an LC tuner and/or a bandpass filter. High rp is of no censequence here.
They're not. I've done two designs, one with 807s and the other with 6BQ6s, both operated as pentodes. They work just great.
A floating grid will pick up electrons, develop a negative charge that'll slowly choke off plate current. The only reason for trioding is to get a triode characteristic. In the "good ol' days" the pentode revolution came on quickly (in ~1934). Since pentodes are a good deal more efficient, this development killed off the use of audio power triodes. The last one developed being the 300B. After that, the only triodes to appear were RF types, series pass regulators, and vertical deflection PAs for TV. Between the smallwatts of 45s, 50s, and 2A3s, and the big one, the 845, there really aren't any audio triodes falling into that gap. Trioded pentodes can.
And let's not forget that 45s, 300Bs, etc, are Audiophool Expen$ive. It's simply a good deal less expensive to triode 6V6s.
This thread gets more and more interesting by the post, it really does.
Here's what I found running pentodes as pentodes, pentodes as triodes and triodes on their own.
For power, Pentodes as pentodes is the best. However, to my ears and others, they can sound ' a bit rough and ready'. By running them UL they sound a lot smoother. This was 6 EL84's in parallel single end with 20% UL tap on the OPT.
I didn't like the ' compromise' by strapping G2 to the anode. It didn't work for me sonically. Probably measured a lot better though.
Parallel single end triodes did it for me. 6SN7's on this occasion. All the sweetness of the valve plus the 'bottom end whack' ( I hate that phrase, I really do!) of a good transistor design. That's not me by the way, that's a tranny amp manufacturer's opinion.
So there you have it guys. Triodes do it for me ( but that's me ). No feedback, straightforward design.. Doesn't mean it's everybody's cup of tea though.
Ricky.
Here's what I found running pentodes as pentodes, pentodes as triodes and triodes on their own.
For power, Pentodes as pentodes is the best. However, to my ears and others, they can sound ' a bit rough and ready'. By running them UL they sound a lot smoother. This was 6 EL84's in parallel single end with 20% UL tap on the OPT.
I didn't like the ' compromise' by strapping G2 to the anode. It didn't work for me sonically. Probably measured a lot better though.
Parallel single end triodes did it for me. 6SN7's on this occasion. All the sweetness of the valve plus the 'bottom end whack' ( I hate that phrase, I really do!) of a good transistor design. That's not me by the way, that's a tranny amp manufacturer's opinion.
So there you have it guys. Triodes do it for me ( but that's me ). No feedback, straightforward design.. Doesn't mean it's everybody's cup of tea though.
Ricky.
My amp uses 6080 in a PP design without feedback. I love the sound of it, very SE, and it covers the whole spectrum down to 10hz. However the one thing it doesn't do brilliantly well is bass attack. I think this is where a bit of feedback would probably help. Unfortunately thats not an option with this design.
I was listening to a solid state PA amp and it reminded me what silicon does best. Bass heavy tracks need good woofer control and that not that easy to achieve with triodes or any valve.
The compromises we have to live with.
Shoog
I was listening to a solid state PA amp and it reminded me what silicon does best. Bass heavy tracks need good woofer control and that not that easy to achieve with triodes or any valve.
The compromises we have to live with.
Shoog
Johan Potgieter said:Ilimzn,
As usual I have to think carefully when replying to a post by you! (compliment).
I hear (OK, read then
Also yes, triode-strapped pentodes are different triodes, but not necessarily less suitable for audio use than original triodes. (I get the impression that what is intimated is that "true" triodes will - must? - sound better. That is not apparent from the transfer characteristics.
The highlighted parts of your reply really are the points i was trying to make, namely:
(1) The curves look like a triode, then it works like a triode. No difference from a 'real' triode, assuming you do the design work properly, at least for audio frequencies and FAR above that.
(2) Very often people assume that if it says 'made for audio amplification' it means it's automatically going to sound better. This can lead to dissapointments. For instance, although the KT88 is well respected in pentode and Ul modes of operation (but, keep in mind there are barely any tubes that you can easily sompare spec for spec, so it may well be a case of 'it's good because it's the only one we have'), it wouldn't be the first pick as a triode.
Yes, and also for horizontal deflection in TV sets, etc. Developing a tube is a very costly endevour, so it is not surprising that as much as possible of the developer's effort is 'recycled' into other versions of the product.
But, as i said above, my point was that people often assign 'magical properties' if it says it was developed for audio use, and more often than not, designers go the route of least resistance simply because some tubes gain great popularity because of this, even though there may be better suitable ones that were 'not made for audio'.
Not to mention the 6S33S, or triodes originally used for RF, like the 211. But, just look at how long it took for people to recognise the 6AS7 (and it's been around forever), or 6S33S.
The point is, it all comes down to having the knowledge and idea how to implement a given tube in a topology that gives good results. 6AS7 is a great example - it is really not particulairly linear if you look at the curves, and it's mu is impractical, but all of this has a completely different meaning if you (1) decide you don't need to use standard output and power transformers, (2) you cancel the nonlinearity using a PP output stage, or using complementary distortion cancelation. If you just lash it up with a 6SN7/6SL7 because those are 'popular', you don't get very far, and rightfully so.
And here we completely agree.
Miles-
Thanks for the reply! The first part of my question was, of course, absurd! What I intended to ask was: When pentodes are operated in triode mode, why are the screen grids always strapped to the plates, rather than left unconnected? ---which you answered as my second question.
No wonder my analyst doesn't understand me!
Thanks for the reply! The first part of my question was, of course, absurd! What I intended to ask was: When pentodes are operated in triode mode, why are the screen grids always strapped to the plates, rather than left unconnected? ---which you answered as my second question.
No wonder my analyst doesn't understand me!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.