My 3 cents:
Triode sound means modulation of Rp that produces mostly 2'nd harmonic distortions. It sounds very natural. I've found that triode - strapped pentodes have triode characteristics and sound like triodes, for example 12L6.
Pentodes designed for RF and vertical deflection in TV are very linear in audio applications and sound nice, for example 6P15P, GU-50, while some triodes are hard to operate, for example 6N7: they can dissipate enough, but deliver very small current when 1'st grid is negative.
Triode sound means modulation of Rp that produces mostly 2'nd harmonic distortions. It sounds very natural. I've found that triode - strapped pentodes have triode characteristics and sound like triodes, for example 12L6.
Pentodes designed for RF and vertical deflection in TV are very linear in audio applications and sound nice, for example 6P15P, GU-50, while some triodes are hard to operate, for example 6N7: they can dissipate enough, but deliver very small current when 1'st grid is negative.
I believe Pentodes get this bad wrap mostly because loudspeakers impedance rises in the upper mid to high frequencies. Figure, if the amp has a higher Zout that means your SPL output will be more when the load impedance is higher. Pentodes can sound excellent when properly applied. If you want to run Pentode without NFB, a Zobel on the driver(s) will make it sound a lot less harsh. If the bass is too much, put a notch filter to flatten impedance in the LF. Also, Pentodes would probably do a good job with Magnepans since they have pretty much flat impedance and require a lot of power.
Hi.
Ok. I think (???) my original question has been answered, and a few other things taken on board besides.
I should like to take this opportunity to thank you all for your valuable input to this thread but, seriously guys. Hasn't the time come to stop talking round in circles and get on with what we do best?? Making music with the equipment we have built??.
Question asked, Answers given, thread closed, as far as I'm concerned.
Thanks guys. I'm outta here.
Ricky.
Ok. I think (???) my original question has been answered, and a few other things taken on board besides.
I should like to take this opportunity to thank you all for your valuable input to this thread but, seriously guys. Hasn't the time come to stop talking round in circles and get on with what we do best?? Making music with the equipment we have built??.
Question asked, Answers given, thread closed, as far as I'm concerned.
Thanks guys. I'm outta here.
Ricky.
ArtG said:
There's another consideration as well. If you connect the screen(s) to the plate, you get a low-u characteristic, with its desireable low rp. You can also simply connect all screens in parallel, and run them all as a control grid. This is done quite often to triode pentodes for RF amps. Connecting all grids together makes a high-u, "zero bias" type triode with the high-u, high rp, and high power sensitivity that you're looking for in an RF power amp.
Since none of that is desireable in an audio triode, you probably won't be seeing it in an audio amp.
Miles Prower said:
Miles,
Sure - perhaps some vagueness from my side. I find (as mentioned further down in my original post) that there is sometimes, so-to-mention, an aura of audioness round a so-called "audio triode". Triodes for audio preferably have the characteristics you mention, but all triodes with such characteristics would be suitable (also triode-connected pentodes), e.g. to a degree also the 6080 etc., not listed as "audio triodes".
No disagreement on this point.
rdf said:
No it doesn't. I checked the supplier I use most often and their price for a 12AU7/ECC82 is $14.00 (with a '*' that indicates "premium", high demand VT) or 12AU7WA/6189 for $9.50. (Which is why I went ahead and used 6FQ7s for the cathode follower grid driver: there was no saving there.) The 6BQ7As I used for my last project go for $2.00.
'AU7s don't exactly cost "nothing". Not surprising since these puppies are found in lots of gee-tah amps.
It was once pointed out to me that if one compares the plate curves of a 300B with those of a triode-strapped EL34, there is a great difference in linearity. Indeed, I could see that was the case. The 300B looks extremely linear whereas the triode-strapped EL34 does not. The point of it all was that if I expected to get 300B-class performance from triode-strapped EL34s, I was going to be disappointed.
Now, I have listened a lot to trriode-strapped EL34s in PP and they sound OK to me. I haven't had any experience of power triodes of any type, either DHT or IDHT (if the latter exist), so I don't know how much better they would sound, but I can see from the curves that 300B, at least, ought to sound better. Maybe one day . . .
Now, I have listened a lot to trriode-strapped EL34s in PP and they sound OK to me. I haven't had any experience of power triodes of any type, either DHT or IDHT (if the latter exist), so I don't know how much better they would sound, but I can see from the curves that 300B, at least, ought to sound better. Maybe one day . . .
ray_moth said:
This is an excellent example where oversimplification of the problem can lead to erroneous conclusions. The point being, a typical SET amp is not only the output stage, but also the driver stage. It's not by accident that there are tons and tons written about what driver tubes to use and in what configuration - though not so often that certain drivers work better with certain output tubes. IF one looks at the driver configuration, it can also become apparent that the driver stage or stages (and this makes a big difference too) topology plays a part as well.
Common sense at first glance dictates that you want to keep all your stages as linear as possible. However, a second glance might reveal that since you can't make them ideally linear, you might want to take advantage of the polarity inversion of certain topologies, to use nonlinearity of one stage to cancel out (some of) the nonlinearity of following stage. So, while it's a very appealing proposition to separately build stages and then combine as needed for experimenting, it is not necesairly the best strategy - it may well be another reason why there is constant debate about triodes vs triode strapped pentodes, or indeed, some triodes vs others. It really comes down to doing a design tailored speciffically for a given output triode (either 'true' or triode strapped pentode, tetrode or BPT). Of course, it may well end up being a design where you can't play around with many different tubes...
It should also be pointed out that you CAN go too far in the attepts to cancel out nonlinearity - the way it is done with triodes cancels out even order distortion, but may make odd order distortion worse - so it's a question of finding the right balance, which puts us back into the subjective realm of what we percieve as 'good sound'.
One other thing that should be mentioned is that when an amp is designed, it is usually done graphically, using load lines. A load line represents a true resistive load, which in reality never happens, amongst other things because the speaker and output transformer do not combine into a purely resistive load. As a result, operating conditions actually vary substantially depending on frequency and amplitude, and the instantaneous operating point can find itself in quite nonlinear regions of operation.
It should be noted that nice parallel and evenly spaced lines that make a tube suitable for linear amplification are more often found in triodes that are used for audio, than in pentodes - in fact, it is more likely the will be found in so-called video pentodes, which are not very large. In general, what obne sees with pentodes or BPTs is a bunching up of curves towards the X axis, and with BPTs there can also be a residual tetrode 'kink' at lower voltages and currents. The trick here is to understand how to properly use such a tube. As a pentode or BPT, bunching up of curves towards the X axis generally means it is best used in a PP amp, class AB (and less often, class A). The reason being that in the area close to the X axis would be the one where both tubes in the PP arrangement work at the same time, compensating for each other's nonlinearity, which may be apparent once the composite curves are generated (A good idea in general, even plotting composite curves of a driver and output stage of a SET amp).
Such a tube may not be the best choice for a single ended pentode amp, even with a purely resistive load, and especially not suitable with a reactive one. So, when doing comparisons, we could be looking at an ill-concieved design. However, such pentodes may make perfectly usable triodes when triode strapped. Perhaps the best example would be TV horizontal sweep tubes. So, we are back to the fact that comparing pentode and triode modes of even the same tube is rarely if at all possible on equal grounds.
Ray, take a look at the 813 and 814 triode curves on Pete Millett's site and prepare to be amazed.
Re: even order cancellation, it's a nice trick under controlled circumstances but I don't see how it can be reliably achieved in the wild without serious local feedback around each stage. The distortion spectra of equivalent tubes from different manufacturers varies way too much to guarantee the 1-2 dB accuracy required for significant harmonic cancellation at one level, much less across a wide operating range. I've given up trying and now aim for minimizing the distortion in voltage gain stages.
Re: even order cancellation, it's a nice trick under controlled circumstances but I don't see how it can be reliably achieved in the wild without serious local feedback around each stage. The distortion spectra of equivalent tubes from different manufacturers varies way too much to guarantee the 1-2 dB accuracy required for significant harmonic cancellation at one level, much less across a wide operating range. I've given up trying and now aim for minimizing the distortion in voltage gain stages.
I couldn't find any curves on Pete Millet's site, rdf. I did find an SE 813 design but that's not my 'bag' - terrified of high voltages and not an SE fan 
I appreciate your thorough treatment in reponse to my post, ilimzn - I was hoping for something like that. I'm sure you're right about the need to look at the driver combined with the OP triode.
I can't help thinking there must be somthing to the 300B phenomenon, otherwise why would people spend so much money on such a funny-looking bottle? I'm only thinking about PP here because I know next to nothing about good SE design.
I appreciate your thorough treatment in reponse to my post, ilimzn - I was hoping for something like that. I'm sure you're right about the need to look at the driver combined with the OP triode.
I can't help thinking there must be somthing to the 300B phenomenon, otherwise why would people spend so much money on such a funny-looking bottle? I'm only thinking about PP here because I know next to nothing about good SE design.
Wondering what everyones preferences are when a pentode or triode is asked to be used as a input/driver in a two stage 300b amp.
The areas of interest for me in trying to decide on an appropriate input/driver are having both maximum drive capability to give non- homogenous sounding tight bass (something the 300b lacks) and retaining the best tone (thinking maybe for pentodes el84, D3a, c3m, C3g and triodes 6c45p, 417,437,46, 6em7).
The areas of interest for me in trying to decide on an appropriate input/driver are having both maximum drive capability to give non- homogenous sounding tight bass (something the 300b lacks) and retaining the best tone (thinking maybe for pentodes el84, D3a, c3m, C3g and triodes 6c45p, 417,437,46, 6em7).
rdf said:
This is indeed correct if you look at the typical tubes. Much easyer to implement with tubes one would consider nonlinear in a predictable way - the 6AS7/6080 comes to mind. Another good example not directly related to output stages would be the 12AT7 or even ECC85 in a LTP phase splitter configuration.
Maybe I should of wrote "to further assist in addition to the other aspects like PS, etc. to make non-compromised tight bass and still have exemplary tone".
I had been scratching my head why on earth the majority of commercial 300b amps including today use the very criticized 6sn7(by many on the various boards due to there poor performance compared to the other tubes mentioned) for driver duty. Were they just stupid? I then read somewhere it was purely for the sound aspect (sacrificing the performance end) which had a better tone then some of the current in vogue low plate resistance/high current triodes or pentodes.
Hence my question being which performs best in a 300b amp in the input/driver position and yet still sounds great.
I had been scratching my head why on earth the majority of commercial 300b amps including today use the very criticized 6sn7(by many on the various boards due to there poor performance compared to the other tubes mentioned) for driver duty. Were they just stupid? I then read somewhere it was purely for the sound aspect (sacrificing the performance end) which had a better tone then some of the current in vogue low plate resistance/high current triodes or pentodes.
Hence my question being which performs best in a 300b amp in the input/driver position and yet still sounds great.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.