Triode versus Ultralinear/Pentode

NO. But that's only because you used the "always" word. :-)

Wiring triode may reduce some distortion, you can probably use less feedback, a good thing IMO. But you end up with a lot less power and that won't set too well with some speakers. At least you could go back if it didn't work.

I have an Audio Space AS-3i with a triode/ultralinear switch on the front panel. This is my story...

I've used both EL34s and KT66s in this amp. With the EL34s there was more power available, of course. I ran the amp all the time in triode mode because I found the sound sweeter and more detailed. With KT66 tubes the difference between triode and ultralinear was less obvious. I used ultralinear for the extra power. FInally, though, it began to appear that triode was the nicer one even with KT66s.

Having a switch on the front panel is really useful. There are some discs--rockers, full orchestras--which get a needed boost in impact with the extra power of ultralinear mode. But in the end it's a bedroom system and I do most listening in triode.

I've never listened to a pure triode designed amp so this is definitely a FWIW.

I have three amps were you can select triode/ultralinear on the fly. One, (a Cary) in triode sound much the same as two other's in ultralinear. In ultralinear it becomes slightly hard. All have low negative feed back.

With the speakers (3 different pairs)I have used with these three amps, best described as relatively neutral 3 way cone designs with a gentle impedence curve, unltralinear produces what I consider natural and 'linear'. When put into triode the mid-range power seems to deminish and the center sound stage recesses. A very relaxed sound but on a lot of music it sounds a bit thin.

I suspect triode might be a mode that would sound best used with more aggressive speakers (no shortage of those in the high end).

I do like to use triode for background music however. Apart from it being 'relaxed' at about half power the tubes and amp run cooler and I suspect last a lot longer.

Switching an amplifier from pentode to triode can be a bit tricky inside a single amplifier. The reason is that triodes tend to have a lower impedance than pentodes and so will want to operate at a different impedance on the output transformer.

That is tricky to set up and a lot of amplifier manufacturers 'fudge' that parameter a bit.

Its also interesting to observe the behavior of a pentode wired in triode as opposed to a real triode- it becomes obvious that real triodes have better performance. It is often difficult to demonstrate this fact because there are so many variables that get introduced, but- if you are thinking that triodes might have something to offer, you would be right. One of them is *not* power :)

You use triodes when the best sound possible is your goal. They offer lower distortion meaning little or no feedback is required, and lower output impedance which means less turns on the primary of the output transformer (that translates to: less distortion and more bandwidth and *that* translates to 'more detail, better highs and lows') or maybe no transformer at all.

The trick is like it always is- get the right speaker to take advantage of those advantages. There are high power triode amplifiers but they tend to be rare.

No. With my Mesa Baron, Triode is preferred for acoustic and smaller scale music, but adding some pentode helps a great deal with rock, orchestral, some jazz, piano and even solo guitar sometimes. and it solidifies imaging a bit. I suspect it's as we've always heard: it's all in the implementation. Unless you listen only to chamber music, there are times when pure triode is subobtimal.

Lloydc, your amplifier might be a good example of what I meant- that even though there is some advantage of the 'triode mode', that mode is not optimized in the amplifier in order to allow for the switching in the first place. Were it optimized, there would be no trade-offs in sonic performance, only power.

So I take it that if I were to convert the Music Reference to triode it should be triode only, optimized for triode, and not a triode/ultralinear via switch approach.

What does Roger Majeski say about the potential result of the conversion? Benefits v liabilities? There is always some trade off - what does he think they are with his amp?

I asked that question on AudioCircle in the Manufacturer's forum and still have not gotten an answer. To some exent I hesitate with the idea of the mod (he did not object to the idea), thinking if it was such a good idea, why didn't he do so from the begining - could be the amp has more universal application run in AB2 Ultralinear. I know by going to triode, he pull back on the NFB which appears to always be a good thing with my Merlins. I agree in principle (with no engineeering knowledge whatsoever) that it is best to have a unit optimized for one type of operation or the other - the logic of that makes sense to me.

Pubul57, Merlins don't need a lot of power. We have had many Merlin customers over the years using our M-60 and I never hear about power issues. If you can get a triode amp that makes at least that much, unless you are in a very big room you should be fine!

With Merlins it all about finesse rather than power- they are plenty revealing- going triode with them will definitely pay off.

See if you can find one of these for sale: (http://www.stereophile.com/tubepoweramps/1200cary/) You won't do much tube tasting as far as the output tube(T-1610) 'cause there's only one company that makes them(KR). Then again there's the price of a pair of those glorious triodes. At least they'll give you 100 watts on peaks. BTW: If you can afford those- Are you interested in adopting a son?

I'm not so much worried about power, as the Music Reference RM9 SE would be 70 watts in triode - way more than enough for my Merlins, I'm thinking about whether converting an amplifier from ultralinear to triode makes much sense if the amp was not desinged for triode operation from the ground up; I just fell there must be some compromise in this apporach rather than being built and spece'd specifically for triode appplication, like my CAT JL2.

I'm 95% there Tvad. It's why I have doubts over the wisdom of what I am considering. I think your logic is impecable. It's not like Roger said, "great idea, I don't know why I didn't think of that." - Although I'm sure he was also interested in bringing something to market with more power than triode would have provided. The whole triode idea is driven by the fact that my speakers (Merlins) like to be driven by amps with very low damping and NFB, and triode operation theoretically permits use of less NFB (or none?), etc. But I do hear the logic behind what you are saying, and I'm leaving well enough alone and going with the amp as designed by Roger.

Taking up atmasphere's point and changing the question. Is it always true that the compromises envolved in making an amp adjustable between Triode and Pentode, mean they will perform worse than an amp optimised for triode or pentode only?
To answer my question and having spoken to a few tube amp designers, it is my belief, that the compromises do, inevitably, have some effect on performance.

I thought that Atmasphere's point was that "switchability" is often a compromise, but not necessarily so and not always - but one should tread carefully.

Pubul57 is correct. Some amplifiers pull it off quite well- others don't. I rebuilt a Marantz 9 from years ago- the amp sounded OK on Pentode, but really, triode mode was the really useful mode if you were listening seriously. Most triode amps today would eat if for breakfast, though. Its a fact that pentodes wired in triode do not have the same linearity of real triodes.