Ultralinear vs. Triode vs. SET


I currently have a Rogue Cronus that I have been quite satisfied with, but I am intrigued with the possibility of a SET amplifier. From both a technical and sonic perspective, what are the differences between a tube amp with switchable ultralinear and triode mode vs. a true SET amplifier?
droz
Hi Droz,
I was in your shoes about 14 months ago. I prefer tubes over SS amps and had been using push-pull(PP)amps with mcuh satisfaction. These amps were pentode/triode switchable and were El34 or KT88/6550 based. Well after reading reviews and articles concerning SET amps my curiosity got the best of me and I purchased the CoincidentMK2 300b monoblocks.I`m absolutely hooked on these amps they`re just fantastic. As other posters have said above a good quaility-well designed SET amplifier just sounds so naturally real and live. The tactile and palpable presence is the best I`ve heard compared to SS amps and even very fine sounding PP amps. The sound is very organic, fast and dynamic, yet it retains such a relaxed musical flow.
I disagree with those who say you give up upper and lower extension for sake of a great midrange NO YOU DON`T! This is a misinformed stereotype Droz. The key to sucess is matching to a appropriate speaker(high effiency and favorable load-ohms). You do not have to settle for rolled off highs or soft/flabby bass. You will have many choices in different DHT tube types and also various speakers. Just be patience in your search and read many reviews of SET amps to increase your knowledge base. The world of SET can be wonderful with the proper approach. I hear nuance harmonic overtones and micro detail that were hidden with my PP amps.
All great information, but not all of it true for all SETs. I love my speakers (87dB) and did not wish to give them up, but did wish to try a true SET. Thankfully, Audio Mirror 45 Watt SET amps are true SETs, yet very powerful, and not lacking frequency extension in any way. If you love your speakers and wish not to change them, you owe it to yourself to try a true, powerful SET. I believe there are a few other manufacturers available with true powerful SETs as well. Swapping the stock driver and input tubes with NOS took my system to a whole new level.
Droz,

1. Pretty much. It's single ended. It will have it's own flavor depending on the tube used. One of my amps is the little Glow Amp that uses 1 EL84 per channel. Nice sweet sound.

2. It would be close. A push pull 300b amp would have a similar flavor to a single ended 300b. I hear a purer tone and better decay with SE.

What do you value more, loud cresendos or smooth decay?
Droz,
If you have not done so already, search the site for high effiency speakers this topic has been well covered on audiogon. The old threads contain much information and many good suggestions, there`s no shortage of very good speakers of this type.In contrast to Sparky`s statement above, in my experience the high effiency speakers sound more real/natural and less HiFi in many cases.
Push-pull amplifiers make more power, but often do so with increased 5th harmonic distortion, which causes them to sound brighter. This is not always the case though, if the amplifier is fully balanced from input to output the only distortion component will be the 3rd harmonic, which is considered one of the lower ordered harmonics and is more innocuous.

A pentode amp will make the most power for the tubes involved, but will also be the least linear, and so will need negative feedback to linearize it. There is a price paid here- you get most forms of distortion reduced, but with an increase in the odd ordered harmonics, which the ear finds unpleasant.

Ultra-Linear is a method of using the output transformer as a very local feedback loop to linearize the output section. The result is an output section that that has nearly triode linearity, and nearly pentode efficiency. This is a very nice way to set up a push-pull amplifier, but if the amp were actually using *real* triodes, it would sound even better as it would be more linear.

Output transformers are such that the bigger you make them, the harder it is to get bandwidth. This is particularly true of SETs, and is a primary (no pun intended) limitation in the higher-powered examples. In most cases, the 7W of a 300b is about as big as you can get and expect bandwidth that might be considered 'hifi'. This is also why the smaller SETs, the 2A3s, 45s and the like have the reputation for sounding better- they have greater bandwidth. Now you can fudge things a little- optimize for bass or highs (and since most high efficiency speakers have troubles making deep bass, the tendency is to go for the highs rather than the bass) since you aren't going to get both.

Push-pull overcame a lot of these issues to a certain degree, with its own set of trade-offs. For example, an SET will have nearly unmeasurable distortion as the signal level heads towards zero, whereas a push-pull amplifier will usually see increased distortion (this is commonly attributed to an additional circuit in P-P amps called a phase splitter; its worthy of note that not *all* P-P amps have to have this as a separate circuit). This is where SETs get the magical midrange that they are known for- when distortion is reduced, detail is revealed.

You still have the issue of bandwidth with P-P, but in general you can have about 10X more power available compared to an SET with equivalent bandwidth.

The DC that the power tube draws through the transformer of an SET can cause saturation (distortion) issues with the transformer. This is a major design consideration of the OPT for an SET. You don't have the DC saturation issue with P-P, as the DC currents used by the power tubes cancel in the core of the transformer. On top of that, even-ordered harmonics are also canceled in the load, making for an amp that should be more transparent at greater power outputs.

The distortion of an SET is mostly lower orders until the output power becomes significant. At that point the higher orders become significant (although measurably insignificant) to the human ear. The 5th, 7th and 9th harmonics are used by the human ear to determine sound pressure, so when they are distorted (even so slightly that it is difficult to measure) the effect is that of increased loudness. Since more power is usually used on transients, what you have is the loudness cues occurring on the transients. The result is that SETs will *seem* to be very dynamic for the amount of power that they have, and often people will say its all the power that they need- they don't want to turn it up any higher.

So now you can see that in conversations about such, that if you substitute the word 'distortion' for that of 'dynamics', the meaning of the conversation will not be altered! It happens that if you can eliminate the odd ordered distortions, the amplifier will have no loudness cues, and the immediate effect will be that of turn the volume up higher, as it does not sound 'loud'.

Now IMO and IME, 'loud' is something that a stereo best not do, **regardless of the actual volume**. That can be a bit of a trick, but is well worthwhile.