I pondered whether to post any information, either pro or negative, but I have some information that virtually no one ever posts, or has seen, so I have decided to post it here. Some time ago, I also posted at some info on AC. Anyway, here goes.
I must add that some two decades ago, there was a fierce marketing campaign on to push SET amps, especially using DHT tubes (along with profits). All kinds of superlatives, specs etc given to DHT amps.
Virtually no one had/has the test equipment, or listening tests, to separate truth from fiction. And when was the last time you heard the recording studio walls when playing an set amp.
The main difference claimed between SET and other amps is the "midrange magic". This is partly due to:
1. The claim is that DHT distortion is predominantly 2nd harmonic in nature. This sounds good except:
A. the distortion spec is virtually no better than any other tube. Below I compare a 300b to my
SET using IDHT, and my PP amps. In both cases, my amps have much much lower distortion, without global
negative feedback.
B. there are 3rd, 5th, and 7th order products produced by DHTs, just like other 3/2 power tubes. Check bottom
of post to view Eimac data. I always have to chuckle when I hear the extravagant claims of DHT distorton is
even order harmonics.
2. The midrange "magic" is also due to frequency roll offs of the extremes (parafeed operation may help the bass).
The roll off of the extreme high and low frequencies are due to:
low transconductance, poor frequency response of the input/driver tubes,
poor quality and small size of coupling capacitors, and
DC current through the output transformer (OPT) reducing the primary inductance and also increasing the distortion
(See RCA Radiotron Designers Handbook, 26 engineers. All these problems minimize both high and low frequency response.)
By the way, the lower the high frequency response, the more musical nuisances are lost. The high frequency response mimics a low pass filter, just as power supply filtering stages minimize 120hz ripple (signal). The lower the high frequency response, the more masking of low level musical nuisances/information/cues. Ever hear of the intimacy described of an SET amp?
That is because the low level musical nuisances has been essentially removed leaving only the main instrument.
3. The SET amplifier is basically a simple build, and has hidden faults inherent in a Class A design.
A. The damping factor varies from zero (at maximum plate resistance), to maximum damping (at minimum plate resistance, tube conducting maximum current), for any type SET amp. PP basically does not have this problem due to one tube maximizing damping while the other is minimizing and visa versa, so basically constant.
B. The intermodulation distortion (IMD) is very generally 3 times that of the harmonic distortion (HD) figure. Thus for 5% HD, the IMD is some 15%. For 10% HD, the IMD is some 30%. Both types of SET amplifiers are known for their high IMD.
To remove the perceived distortions, all the low level musical information is essentially removed leaving only the main instrument.
4. The lowest distortion driver tubes ever produced are the new JJ brand 6dj8/12axx line of tubes, and
by a wide margin. We are talking some 1/9th or less distortion, or -18db less than any other tube, such as a
6SN7xxx series. A tube with that low of distortion is difficult to work with but effective.
Below is the distortion products (.pdf) of a typical 3/2 power tube, either DHT or IDHT. Notice the 3rd, 5th, and 7th orders, distortion claims the DHT crowd refuse to acknowledge. (More later.)
5. I have heard claims that the musical signal is somehow "taken apart and put back together again" in push pull designs. That is utter nonsense. The signal is simply divided in two, one for each output tube. There is no magic trick involved. I guess such marketing works though.
6. One of the easiest things to design is a limited bandwidth audio amplifier. But how does a limited bandwidth amp realistically produce music? It cannot by definition since so much of the music is deleted. It takes a wide bandwidth system to accurately reproduce realistic music with dynamics/rise times, spacial characteristics, proper harmonics etc.
7. Distortion from a different view. Below is distortion data of 300b vs power output vs RL.
Just the 300b tube, 4% harmonic distortion (HD) at 6.5 watts output. (see graph below) The tube is lucky to produce
less than 1% at 2 watts.
A comparison of my 20 watt PP triode KT88 amp, the entire amp measured ~2% at 7 watts and 0,05% at 1 watt without no global negative feedback.
We also need to consider is the combination of input/driver/output tubes in an amp and the distortions produced.
300b needs up to 200 volts p-p drive signal for full output. At least three stages (which includes the step up transformer). KT88 needs less than half that.
6sn7 types have limited gain, so two stages are necessary, with much much higher distortion and harmonic orders created. Besides that, more stages means limiting the bandwidth even more.
In order to mask all the distortion, the operative is to eliminate low level musical information so the instrument is all one hears. Then call it "intimate". But where is the natural sound stage, the echos of the recording studio?
8. For the number guys, distortions combine to create new orders. For instance, 2 stages, 2nd harmonic from X tube combines with the 2nd and 3rd from Y tube to create 4th and 6th harmonics. 3rd order combines with 2nd and 3rd to create 6th and 9th harmonics. Notice the 7th harmonic is created in a 3/2 power tube so we can obtain 3rd and 7th for 21st order harmonics etc. But one never sees those harmonics in the measurements. (The higher the harmonic/order the more sensitive the ear is.) Rid the higher order harmonics by eliminating the inner detail of the music.
9. Another problem with SETs in general is the need for chokes/inductors in the power supply to minimize hum (which is a signal). However, from the musical stand point from the musical signal end, chokes cause a large non-linearity that interferes with the musical honesty. Yet another example of an SET amp artificially manipulating the music in various ways. Anyone can manipulate the music at much less cost, but lots of expensive "iron" and tubes makes for marketing. More cost, more profit.
10. Don't want to forget the IMD distortion products, which are sum and difference products. As a simple example, say we have a 1khz and 3khz (3rd harmonic), we have products of 4khz, original 1khz, original 3khz harmonic, and 2khz at a minimum. But the 4khz and 2khz are not part of the original musical signal. Try playing an orchestra with all those musical signals and harmonics.
11. One of the main advantages claimed by DHT fans is "intimacy". However, as we have seen above, manipulating the music in various ways means one is not hearing the real music, what the artist intended. It is not really intimacy.
Have you ever heard the recording studio walls when playing any SET amplifier? Of course not. Yet, it occurs with PP, OTL amps all the time. How about Pro Musica 2nd/3rd reference upgraded ST-70 amps, DeHavilland, Herron etc.
I have been in the audio world for over 40 years, and I will take the real deal any day. Knowledge is power to understand the difference between reality and hype.
steve
