• AudioFanzine : Home Studio, Computer Based Music, Guitars, Basses, Live Sound & DJ
  • Audiofanzine en


Tubes & Transistors - Part III

Greetings and salutations, as always, dear audiofanziners! Even though we are still on the subject of tubes and transistors....I am going to start 'broadening' by further examples of actual usage as well as melange / hybridization of both technologies...I feel sure that Actual Application of certain equipment you may actually use in your home-studio or hi-fi listening-room is more interesting that going further back into "history of development and drawings of internal construction of both technologies"…In any case, since this is a truly democratique forum/zine, you are free to pose questions and ask for extra details... And our éditeur en chef, will certainly decide what is worthy of inclusion.

So now I will start to discuss topology/usage of both technologies... 'Topology' defines Recto View of a device, normally shown or described in "block diagram" sections rather than full circuits with every component drawn exactly with its detailed values/part numbers etc. And I am going to move toward description of tubes and solid-state (to embrace all transistor usage/combinations)

About single-ended amps...

Let us begin, for a change, at the end of the chaine... The power ampli that drives your loudspeakers... Analogue transducers that your (also analogue) ears actually hear. So, topology-wise, we immediately get into the subject of "single-ended/lampe unique" as applies to tubes... But also available, though not from very many manufacturers, single-ended solid-state amplis.

Single-ended amp

And, by the way, though 'single-ended' specifically refers to the output stage of ampli's, the topology exists equally in preamplis but then is described as 'cascade’ topology... A chain of single devices following like 'ducks in a row' - 1,2,3 etc (tube or S-S) as opposed to 'differential / symmetrique / push-pull' topology... In preamplis both have their usages. While in amplis - I have to be frank and honest, sorry! - It is only my opinion but I deeply believe single-ended is totally without any merit whatsoever and will go on to explain why I know this to be so...(for historical purposes, I record that the very first amplis of perhaps 2 or 3 watts were perforce single-ended and first used in radios and low-power transmitters... Going right back to when Lee de Forrest added the 3rd electrode, the grid/grille, and tubes of amplification-value had arrived!) I will concede that 'those who love only single ended amplis are entitled to their choice/preference just as they are to their choice of violin against guitar/classical against jazz/ rock against jazz etc. Democracy must prevail.

There is a big following for single-ended amplis in Asia... Starting from Japan (which has never diminished hobby-wise) and has spread to certain other parts including some parts of Europe and even, though in very small proportion, the United States of America... The circumstances differ considerably, starting with VERY SMALL rooms in Japan and very large rooms (averagely speaking) in USA...

Horn-speakers

And then comes also as a "natural given" that these SE/LU amplis are only capable of very low power output (as low as 1 (ONE!) watt with a single 6B4 and 3,8 watts (single 2A3) and 5-7 watts with a single and very expensive 300B as can be found in a catalogue from "SAN-EI" (now long closed) where, in their defence, I must also mention that they employ negative-feedback and achieve distortion levels as low as 0,5%...! all of which leads to the next "natural given" being that it is essential to use hyper-sensibility loudspeakers ranging from 98dB to perhaps 105(!)dB (measured at one metre/one watt)… And then the next "natural given" that these speakers are not what one would choose in musical terms of being hyper-efficient "horns" or other variants that "shout in your face"... Again, another example of 'different strokes for different folks'... In other words the democratic tastes of various peoples.

Nowadays, with the 'changes of fashion' that come and go in the Hi-Fi hobby, the whole idea of negative feedback has been ruled as a "NO-NO/BAD BOY" defect which gives rise to 'risible' performance results like 8 watts with 5,7 and even up to 10(!) per cent distortion (a melange of all forms - harmonic, inter-modulation and frequency distortion when used at their full 8 watts power!(?) Do you honestly believe I have devoted my entire life in audio to achieving designs of amplis with this kind of performance? (Much worse than 'numerique' and excellent "France Telecom"!) But the SE/LU devotees tell me they "perceive a delicacy and purity" that is not available with push-pull amplis at 0,1% at full power output which was what Harold J. Leak was famous for in the '50's with his 12 watt Williamson design... And always these "delicate and pure" SE/LU amplis are demonstrated with solo guitar or string trio with the cello being the lowest frequency-content instrument. Just try the superb Ravel orchestration of Moussorgsky's "Pictures at an Exhibition" at room-filling volume... You will not listen for long, I promise! – even less long with "Dark Side of the Moon"! And here comes the whole "joke" of SE/LU lovers : mainly they prefer to play Vinyl 30cm 'disques noir'. OK democracy again, But not one of them has a single LP that was cut/mastered with a no-NFB/CR SE/LU amplifier driving the cutter-head (NEUMANN OR ORTOFON) because the whole system of modern electrical disc-recording depends on PUSH-PULL amplifiers of considerable power with NFB/CR driving a special and separate NFB/CR winding in the cutter-head! (I promise never to say one more word about SE/LU power amplis ever again... I rest my case.)

The push-pull topology

Now, moving on to push-pull topology in power amplis... I have previously mentioned the cancellation and rejection of AC/secteur power/heater noise (when tubes apply) and also "even - 2nd, 4th, 6th) harmonic distortion... I come now to two other output topologies/styles that occur in both the ampli AND the 'room positioning or 'layout' of equipment: (but quick interjection first please re solid-state amplis... Are you aware that the famous McINTOSH amplis continued to use output transformers when they moved from tubes to Solid-state?) The two topologies I mentioned as "up next" can both be given 'labels' of "European style" and "USA Style"… And the USA style is the better and more correct one in both cases - as I will prove and you will immediately 'see'; FIRST the ampli output stage: almost all (and I'm going back a little in time to include D.T.N. Williamson's seminal resistor/capicitor-coupled amplifier circa 1955... The first true high fidelity amplifier not utilising inter-stage transformers etc. And from which followed the leak, quad, rogersS, pye, goodsell etc. commercially-available models: all these used European 'thinking' which uses 'self-bias' as defined by either a single or a pair of cathode resistors varying in value from around 150 to 300 ohms...

This was and is a poor idea for at least three reasons : (A) about 40% of the push-pull maximum output power is 'lost/thrown away' and (B) with higher distortion resultantly and finally (C)... The cathode resistors must eventually burn-out/break-down as the tubes age and 'ask' greater current(unless one used, say, 50 WATT cathode resistors... Never seen by me in any amplifier!)... The reason for this commonly-used European 'self-bias' design I have never fully figured except for cost-saving possibility of the better and commonly-used USA concept, termed 'fixed-bias' which requires an extra MINUS/NEGATIVE power-supply of 35 to 65 volts to supply 'grid-bias'(in European 'self-bias' the input or control-grid derives its negative grid-bias VIA the cathode resistor/s.) So in the USA 'fixed-bias' mode(and, yes, a few, very few, USA models also used 'self-bias) All the above-mentioned 'down-sides' A, B, and C are fully REVERSED plus with the additional benefit of (D) longer tube-life... Because the amount of bias voltage can be constantly adjusted upwards so the tube draws/gets the optimum current throughout its much-extended life... You see what I mean by "Better idea?"

Find the right place for your amp

Find the right place for your amp

Then, moving away fro the Amplis themselves (SS or TUBE) and 'into room layout'... The commonly-used European layout puts the amplis at/near/alongside the pre-ampli/CD/turntable/tuner or 'listening position' thereby using long loudspeaker cables... Again a poor method when compared to the commonly-used USA plan where the amplis (whether monoblocs or stereo-version, and whether TUBE or SS) are placed near/next to or behind the loudspeakers... Much better idea altogether! Short speaker cables (particularly preferable with SS amplis which are current/ampere favoured rather than tubes which are VOLTAGE favoured) used with long interconnects between preampli and ampli... The benefits are very obvious indeed

  • A. Less space (and heat!) needed at the listening-position (may be on book-shelf close to/with vinyl discs that will Warp from heat)
  • B. Amplis out-of-sight (if behind loudspeakers)
  • C. Better current-transfer from amplis' output (ESPECIALLY SS!)
  • D. Much less expensive if "Golden-ear" high-cost high quality speaker cables are in use…

Here again, I offer my supposed reasoning behind the 'history' of the two methods : commonly, the European preamplis were, again for cost-savings reasons, High-impedance output from the anode/plate of the output tube, so they could not drive much more than one metre of cable without HUM pick-up as well as high-frequency loss (which ocurred to a small extent even with one-metre interconnects... Indeed, and again for cost-saving reasons, the one meter cable from preampli contained also in same cable, the voltages needed for the preampli itself and thereby getting its power from the main ampli and avoiding the cost of the preampli having its own internal power-supply. Lousy/Rotten/Poorly concepted idea!!! Again, predominant in USA models like McIntosh, Marantz, Harmon-Kardon, Grommes and even 'Heathkit' preamplis had own power supplies and low-impedance cathode-follower outputs... Again much better concept, no? And just in the case you’re thinking "David is dealing in the past again!" - you would be wrong in your thoughts : have you not noticed how over last few years it is very common to have the ampli inside/integrated in the loudspeaker cabinet? "self-powered monitors" as commonly termed. And again, you'd be wrong if you thought "great new/modern concept due/possible only to miraculous SS amplis!"... Because this concept has been used as long as I have been in the industry (50 years!) with tube amplis fully installed (in a separate compartment, normally underneath - like where we use stands/pieds today) in proper professional studio usage as traditionally correct at EMI/DECCA/DGG/BBC etc.

About the cascade design

Now continuing with both solid state (not OP-AMP for the moment) and tubes, but now reverting to pre-amplis use, again the two systems apply.....in an SE/LU way of thinking we call it cascade and cascOde whether SS or tubes, as I mentioned at the beginning... This is the comparison of push-pull in an output usage but called 'differential' or 'symetrique' when used in preamplis or front-end of main amplis. (I exclude op-amps because they are all differential in the input....but not purely balanced in exacts terms (in my opinion!) because their enforced use of vast quantities of NFB/CR has to be returned to 'one leg/pole' of the supposedly balanced input...)

So I am referring to discrete 3-pin transistors - which can also be Field Effect Transistors - when used in a fully symetric mode as can be equally executed with tubes... And most commonly WITH an input transformer. Op-amps can, and frequently are, used 'transformerless' in both input and output - which gives rise to the terms "electronically balanced" and "transformerless" outputs most often found in condenser microphones of which I, and many professional recording engineers I know, are disctingly not enamoured... Again it is a case of cost-saving 'pumped up' in marketing-speak so as to appear 'more modern and better' - If you can find a similar pair of XYZ brand of condenser micros - one being 'transformer-coupled output' and the other being 'transformerless'... Try them both with long cables and judge for yourself please.

The 'sound' of differential is distinctly different from that of cascade... But this is truly a case of preference as it affects different micros in different acoustics with different instruments/voices. Broadly speaking, the symetric will handle even bigger swings of input signal and also gives a wider 'stereo stage'... But there is no downside to cascade in a preampli sense such as the veritable downside to SE/LU output circuits as discussed in this article. We will discuss further under title of "micro-preamplis" later in this series, I promise.

But in the next article we will continue with power amplis and will go further into the advantages and DIS-advantages of both types of amplis....SS and TUBE....until then 'au revoir' from David.

David Manley, Paris

Voir le site de David Manley Paris