Valves versus transistors

"Not all of the considerations of valves versus transistors relate solely to performance. It is worth bearing in mind that products involving obsolete technology will be disproportionately expensive, difficult to obtain and possibly of inferior quality.

Valves can also vary in operating characteristics from sample to sample - especially where two valves of the same type are obtained from different sources. Characteristics that can vary are mutual conductance, operating grid bias, anode current impedance, and even usable anode voltage.

By comparison, the performance characteristics of, say, a range of 2N3055 epitaxial base output transistors are almost identical, whether made in the Philippines or in Toulouse.

Again, all valves deteriorate in use, exhibiting a gradual loss of cathode emission over a typical 3000 hour service life. If a valve is persistently over-driven, the heating of the anode may cause the metal to out-gas. This impairs the vacuum essential to proper operation, and shortens the valves' life.

A further consideration is that valves are high voltage devices, which can be dangerous. And the need for high working voltages can lead to more rapid failure of other components in the circuit - especially capacitors."

John Linsley Hood, Electronics World, September '96

Concurring with all of the above, many designs lack a good output transformer which is vital to a valve amplifiers' 'sound', a factor ignored by the majority of users, being a major source of transfer and frequency response non-linearities. Take for an example, the 'Williamson' design (below).

This had a frequency response from 20 to 200kHz giving an output of 10 watts, with 0.1% THD @ 15W, over most of the range, a hard act to follow - in 1947/9. A fundamental component of this amplifier was the output transformer designed by Dr Partridge (a 'Transistor driver for valve amplifiers' that considered this design was proposed by Seth Berglund, WW, April 1976). Some 'modern' designs encountered can experience difficulty reaching some 5kHz, let alone 10, 20 or 200. One design included a 14kHz filter before the driver stage to prevent oscillation. Guitarists readily compensate for these deficiencies with tone controls which add further colouration, whilst 'audiophiles' appear unaware of any deficiencies.

Linsley Hood also notes that;
"In general, the transformer is the most difficult and expensive part of the system to design and construct. This is because of the following conflicting demands:

For a low leakage reactance combining both leakage inductance and inter-winding capacitance from the primary to the secondary windings, to avoid loss or impairment of high frequency signal components.
For a low level of leakage inductance from one half of the primary to the other, to reduce the discontinuities due to push-pull operation, and the odd-order harmonic distortion resulting from these.
For a high primary inductance, to give a good low-frequency response.
For a low winding resistance, to avoid power losses.
For a good quality grade of core laminations to ensure a low level of core-induced distortion, due to magnetic hysteresis and similar effects."

It is not necessarily the use of an output transformer that limits the audible range, but its' quality of construction. A reasonable frequency response of say 25Hz to 75kHz ±3dB @ 20Wrms and <10Hz to >90kHz ±1dB @ 1Wrms with a natural resonance of 80kHz can be expected from a decent transformer. Solid-state amplifiers driving 100V lines in critical areas might inject some 20kHz into the signal, the speaker loop then being fed back to a detector circuit which monitors the loop for open circuits by an absence of the 20kHz tone. If the output transformer is sound then, since the manufacturer is likely to source from the same company, the mains transformer, another critical component, might be considered sound as well.

Low power single-ended outputs gave demonstrably better performance, however, to prevent driving HT voltages to prohibitive levels with outputs exceeding only a few watts, push-pull outputs (as seen above) were required. This relatively sophisticated 1953/4 model demonstrates what has been described as an 'ideal musical standard'.

Of the same vintage, a push-pull 'hi-fi' approach, the more sohisticated 9-valve RGD ARG510 auto-radiogramophone, opted for conventional windings that were, however, screened. Apart from a four-waveband AM tuner, this unit used the Garrard RC75 auto-changer fitted with the HGP33C turnover crystal head (cartridge).

Note the cross-over and arrangement for multiple speakers not normally then seen in domestic audio, although the polarity of speakers appears to have been considered unimportant at this time (this is still the case today with some upright organ manufacturers, although with some cabinets the phase arrangements are found empirically to give the best subjective results in a showroom). Similarly, the application of negative feedback was rare (as in this case), although some manufacturers provided this via its' own output transformer winding, like the Philips, adding complexity and thus reducing interchangability.

With the advent of stereo, some record and tape manufacturers did not understand the importance of phasing correctly. Some recordings would change phase from track to track, even with those intended for demonstration purposes! A simple DPDT switch allowing a speakers' connections to be reversed could be used by an enthusiast to remove the deficiency in bass reproduction and the 'hole in the middle' of the soundfield. Most people however, including the experts, were unaware of any problem.

The importance of the output transformers' colourations has been acknowledged in some designs, like the 1987 Seymour Duncan Convertible (below),

although it can be seen that the inclusion of screen grid taps (usually 43% from CT for a distributed-load or "ultra-linear" configuration) can introduce difficulties in the supply of replacements. Although necessary, switched HT lines can also be incredibly noisy, or even cut-out altogether. How many guitarists (and others!) have relied on hearing the 'mush' that comes out of a valve-driven speaker to inform them that all is well?

Varying the damping, some users found interesting and useful, although it must be realised that a high quality valve output transformer might quote a damping factor of about 50. Today, this has to compare with some domestic solid-state transformerless designs that can offer a far superior 700, or 14 times better! Even ICs offering similar power outputs can offer some 4½ times better, even at 20kHz.

As with solid-state amplifiers, power supplies are important. Much of the hum inherent in many valve designs is caused by inadequate smoothing and screening. The 575V HT in the Seymour Duncan above was smoothed with a generous 100µF (700V), a microscopic amount by solid-state standards. Some early 'hi-fi' designs (eg; Ferguson 500RG) coupled the gramophone pick-up with a capacitor rated at 1kV! (newly arrived 'microgroove' records had yet to see a transistor). Considerate manufacturers would fit screens for low-level valves.

Valve heater supplies can be centre-tapped to ground or rectified, smoothed and regulated to assist hum reduction. Ramping these (on power up and down) will reduce incidence of heater failure, some designs employing thermistors. An inadequate supply will result in 'overdrive' more easily than a meatier one, most guitarists ignoring completely any concept of dynamic range, the most, if only, meaningful specification sought being the output power.

One constructor had this, amongst other things, to say about the Williamsons';

'This supply had serious deficiencies. The audio circuit had a direct-coupled first stage. On warm-up, the B+ supply initiated high voltage to the bus before the output tubes warmed up. This caused a very positive voltage to appear on the grid of the second stage during warm-up. This caused current to flow from the second stage cathode to the second stage grid. The likelihood of grid damage was high during warm-up as grid wires are very small and won't carry much current.

When the Williamson was turned off, it fed a large very low frequency power pulse to the speaker. Many Hi Fi speakers couldn't take this very low frequency pulse and blew out. This turn off power pulse showed that the power supply was unstable, poorly decoupled, poorly regulated and prone to motor boating, a very nasty instability problem. The capacitor-input system rendered marginal the use of the first choke as a filter element. Capacitor input systems defeat the advantage of using chokes in a power supply.'

With a valve amplifiers' HT at some hundreds of volts above the mains in, it is obvious that any incoming spikes will be similarly increased, then stressing the other components in the supply, the smoothing caps being contenders to fail first. Experience with these and the annoying variance in valve spec, which could mean that a valve would work in one unit and not in another, led to a welcome embrace for solid-state, limited though germanium was at that time ("Look, no heaters or output transformers!"). Although a matching transformer often drove a germanium output pair, much-welcomed transformerless designs soon appeared.

Users often had to accept other instabilities that were built into a design. Taking the Ferguson 500RG mentioned above, one manual stated 'the rated output is 14 watts, and this will generally be reached at about position 5 of the volume control; over-driving will cause severe distortion, and the onset of this will be sudden'. A shame when trouble had been taken to include multiple speakers, cross-over and an airtight enclosure, all rare for the time, and a single resistor could cure it.

At the same time, the use of high-value (500k+) potentiometers could introduce problems because of the stray capacitances between the wiper and the tracks' ends. At low volume levels, a few picofarads across the upper arm could boost the treble response by several dB at 12kHz or so, or introduce a 'spikiness' to square waves. Clearly audible, and visible on a scope, this phenomenon is undetectable by the majority of valve enthusiasts encountered.

All electronic circuitry is stressed when it functions, the degree of which is determined by the technology used. The level of voltage and thermal stress that must occur before a valve configuration will function can be profound.

Measurements of output / power consumption, including supply losses, have shown efficiencies of about 6% for some valve designs compared to 15-25% for class A solid-state designs, 36-54% for class AB, B and >70% for class D, when driven at similar levels. The inclusion of a heavy piece of metal in the shape of an output transformer can add significant cost and weight. Power consumption can become important where supply is limited.

Some custom valve designs seen (especially those built for reggae bass) employing multiple arrays of KT88s and 1kVA+ output transformers, can be very impressive to look at, when made for this (see the Carver Silver Seven). Noticeable, is the propensity to dim street-lamps for miles around when switched on (it was felt then that the only current limiting in the entire system was that represented by the impedance of the mains supply!). However, given the choice, it is felt that the 'good' attributes applied to the sound produced are due more to the speakers used, rather than the amplifier.

The high current slewing conditions when transient signals are fed into difficult loads can produce higher order harmonics as some solid-state output stages 'catch-up' with the demands made. Caused by a transistors' parameters inconvenient habit of changing with collector current, these tend to 'harden' the overall sound. The gradual non-linearities in class A designs where the output devices continuously operate in the linear collector region are held to produce lower order (2nd and 3rd) harmonics commonly found in musical instruments which are easier on the ear, as Linsley Hood found with his 1969 class A design. However, because of the high quiescent currents involved, quite substantial hardware arrangements are required for outputs of 10W or more, compared to class B operation.

Below is a comparison, in general, of the limiting characteristics of both soft-limiting thermionic and hard-limiting bipolar amplifiers. Opinions exist that hold these curves to be the whole truth as to why valve amplifiers are preferred by many guitarists.

What appears to predominate is the desire to hear the distortion caused when hardware is in the process of being stressed and irreparably damaged - 'there's nothing like the sound of hot valves (and transformers!) - which, in abundance, can mask poor technique. This tendency, or habit, probably arose from the use of the excellent Vox AC30 (rated at about 30W, with a distinctive sound all of its' own) and the Marshall 100-watt 'heads' (then the largest amplifiers around) which could be overdriven to an alarming degree (as with any thermionic device compared to a semiconductor type).

Sometimes, as seen above, distortion was deliberately introduced as a design feature. Jim Marshalls' designs had a 'dirtiness' that 'clean' American amplifiers did not, for example. The 'feedback' sought after by some artists could often result from a valve stages' microphony, especially if an amp was placed on top of a 'speaker stack', thus the popularity of the Marshall arrangement for 'rock'. Given this combination, it can be argued that it is not just the guitar which is the musical instrument. Attempts have been made to introduce 'overdrive' by reducing heater current, but this met limited success due to the hysteresis inherent in the heater coils' temperature.

Some combo users sensitive to 'the valve sound' may oddly remain unaware of the parlous state of their speakers' cones or the fact that vital links in the audio chain such as mixers, effects and high power amplifiers are exclusively solid-state, thus implying that all of these last, without exception, are acoustically transparent, ie: perfect. Similarly, American constructors tend to prefer the foreign EL34 and the Europeans the 6CA7. The Russian Svetlana factory that manufactures a modern EL34 equivalent does not distinguish between the two. Since, clearly, a constructors' ear can be expected to have suffered comparably less damage than that of, say, a rock guitarist, there can be no tangible or audible difference. Combos and amps in the UK used to be 'valved', now, as the Americans say, they are 'tubed'.

In the authors' view the EL34, although of a higher electrical spec, was not as physically robust as say the later EL84. Although the American 6L6, the later (Phillips?) 6L6GB or the 6V6 were 'tougher' and less prone to microphony, it is generally held that the 'European' ones sounded better.

It has also been noted, how little or no advice is offered to a purchaser of a new combo/amplifier with regards to the fragility of valves when hot. One should consider what happens when, after a gig, hot gear is piled into the back of a van which then hits a bump making the load temporarily airborne. The retaining screens and springs often seen in quality portable valve gear are frequently missing from PA designs. When and where possible, the author has used mattresses on a vehicle floor to help protect equipment, which can also prove useful in other ways.

Similarly, although users might be aware of the dire consequences of attempting to drive a short-circuit with a solid-state amplifier, few if any will know how hot valves can get without a load. This can be important if multiple speakers, say in different rooms, are switchable without load resistors.

In blind tests, where 'the valve sound' was explored, an harmonic enhancer (high pass filter of 1-5kHz mixed into the original signal, <10%) gave that 'unmistakable' warmer and mellow 'richness' with any solid-state design used (perhaps super-fi manufacturers could consider the inclusion of a 'niceness' knob!). The author invites suggestions to test 'the transformer sound'.

It was a joke that in military circles artillerymen could be identified by their shouted conversations. The same can safely be said about some musicians. One well-known vocalist could not hear himself without the harmonics of clipping. Another, a self-styled aficionado, waxed lyrical and at length of the distinctive qualites of 'the valve sound' emanating from an antique record player unaware that the unit he had listened to was in fact clockwork and thus devoid of any electrical components. It is ventured that what was really being admired was the bass presence given by a substantial wooden cabinet (a presence between 64-128Hz gives a 'warmth and body' to sound).

The (1965?) Decca RP205, a three stage two valve record player was a case in point. The second 'microgroove' record player my family owned, this could impress with its' detail, loudness and bass when the heavy and padded lid was lowered, despite the relative simplicity that was not reflected in the cost (expensive Decca Deram ceramic cartridge). Although lacking a radiogrammes' deep bass it made other later makes, like Dansette, sound simply awful. As soon as an open-reel arrived, the Decca was modded so that the tape deck could play back through the larger speaker.

This unit was still functional after more than 40 years and been hammered by two teenagers. Notable feature, it did not feedback. Use with a new album would have been to place the record on the platter leaving the autochanger arm in the up position, the record then replaying until turned over.

The cabinets of early semiconductor designs, made of 'modern' plastic, resonated badly contributing to the 'cheap and tinny' sound produced, thus the euphemism 'tinny transistor', and association with poor quality sound. The influx of cheap Japanese units also suffered from a denigration arising from Western post-war opinion. Mass production of injection-moulded plastic parts was new and not understood as it is today. Some speaker baffle boards were no more than hardboard. If however, the internal speaker could be bypassed to a substantial external one using, say, the earphone socket, results could surprise.

At the same time, in early solid-state, there was a reliance on previous valve design topologies (push-pull) utilising transformers for stage coupling, phase-splitting and no feedback. These 'matching' transformers could have a usable frequency response of some 200Hz to 15kHz, or less, at trifling output levels.

The limited maximum temperature range of germanium (60-100°C, compared to 125-200°C for silicon) and commercial electrolytic caps meant that many portable radios, placed on the back window sills of cars, 'melted'. However, the advantages of enhanced portability and instant warm-up over-rode prejudices and solid-state took off.

Cross-over and harmonic distortion, damping factors and noise figures of early solid-state amplifiers did merit improvement, however the valve-biased fraternity who poured derision on these attempts ignored deficiencies in signal sources, such as AM tuners and record player cartridges, let alone speakers. To put the technology into perspective, one designer found that using a Decca "ffss" MkI cartridge with a Decca SXL 2057 test record produced a channel difference of only 6dB @ 2kHz when only one channel contained recorded information. This was at a time when the UKs' considerable lead in audio and 'high-fidelity' design included the countrys' FM network and that 'thermionic' technology had had some 50 years to mature.

Many were comforted by a bulb-lit dial that gave nostalgic warmth on lonely nights. Many were also to lack the vision of what an oblong box with nothing but a row of switches and lights would be doing one day.

The post-WWII commercial boom benefited from many valuable lessons learned about the construction of rugged valved gear. Almost invariably, these used tag-strips to link the components and valve sockets with wires and other external switches, controls, pots, etc, sometimes called "scramble" or "point to point" wiring. Some found it hard to cope with the sudden miniaturisation and the need for new mounting and wiring techniques, like PCBs.

Commercial designs can be very cheap and decidedly uncheerful and this aspect was the principal factor in determining the authors' attitude to valve designs, even towards very well-known makes intended for PA or stage use. At least one manufacturer started with good designs, but then allowed their production team to remove one component after another till the prototype stopped working. The last component removed was then replaced. Voltage and power ratings were then pared to the minimum possible, the components then been sourced from the cheapest supplier. Another, otherwise sensibly built, popular and very well-known make permanently damaged their (hefty) mains transformers before fitting them, thus guaranteeing an early demise.

It can be seen that, possibly in keeping with rocks' 'hard' image, it might be fashionable if an amp appeared to have been built in a garage, with an attendant complete ignorance of fundamental electronic principles!

A few modern designs offer a very high-quality build compared to some commercial efforts, employing toroids and soft-starts, for example. However, these can be surprisingly complex (see the Velleman K8010/1 kits and Troels Gravesens' Ruska RIAA, for example). Those, where the valves are left exposed and unshielded, are often simply intended as eye-catching centrepieces and are neither child or drunk-proof, or mobile. A smaller (1979) design, the Brimar 25P1 25W amplifier, could be considered.

Few, even amongst professionals, have perfect pitch. For those who do, the music world must be an awful place. Prolonged exposure to high sound pressure levels inevitably results in permanent ear damage, as witnessed by the ringing in ones' ears after a loud concert. Sensitivity to higher frequencies diminishes with age, and no one can guarantee that at any given time their auditory faculties are certified to traceable standard to be in perfect working order. Add to that the brains' ability to interact with the inner ear, thus subjectively reducing extraneous noise, and it can be seen that hearing is mostly in the brain of the listener (see Sound). Just like sight.

An individuals' attribution of 'valve sound' qualities to solid-state designs is invariably independent of the circuit topologies employed, though differing widely. For example, a class A solid-state design will be compared with a push-pull thermionic one (JLH class A), the use of FETs may be discussed (class B Yamaha B-1) or the use of feedback techniques (Quad 405). JFETs can exhibit a Vgs curve 'kind of like a triode' at low Id and Pass has demonstrated (Zen variations) class A designs that can output 1W with virtually no regard for efficiency. Invariably, however, there will be an element of intangibility that blind tests can dispel.

Perhaps this explains why, to a valve-audiophile, it has no bearing whatsoever on the audible outcome whether a design is single-ended or push-pull (although some will know the difference, but cannot distinguish between them) and why then that the limitations of all known output transformers, simply by virtue of being driven by thermionic means, automatically become acoustically transparent and therefore perfect! Unfortunately, for this line of reasoning, there exists a concept known as 'the real world'.

Over the years, experience has suggested that often some of the most sweetest sounds have come from outputs driven by a single (and large) transistor, especially if the speaker load is 'notorious' for its' unstable nature, and reflections.

Valves had their time (the term thermionic being mooted in 1909) and were necessary, but even on an environmental basis can today be deemed ethically unacceptable. Such richness of digital manipulation is available that the power previously consumed can now largely be used for something else.

Other observations

Class D amplifiers

Contact me at
especially if you want additional content to this page
or if you find any links that don't work. Don't forget
to add the page title or URL. Take care!

Google search for John Linsley Hood.

Some JLH designs.
Back to index, sound, tips or home.