The JLH Gateway to Class-A :: 3
Building the JLH Classic -- Milestones of the Era
Happily the JLH Class-A was not just another 'typical design' of the period -- it had nothing that went before it, and on which it was based. Not only that, the design had many unique features despite appearing to be quite basic. Suffice to say, its apparent simplicity hid within itself many subtleties of design not easily revealed to the casual investigator or circuit hacker. No wonder it has defied the efforts of many talented designers to 'pin it down' and analyze it threadbare. GOM and old timer Nigel Pearson has put it wonderfully thus: "It makes my blood boil when people say "JLH was not bad for it's day". It's excellent ... and that's the only truth worth saying..." The design JLH had finalized was not the result of casual thought, but of deep reflection, inspired creativity and absolute originality.
One should be prepared to delve somewhat deeper into the state of audio electronics and the prevalent design thinking of the late 1950s and 1960s before one could visualize the proper context and background in which the JLH classic amplifier should be viewed and assessed.
Design Evolutions
The late 1950s saw the emergence of transistors widely in commercial products, to an extent marking the beginning of the end of the monopoly of thermionic valves/tubes. These devices were all fabricated with Germanium, with its attendant thermal limitations. If the diffused junction regions of the (mostly PNP only) transistors got 'sufficiently hot', that would lead to the diffusion process itself to continue within the device, altering its characteristics alarmingly, leading to instant failure. (As one wag put it, "... if the device was hotter than a cup of tepid tea, well, it was bye-bye ...") Amplification continued to be the domain of the sturdy valves/tubes and they were the "hot favourites" of 'hi-fi' enthusiasts.
 |
| Germanium Transistor Radio Circuit |
But very soon, by and large, the new-fangled transistorized portable radios jostled their cousins, the heavy-weight table radios, off the domestic market on either side of the Atlantic. Even makers of 'pro' class "explorer" radios like Zenith (Trans-Oceanic) in the US and Grundig and Nordmende etc in the Continent moved from miniature battery valves/tubes to "solid state" devices. (The term 'solid state' , as we know it and refer to it today, had not emerged till then, nor had it begun to assume any 'weight' in the electronics field.) Even the humble teen DIYer had his "standard" 6-transistor radio kit, putting out about 100 -- 200 mW of audio power driving a small 2 inch loudspeaker. In no time we witnessed a flood of small personal "pocket radios", which arguably reached its smallest size and peak of performance with the Sinclair (UK) 'Micromatic' of the late 1960s.
 |
| Germanium Power Amp featuring AD149 |
Research was proceeding apace and we got to see Germanium power devices like the AD149 and AD161/162 (wonder of wonders ... a complementary pair!) from the likes of Mullard (UK subsidiary of Philips) and others. But amplifier design, whether it was for a flea power portable radio, or a "high power" (all of 4 Watts!) amplifier, largely followed the framework of established tube design practices. They invariably had a driver stage transformer and an output transformer, and as a result quality per se was neither here nor there.
 |
| Germanium Amplifier with input and output transformers |
The Solid State Era
But the 1960s saw the emergence of a full range of silicon bipolar transistors, and their better specs and performance (chiefly higher working voltages and temperatures) made them the newfound loves of designers. The stage was set for the emergence of the modern era amplifiers. But in those early years the very appearance of even "cutting edge" amplifier designs borrowed much from their valve/tube counterparts.
 |
| Typical Valve Amplifier |
Take a look at the "Brute-70', a solid state amplifier design for advanced DIYers; at first glance, it was not easy to differentiate it from a typical valve amp of the era. But slowly all that changed, and solid state circuit topologies established their strong bases. But not before a slew of insightful designers, including JLH, put paid to the prevalent notions of "transistor sound" with their iconic circuit topologies.
 |
| An Early Solid State Amp |
As most of us somewhat long in the tooth are sure to recall, the "modern" transformerless amplifier topology owed a lot to the circuit framework proposed by H.C. Lin, a researcher at RCA in 1956. The design had three distinct stages: an input stage, a voltage amplifier stage (VAS) and an output stage, which evolved to become the dominant circuit topology in modern times. He employed what was called a "quasi-complementary pair" of output devices to drive a loudspeaker directly -- the upper pair was a conventional Darlington, while the lower set of devices comprised a compound emitter- follower.
 |
| The Lin Amp Topology |
From then on modern amplifiers have followed this topology broadly. Evidently the asymmetry inherent in this design did result in a slew of issues which were subsequently addressed by designers. But till that happened, there were reports of 'listener fatigue' (when compared to valve amplifiers), and very soon two divergent camps emerged, with the established valve/tube aficionados alleging that the "transistor sound" was thin and unrealistic. Of course, even in the case of amplifier designs of the period that did measure well on par with some of the best valve gear, in subjective assessments of "fidelity" they fared poorly. Many designers who had assiduously followed this lack of performance soon discovered the weakness of the solid state topologies and endeavoured to correct them. Names of JLH and Peter Walker (Quad designer) need particular mention in this regard.
 |
| The Leak Amplifier |
A typical solid state amplifier with silicon devices and acceptable performance of the time was the Leak Delta-70, a much-imitated commercially successful design of the mid 1970s. While class-A amplifers did exhibit deterioration of THD at the upper end of their power limit, their performance at low power levels was sterling. Not so with most of the class-B designs; they performed and measured rather well at higher power levels, but as output power was reduced, their distortion figures climbed. (Today it is a given that it is the "First Watt" that really counts.) Again, tube amplifiers employed only modest amounts of NFB, while the solid state ampilifers had it in loads in order to improve their specs. This also ultimately created its share of issues in subjective quality and performance. Designers were only becoming aware of these rather strange phenomena, and were grappling with circuit techniques to tame these irritants.
Adding to these ills were the instabilities that resulted when the amplifier was tasked with driving a "real world" loudspeaker, a highly complex and temperamental load, instead of a fixed "dummy load" resistor in the lab. Amplifiers that were stable and 'perfect' on the test bench, when playing back music often punctuated the signal with bursts of high frequency oscillation, not in itself often audible, but which did irritate the ears of the listeners.
 |
| The Peter Walker Triplet |
Stalwarts of Design
Improving the symmetry of the output pair was the path taken by Peter Walker, whose design the Quad-303 power amplifier WW, Apr 1968) displayed near perfect symmetry with output triplets. JLH had also proposed a design that employed output triplet transistors with much improved symmetry. The JLH triplets were virtually identical under DC or AC conditions, and the optimum quiescent currents for the upper and lower sets were identical. Moreover the THD performance approached thresholds of measurability at the time over the full power band from about 10 mW to the maximum output at 30 W. It was indeed a milestone in amplifier design!
 |
| The JLH Triplet |
Later I.M. Shaw and Peter Baxandall (of the Baxandall tone control fame) proposed a simpler approach to achieve symmetry by employing an additional diode. JLH too liked this simpler approach, and he was able to improve the performance of the circuit with the addition of a parallel capacitor t osimulate to simulate the output device's junction capacitance. This very effective topology was presented as a 75 W DIY project in 1972, which broke some records for popularity. What perhaps is more important for JLH fans is the fact that the circuit employed many innovative features which have become the hallmark of most modern amplifier designs. The amplifier gathered accolades for its measured and subjective quality and its load agnostic nature and unconditional stability.
 |
| The Shaw Diode Circuit |
 |
| The Baxandall Diode Circuit |
 |
| The JLH Diode Version |
JLH and his Class-A Amplifier
The complexity and the inherent drawbacks of various designs of the period had probably influenced JLH in deciding upon a simple Class-A design. He had initially designed and built it for his own enjoyment, but succembed to gentle persuasion and sent it for publication, as he had clarified at the time , "... because the use of output transistors in Class-A had become a matter of topical interest ...principally because commercial designs like the Sugden Class-A amplifier had attracted favourable reviews in the hi-fi press, who applauded the design for its freedom from "transistor sound" ..." And that was how the JLH Classic saw the light of day in the pages of Wireless World in April, 1969.
The design, with its disarming simplicity, attracted a huge following among the DIY community. Built with off-the-shelf 'standard' parts, its subjective sound quality put to shame many commercial designs with "high end" pretenses.
JLH explains his classic design thus:
"The structure of the circuit is very simple, with Q1 acting as a grounded-emitter amplifier stage, with Q2 as an active collector load, driven in phase opposition to Q1 by Q3. The loop gain of the amplifier is increased by bootstrapping the load resistor for Q3 by C1. Because the transition frequency of the output transistors is of the order of 4 MHz, whereas those of Q3 and Q4 are in the 400 MHz range, the circuit has an in-built dominant lad in its loop NFB characteristics. This ensures that the loop gain has fallen below unity before the loop phase angle reaches 180 degrees. No additional HF compensation networks are therefore necessary to ensure complete loop stability, even with reactive loads." What a wonderfully brief and succinct explication!
Of course, the coming years saw many competent designs that employed the newly availble PNP complements of the existing NPN output devices. A notable design in this format was that by A.R. Bailley presented in the WW, May, 1968 issue. But it has to be admitted that even when such designs approached theoretical perfection, there were significant factors affecting performance, since in spite of being nominal equivalents, PNP and NPN devices were marked by inherent differences. The most obvious was that the current carriers in NPN devices were the faster electrons, while in the PNP devices, the duty was done by comparatively slower holes. Also, PNP devices have proved in practice to be more failure prone, despite their SOAR data matching that of the NPNs. Performance difference in these situations are supposed to show up in the HF performance of amplifiers. All said, the Bailley amplifier was perhaps the first design that demonstrably could perform equally well with a square wave test signal into resitive or reactive loads, which capability, according to JLH, is an important factor that decides the overall sound quality of an audio amplifier. Another notable design of the era was the ultra-low distortion Nelson-Jones class-A amplifier (WW, March, 1970)(see links below)
 |
| Nelson-Jones Class-A Amplifier |
JLH, as is often made out by half-baked "experts", was not merely the designer of a simple amplifier that "luckily" managed to tick all the boxes. He was instrumental in leading the progress of the art and science of solid state amplifier design over the decisive decades that established solid state amplifiers on a pedestal equalling or surpassing that of the tube behemoths. His signature could be found in a host of designs and improvements, and most of which have been adopted and adapted by designers till this date.
Links to Archives under reference:
1. Bowes Germanium amplifier -- WW, Jul 1961
https://drive.google.com/file/d/1wwMnFC7xKNVw0E1SQQNDHpBhZsAdSGWN/view?usp=drive_link
2. Dinsdale power amplifier -- WW, Nov 1961
https://drive.google.com/file/d/1UBATtBWy1zhCbkiyNaSjnoW8G-dBlEqB/view?usp=drive_link
3. Dinsdale preamplifier -- WW, Dec 1961
https://drive.google.com/file/d/1aSEzQ168cLmms57R2EbVK4Ujba4dlOSk/view?usp=drive_link
4. Dinsdale amplifier - Mk-2 -- WW, Jan 1965
https://drive.google.com/file/d/1Dfe2fYaS81_Dq4cI2_6ijO8P_DmNet3T/view?usp=drive_link
4-a. Dinsdale Retrospective of his amplifier designs -- WW, Nov 1969
https://drive.google.com/file/d/1QkKE9WJhm11diIP_t25Lt-8ymykBM7RB/view?usp=sharing
5. JLH mods to Dinsdale amplifier -- WW, Feb 1970
https://drive.google.com/file/d/1-XFeJT29o3Rvy68revtMouLXOtFAe3No/view?usp=drive_link
6. Bailley 30 W amplifier -- WW, May 1968
https://drive.google.com/file/d/1xm6tagy78dm2EWEc2KC2zbyaW5Vww8cD/view?usp=drive_link
7. Nelson-Jones Class-A amplifier - WW - March 1970
https://drive.google.com/file/d/17x_caZKnEgEqxMsdKrTH6LBHHarCKI35/view?usp=sharing
In a subsequent post we shall take a detailed look at most of the published designs of John Linsley Hood that followed the publication of the timeless 'Classic'.
In the meantime, do read, comment, share -- and build and listen!!
* * * * * * * * * * * *
Comments
Post a Comment