TelluriumQ Iridium: 20W power amp
Class A transistor amplifiers are a fascinating leit-motif in the big story of hi-fi amplifiers. The group consists of implementations, which are characterized by tube amp -like simplicity and zero feedback. These include such amps as Jean Hiraga Kaneda, Hawk Audio A18 or even Plinius SA-Reference.
Among Class A amplifiers are amps whose power amp stage contains a non-switching bias structure with which the transistor is forced to remain on at a small bias current, and to be conductive over the entire signal sequence. Threshold S150 Stasis was such an amp.
There are Class A designs, in which the power supply, operating as an active load, is controllable and the supply voltage constant, but the current is modulated depending on what the load requires. The purpose is to keep the output stage all the time in Class A, not letting the continuous bias current become too high. Pass Aleph is such an amp, Sugden A21 as well.
The majority of Class A amplifiers are, however, usual push-pull amps in which bias current has been increased until the amplifier works in Class A, thereby decreasing crossover/inter-modulation distortion. By making the amp operate in Class A other benefits are achieved too such as a smoother power supply load, improved linearity of the transistors as well as better
transconductance of the transistors.
Such Class A push-pull amps work in Class A up to a particular power point (load current), determined by the bias, after which the amp continues to operate in class AB. The difference between a 'normal' Class AB amplifier and an AB amp biased into Class A is that the tolerances of the components are designed to withstand a big idle current.
Musical Fidelity's "Class A" cult amplifier A1 serves as an example. The 20 watt amplifier produced 8 watts into 8 ohms in Class A, but then moved quietly to Class AB operation, despite the objections of the designer Tim de Paravicini. Nevertheless, it was an excellent amplifier. And run terribly hot!
If a push-pull amplifier is "fairly" biased into Class A, it might be considered as a genuine Class A amp. If the amplifier operates in Class A only, say, up to 2-3 watts, there’s hardly point in marketing it, as is sometimes seen, as a Class A amplifier, despite the fact that listening to music takes place mainly at low power levels.
Single-ended amplifiers operate, by definition, in Class A. There seem to be no end to the 100y triumph of single-ended tube amplifiers, but commercial transistor amps that use one-sided single-end topology are still quite rare.
Nelson Pass’s some designs are perhaps the most famous ones. In Jean Hiraga’s Nemesis amplifier only the tube had been replaced with transistors. The latest daredevil is the British cable company TelluriumQ with its single-end Iridium Power Amplifier.
The advantages of a single-ended structure are very much the same as the advantages of Class A amplifiers in general. The disadvantages are also typical to Class A amplifiers: poor efficiency and high idle power, which makes the amp work terribly hot. Single-end amps which do not heat up appreciably are noticeably low in power.
For the Iridium’s designer Colin Wonfor the power supply ripple is a paradigm problem in single-ended designs. Wonfor’s recipe to clean-up the ripple is a wise use of filter capacitors and the power supply regulation. At the same time efforts have been made to improve the amplifier’s noise ratio and transient accuracy, as well as to extend the dynamics of the music playback.
To avoid phase shift problems and to better the bass performance Wonfor has removed dc blocking capacitor at the output. It has been replaced with the power supply dc-servo, which keeps the dc as close as possible to zero at the output and allows OTL-like straight-dc connection to loudspeakers.
Wonfor also speaks of Iridium’s "tracking power supply”. The term generally refers to a structure, where the power supply voltage closely follows the signal voltage in order to obtain Class A performance with less power loss and low bias. There are other possible explanations too.
The layout inside the Iridium is dual-mono. Both segments have their own mains transformer. In order to minimize crosstalk and to protect against ground loops the controlling signals are opto separated.
The signal amplification has been carried out, inter alia, with power 55A MOSFETs. According to the manufacturer Iridium delivers 20 watts into 8 ohms with less than 1% distortion just before clipping. The amount of feedback is minimized eg. by means of an open loop gain.
TelluriumQ claims frequency response (15W/8ohm) 5Hz - 200kHz / ± 0.1dB and 1Hz - 500kHz / - 3dB. Phase shift between 20Hz - 100KHz is said to be equal to 0º. The figures are impressive, but not impossible. For a 20W power amp, the 22kg Iiridium has a good size: 30cm x 43cm x 22cm. The front panel is of 5mm aluminum, the rest of the case is made of black painted steel. The connectors are of good quality.
Despite certain soft power start precautions, there were a few nasty snaps when powering up the Iridium. Iridium has massive heat sinks but still run very, very hot. But that, as said, is typical to this type of designs, and there was no reason to doubt that Iridium would not accept such high temperatures.
There was a small internal operating noise coming from inside the Iridium. I could hear it from a distance of two three meters but never during listening sessions.
The single-ended construction does not make a transistor amplifier sound like a tube amp, no matter how much manufacturers might secretly wish it. The tubes and the transistors have such structural and electronic differences that, regardless of the topology of the power stage, the sound tends to be different.
Iridium did not change my view in this regard. I still think that it is unfair to single-end transistor amps to compare them with the no-nonsense SET amps or even well-made Class A/Class AB push-pull tube amps. My 6L6G SE and 6V6G PP amps, for example, sounded both slightly more "holographic" as well as "softer" than the Iridium. And with “softness” I'm not referring to the tonal balance or the possible differences in the frequency response.
Iridium's voice was more powerful and effective, slightly more forward-projecting and very low-noise, very quiet: as if the sound came “from nowhere". The middle range was nicely presented, but not in the way SET amplifiers do it. Iridium was not lame in bass reproduction, but also not significantly more kicking than, for example, my 6V6G PP amp.
A more relevant point of reference for a single-ended transistor amp would be a standard push-pull transistor amplifier, either one biased into Class A or just technically and vocally strong Class AB performers.
For the purpose I borrowed a similarly priced Accuphase integrated amp whose output stage and power supply have a lot of technical subtleties. I connecteded the Iridium to Accuphase’s pre out, and listened alternately to both the power stage of the integrated amp and the Iridium.
Donald Fagen’s "The Nightfly" album via the Iridium sounded "cleaner", with less background noise, and more accurately locating the sound sources on the lateral plane. Nana Mouskouri’s voice while she was “singing for freedom" was breathing better with the help of Iridium. Accuphase’s sound was less delicate, Iridium sounded more sensitive. There was no big difference in the depth of the sound or in the overall tonal balance: both sounded risk-free neutral.
Karajan/Berlin Philharmonics performing the Overture of The Mastersingers of Nuremberg, the Iridium kept the instrument groups distinct from each other, and generally followed the internal fluctuations of the music sharper than the integrated. Kunzel/Cincinnati Pops Orchestra's playing the James Bond theme, the Iridiuim had a good transient spreed but the Accuphase was more powerful. Mary Black's "No Frontiers" album sounded less grainy through the Iridium, and her voice was more subtle than when the Accuphase supported it.
Again, the differences were not shaking, and Iridium's sound had a certain understandable weakness compared to that of the Accuphase. Yet, the Iridium brought to the music some positive additional value so that if I was asked which one would I have chosen, it would have been the Accuphase preamp section together with the TelluriumQ Iridium power amp.
TelliriumQ Iridium was not the first single-ended transistor amplifier that I was testing: I’ve heard a couple of them. Generalizations are always dangerous, but for some strange reason the SE transistor amps tend to sound, to my ears at least, in a way that appears to distinguish them sonically from other amplifier types.
One unifying and truly captivating feature is the feeling of complete lack of noise, which is related to but not the same as the immateriality of the sound of best tube amplifiers. The second is the ability to identify sound sources within the sound stage, at the same time emphasizing the details of the recording. And this is exactly how the TelluriumQ's Iridium behaved during my listening sessions. If these features are something that you value more than others in an amp, Iridium is an excellent choice. The price is, however, quite hard 6000 euros. In order to get the maximum out of Iridium, everything else in the system must be up to it.
For example, the preamp part of an integrated amplifier hardly does justice to the Iridium. I would also be careful with passive controllers, because the sensitivity is only 1V. However, I’d not exclude the possibility that a high quality CD player/pre amp could succeed with Iridium.
Still, a separate active preamp with 10-20 dB of gain would be the ideal choice. Iridium’s low 10 kilo-ohm input impedance can cause technical problems for some tube preamps, but sound-wise a modern tube preamp would go fine with the Iridium.
Twenty watts is a handsome number of watts for a single-end transistor amp. For some reason, Iridium didn’t sound as powerful as I’ve used to with 20-watt push-pull amps (4-8 ohms). Rather, it drove the speakers like a ten-watter tube amp, differences in sensitivity accounted for. It’s always hard to tell which speaker goes well with a particular amplifier but this time I’d recommend quite sensitive speakers, and such that the impedance does not drop down below 4 ohms.