Opera Audio 300B PSE and PP Amps

The Answer to the Ultimate Question - Single-Ended or Push-Pull?

Products: Opera Audio 300b PSE and PP Amps.
Manufacturer: Opera Audio - Made in China
Cost, approx: 9000 Euros a pair
Reviewer: Geoff Husband - TNT France
Reviewed: January, 2007

Introduction

25 years ago valve amps were a tiny backwater even in high-end hi-fi, an obsession followed by a few collectors (frequently in the Far East) and the odd tiny specialist manufacturer. Within this niche there lingered an even more obsessive bunch of 'bottle-heads', those who pursued the Holy Grail of valves - the single ended triode.

Now valve amps are still hardly mainstream, but it seems that to the serious hi-fi addict they are an equally valid alternative to the transistor amplifiers that once almost swept them from the planet, here at TNT we seem to have another on test every week. This avalanche of new designs has spawned many high-efficiency speakers to help get the most from these amps and this inevitably has increased the practicality of owning the weediest, though arguably the purest for of the breed, the single-ended (SE) triode amplifier.

But is the hype justified? SE amps have some disadvantages over other valve amps, but converts swear that the sound quality is worth it - but is that true? Are SE amps that special? Do they justify their existence? The problem is that just picking two amplifiers, one a push-pull (PP) and the other an SE tells you nothing beyond that the two amps are different - there are many possible reasons for one amps superiority beyond the intrinsic superiority of one technology over the other? Here I've done such comparisons but in a rather desultory way and with no strong conclusion beyond a feeling that SE amps do bring something special to the art - but that could just be differences in manufacture or my own wishful thinking. That said my system has been SE based for the last two years.

Many years ago AudioNote UK brought out two cheap valve amps in their Zero series (I seem to remember), one SE the other PP; the latter, it was generally accepted, was more powerful, the former better sounding, but they were different in too many ways and hardly used state-of-the-art components or valves. Inconclusive for me at least.

Then something happened that seemed a heaven sent opportunity.

I'd tested the Opera Audio CDP 5.0 CD player and the LP 5.0 record player and been impressed by both. Searching their website I found that they had introduce a series of high-end valve amps including a 300B SE that would be a perfect head-to-head comparison with my own AudioNote Quest Silvers. So the usual request to Opera and they said they'd send a pair off - two weeks later* two very heavy boxes arrived.

The contents were not what I expected, each amp having not one, but two 300B valves - a parallel 300B SE (PSE) no less, and even more exotic.

But they'd screwed up, later that week Opera emailed to say that they'd sent a pair of 300B PP (push-pull) amps by mistake. They look identical to the 300B PSE, had the same valve compliment and I had to check the wiring to be sure, but yup, they were PP's. Damn. "No worries says Mawei at Opera, would you like us to send the PSE instead?"

Cue lightbulb hovering over jaded part-time hi-fi scribbler's head:-) "Er could I keep the PP's as well so I could do a true back-to-back, PSE vs PP?"

Yessss!

So not only do I get to play with two sets of very nice amps, I also get to answer a question that's been bugging both me and a lot of other hi-fi nuts...

Theory

OK you all know my knowledge of electronics could be engraved on the top of a drawing pin with a pneumatic drill so here your not going to get an in-depth description of the differences between SE's and PP's - just do a quick search and you'll find all the details. The differences we're interested are those that may effect sound and compatibility. The latter parameter gives all the aces to the PP design, PP gives much more power, generally around twice a SE amp with the same valve complement, better power into low impedance loads and in theory a larger bandwidth as an SE "droops" at the frequency extremes. A PP also draws less current at idle (cheaper to run) and even in this case, is cheaper to buy as the transformers in particular don't have to be so specialised. SE's also don't reject power supply hum etc and so need more extensive mains filtering and they chuck out much more general distortion, though this is mostly even-order and so relatively benign.

So SE's suck? Well they have one thing up their sleeve - they are direct coupled to that expensive output transformer and draw half their rating all-the-time, it means changes in current due to signals are done very quickly and with very little distortion. It sounds good but what does it mean.

Construction

Someone at Opera (apparently he's called Liu) is an industrial designer who manages to make his creations good enough to be called works-of-art. The Droplet range, and now these amps are blessed with good looks well beyond their price bracket. Beauty is in the eye of the beholder of course, but these amps not only look gorgeous, with build quality beyond reproach, they are also original in design and practical. Lifting valve amps weighing 30 kgs is no joke, but those two long handles along the top make the job easy and offer much needed protection to the potentially lethal (and expensive) valves below. The mix of wood, brushed aluminium slabs and acrylic is beautifully done, the valves and transformer cover neatly integrated and the choice of TJ Mesh-plates (the prettiest 300B because of the glow!) inspired. To see these in the flesh, is to want them - it's as simple as that. My only moan is that the on-off switch is hidden at the back and almost impossible to reach without burning your arm on those valves - either they have to be left on all the time or you pull the mains fuse at the end of every listening session...

And they are identical... There is no way of telling them apart, they use the same chassis and valves and inside you'll see the same component quality - the difference is purely the circuit and the transformers, which although of the same quality are inevitably different. And on the subject of transformers, Opera tell me that the transformer laminations they use are from old army-surplus tanks. Imagine - in the '50's and 60's these Chinese tanks must have had hefty valve radio gear with equally robust transformers. Apart (I imagine) from their low cost these laminations will be "aged", something affectionado's of transformers say is a "good thing". They are also more likely to have been made from pure rather than recycled iron - apparently it is now virtually impossible to obtain pure iron any more, and the contaminants in recycled iron have a detrimental effect on performance. Who knows??? What I can tell you is that the transformers are big... Notably that of the 300B SE which is the purest of the breed, zero feedback, direct heated, in fact all the buzz words.

Lies damn lies and incompatibility...

The specs of both make very interesting reading. Opera's website is one of the most Hype-free I've come across, anyone looking for a philosophy course should go elsewhere. That lack of hype seems to be extended to the specification claims for the amps. As a rule of thumb manufacturers seem to claim between 10 and 12 Watts for a single 300B in SE mode and 18 - 20 for a parallel pair. Now it might be possible to justify such claim, but only if the associated distortion is much higher than normal and the measurement taken at 1000hz.

Once you build a 300B push-pull these problems reduce as power is generally around twice that of a similarly equipped SE and is more load tolerant. The "going rate" for a 300B push-pull amp seems to be around 30 Watts.

A quick look at the opera site shows them claiming only 15 Watts for the 300B SE and 20 Watts for the 300B PP. Both are much more realistic than normal, but the PP's 5 Watt advantage is minimal, it equates to about 1 extra decibel in acoustic output - inaudible and I'd suggest the figure is over conservative for once.

In use and sound quality, Push-Pull first...

So now to the crunch. Two amps from the same manufacturer, in the same chassis with the same quality of components - the only difference being that one is push-pull, the other single-ended triode. My aim was to find the difference in ultimate sound quality and so almost all the review was done using my Loth-x Polaris horns, claimed 104 dbl (so probably around 100dbl...) and designed specifically to partner a 300B SE amplifier. With these speakers I would never test much more than the first Watt. Neither amp would ever be put under even the slightest strain and so sound quality alone would be the decider. However the Polaris are very revealing, and with such efficiency any residual noise from the amps would be very noticeable. This might seem a minor point but when testing SE amps a couple of years ago I returned two untested because the mains induced buzz was so bad as to make them unusable. In this case I didn't need to worry, though the PP was very quiet, the faint hum of the PSE wasn't objectionable either. In this respect it was slightly quieter than my own Audionote Quest Silvers, though the main reason for this was that the AN's are more sensitive, and so to drag out a motoring analogy, their tick-over is faster...

As they arrived first the 300B PP's were first on the stage, in fact when I first heard them I was still under the impression they were SE amps and were judged as such - a salutary lesson...

I'm a sucker for a pretty face and so was already in love with the amps. Their great weight (30kgs) was reassuring, though it should be noted that the Quests were very nearly as heavy so equally good value in the £ per lb. equation:-)

And I wasn't disappointed. I don't look forward to taking my Quests out of circuit for reviews as it's inevitably a downgrade, but here I wasn't so sure. The general balance was quite similar, detail and soundstaging excellent and the musical event coherant and well presented. I did think they were a little "harder", a little "clattery" and mechanical in back-to-back listening but not to an objectionable level. I put this down to their newness and looked forward to some improvement in the following months.

These were initial snap observations and over the next month I lived with them on a daily basis. After a few days I learnt of the mix-up and so realised that here I had an imposter:-) To my surprise this increased my respect for the design. This was a PP that could certainly fool me into thinking it was an SE. It gave that lovely low level detail, the coherance and the hear-through midband.

But knowing now that it was a PP I pulled out my old IPL S3MTL transmission line speakers (87 dbl and an easy load). My Quests will drive these at moderate listening levels happily enough, but never with the control in the bass that my old Audion EL34 PP amps had (30 Watt per channel), and winding up the wick to party soon had them distorting horribly. The Opera 300B PP's really made these speakers sing. OK there wasn't the (cue cliche) vice-like-grip the Korato pairing had over them, but it was pretty impressive stuff - they also went as loud as I remember the Audions doing (i.e. too loud). This is interesting as the small claimed power advantage of the PP's over the PSE's shouldn't do this, again I suspect the PP's are pushing out nearer 30 than 20 watts. Certainly there's no getting away from the fact that sonically, the IPL's, good though they are, are no match for the Polaris, but it was an interesting experiment. This showed that whilst the Quests still had an edge over the Opera's with ideal speakers, give them more conventional speakers and the magic that made them so special disappeared and the Opera PP's took the prize by quite a noticeable margin. As always the SE just needs more care in partnering.

PP or SE?

Now to one of the main purposes of the review. Open that other, identical pair of boxes, slot in all the valves and see what the full-fat 300b PSE's could do.

First - "chapeau!" (hats off!) to Opera, two amps using different circuits but with exactly the same overall tonal balance - I like that - it shows the designer is aiming for a particular sound and is able to get it.

Having lived with, and really liked the PP's for a month it was the moment of truth. And in truth there was little in it, but that little was very important. First let's dismiss the myths - the bass control and treble extension were equally good. If anything the SE's sounded that little bit faster and more tuneful in the bottom octave, with the PP's a little drier and more controlled, I could say "sat on" but that's a negative and it's not what I want to put across - it was good bass but perhaps a little more (too?) damped, the SE seeming to breath a little better. With heavy thuddy bass workouts like Madanna's "Ray of Light" that extra damping was useful, but for more realistic and especially acoustic instruments the SE felt more realistic.

This ease in the bass also seemed to give the music a little more space in which to generate an acoustic. To give an example, often quoted, of the back-wall effect on "Take 5" - it was as if there was clearer air between the drumkit and that wall, the PP making it just that bit "foggy". I want to emphasis that we're talking about small differences here, differences unlikely to make you buy one over the other.

But the well-known strength of SE amps is supposed to be the midband, and it has to be said that if there was anywhere where the SE really pulled out a lead it was here. It's not the sort of difference that smacks you in the face, just a gentle awareness of nuance and shape of instruments, the clarity of vocal in particular, especially when in harmony. On really well recorded stuff this effect just sucked you into the music, it's not that the PP was bad, I'd say it was better than some SE 300B's I've had here, it's just that it sounded ever so slightly cold and clinical in comparison, "clipped" perhaps. The PP sounded as if you were listening to a master-tape one-generation back from the original. Even with less well-recorded music the PSE gave an ease and flow that actually made the music easier to listen too especially at high volume - heavy metal with an SE amp? Well why on earth not? I'm not a great fan of pulling out phrases like "you could tell why this was the composers favourite piece on X" but in the absence a better vocabulary I'd say the midband was like a clear mountain stream compared to clear, good tap water (sheesh! maybe I'll lose that...). Perhaps it would be better to compare it with the way vinyl sometimes just thrashes the CD but damned if you can put your finger on why - when you listen for details and soundstaging they are indistinguishable, but you just want to leave the former on longer - so it is with the SE amp.

It's hard to separate the top end from this because the midband just rises up seamlessly to meet it. Those of you who own speakers with crossovers and fundamentally different drivers for each frequency range may not understand this, lumpy response they may have, but the full-range driver is seamless in a way that only another crossoverless design can be, whether panel, dynamic or other technology.

Both PP and PSE pull this trick, but the PSE is so sweet, cymbals having a real shimmer and size when struck. I know critics are going to point to shed loads of even-order harmonics sugaring the sound, and if that's the case fair enough - I have a sweet tooth.

Trying both with the IPL's made the case for the PP amp, here the PSE did begin to sound a bit flabby in the bass, especially if the wick were turned up and the less revealing delivery gave no chance for the superior midband to shine through - in this case I'd rate the PP as a comfortable winner.

So, given my speakers and system the PSE wins the day, but now was the time to stack it up against my own, similarly priced Audionotes. Well hand-on-heart I'd be pushed to tell them apart. The Audionotes are a little warmer but given a day to acclimatise I'd not be able to identify one from the other blind, not even over a long period of listening. Sometimes the differences between amps become apparent when I find myself putting on an extra record - it sounds so good. In this comparison I was equally happy to listen to either, and that's high praise, Audionote have been a flagwaver for SE amps for a long time and many companies use their transformers, for Opera to operate in the same rarefied atmosphere is cause for congratulations.

Of course in this comparison the Opera's have the advantage of giving nearly twice the power and driving ability. That means that where I'd say the AN's need speakers of 95 dbl efficiency, the Opera would give the same level and headroom with a speaker of 92 dbl. That brings a hell of a lot more speakers within its range. Yes the equivalent Audionote PSE, the Conquest would match it and for less money, but it wouldn't have the 'Silver' spec of my Quests and so should trade off with a lower quality of sound overall. The Conquest Silver, which I suspect would be a match, costs more - you get the idea.

Conclusion

So to the first question SE or PP? The test was, for me at least, conclusive. The SE version of the amp was superior - period. The PP was excellent and in other company, even against other SE amps would have shone, but here it was always shaded by the PSE. It's often said that some people will prefer the PP sound to the SE, I have to disagree. Given two near identical amps it is the SE that consistently produces the best music, the difference wasn't a gulf but I think you'd need a perverse set of priorities to put the PP first.

The PSE amp wins but only in ideal conditions. And that caveat "ideal conditions" is important. I love my high efficiency horn speakers, but there are inevitable compromises involved - limited bass extension (I use a big REL sub to help here), uneven frequency response, huge size. If your set of priorities call for different speakers then the PP amp may well win the day.

The second question applies of course to how good either amp is in a more general marketplace. At the high price there's a lot of competition, both valve and transistor, but even then the PSE's make a lot of sense. They look fabulous and they give all you'd expect from a quality valve amp along with reasonable driving ability.

The PSE version of the amp matched my Audionotes in sound quality, build quality and I have to say looked more striking, but retained the power advantage. As the Audionotes had been my choice over similar amps from the likes of Loth-x, Audion and Wavac that puts the Opera PSE right in the front rank.

*Opera deliver review items when they say they will - other manufacturers please note, it's why I like reviewing their gear...