Manufacturer: Shuguang
Product: Shuguang Treasures CV181Z valve (similar, but not identical to 6SN7 valve)
Supplier: Grant Fidelity Price: $320 tested matched pair, YMMV due to currency volatility
Support/Feet: Octal vibration isolators not yet available
Reviewer: Mark Wheeler - TNT UK
Reviewed: May-June 2009

Whenever I have mentioned the 6SN7GT valve (tube) in my reference modified Assemblage SET300B Signature Edition, I have hinted at a preferred NOS (New Old Stock) example that I dare not name as it has never become fashionable among tube-rollers and I did not wish to scare the price high. Tube-rollers are those audiophools, like me, who find optimising valve manufacturer and vintage has more sonic effect than cables, capacitors or mains connectors. My secrecy enabled me to continue buying from my regular supplier Jim Fish, proprietor of Wilson valves. I feared that if TNT-audio.com, the most respected source of audio information on the world-wide-web (as reported by independent auditors) named a particular old stock valve as the best, the price of remaining stocks would rapidly and dramatically increase. I detest b*llsh*t prices for audio equipment because the mood enhancing delights of great audio should be available to all. I feared that if one brand/vintage was named as the best example of one of the most popular octal base double-triodes of all time, the value would increase beyond my own pocket. I have only seen this particular brand/type/vintage named on one website and no internet forum, nor in any magazine, so it must not be widely known. When I had ordered samples of all the different types of 6SN7GT that Jim had in stock to try, it was because the 6SN7, despite being one of the most linear double triodes ever made, can be one of the most difficult to apply musically in high-end audio applications.

"The old fool's off again," Argue plebs, stage left, "He's contradicting himself and rambling into his beer. First he says the 6SN7GT is really popular 'cos it's so linear, then he says it's difficult to use; which one is true?"

Well, dear readers, both are true and neither are mutually exclusive. As the graphs show, the curves for this valve lack curves, if you get my drift (not bias drift, hope the plebs). The GT suffix on 6SN7GT denotes only that it is enclosed in a Glass Tube (rather than metal clad types), not that it is in any way a Grand Tourer. The suffix "G" alone usually describes the larger shouldered baloon style of Glass envelope and the final "T" suffix denotes the 'Bantam' tubular type of glass envelope. Those of us residing on the Eastern Shores of the Atlantic Ocean who continue to resist the term tube to refer to any other type of valve are correct if pedantic; the only valve that is truly a 'tube' is one with a GT suffix. An electrical valve controls the flow of electrons like a tap controls the flow of water molecules so, not having the GT suffix, the CV181Z is a valve. In case you wondered, the WGT suffix on valves denotes an extra strengthener for ruggedised military types whose extra ruggedness might just as easily detract as enhance sound quality (hence the legendary red base 6SN7 RCA5692 is reputed to sound similar to the Sylvania JAN CHS 6SN7WGT) and GTA denotes higher power dissipation and GTB denotes controlled warm up. From now on for simplicity we shall simply use the nomenclature 6SN7 becuse I type alphanumerics too hard and destroy keyboards and bruise fingertips.

The CV181Z and the 6SN7GT

The 6SN7 valve is duly popular because it can work in circuits with a wide range of B+ volatges on the anode: it can be juiced up at low B+ voltage while running high currents; it can swing on a wide range of input signal, so it is a justifyably popular with designers. However, like many enthusiasts' choices (e.g. so called driver's cars or riders' bikes) the 6SN7 needs skillful handling to get the best from it or it might bite back. Some samples have an empty lacklustre quality to the sound, while others suffer washed out but edgy and grating sonics. None of us would wish these from a valve in either high-end audio or instrument (e.g. guitar) amplification.

Of the dozen or so samples sent to me by Jim Fish, way back when, the worst was branded 'Pinnacle' (not to be confused with the music distribution company famous for dance vinyl) rendering all that passed through it to sound like a badly set up Linn, sited on a slope, with Asak at 1.5g. No cable could ever upset a system so much. 'Pinnacle' was a valve import distribution brand over 30 years ago, who rebranded diverse far eastern & Soviet valves, and I'm told were also unrelated to the blister-pack audio accessories brand of the same name. Of those I have tried the best 6SN7 is a Brimar. Not the more prized Tungsol, Sylvania, Mullard, not even the Sylvania JAN (Joint Army Navy) milspec USA manufactured samples (which were very VERY good and only a couple of points behind on my rating scale. Even the Brimar clear glass (GT) samples, which one might expect to be very similarly designed and build, didn't compare with the carbon coated Brimars. The black bottle Brimar has always been a head and a neck in front of all competitors.

I have discussed this on a few occasions when encountering other tube-rollers and everyone who has tried this example of the 6SN7 genre agrees. I was so impressed by the pair of black Brimar 6SN7GTs from Wilson Valves that I bought all they had left (in the picture above). This Huddersfield stalwart of the amateur radio community, who almost singlehandedly kept so many transmitters aglow for so many years, searched through his stock and sent me four more examples. There are none left to my knowledge, so I have kept very quiet about this gorgeous sounding valve.

Until now, that is.

Two things have happenned to prompt this information being made public; Jim retired a while ago and his successor was unable to keep going so sold the remaining Wilson Thermionic hoard. Secondly, I was offered the chance to review some examples of various valves distributed by Grant Fidelity, including the Shuagang Treasure series. Grant Fidelity are the Calgary based North American distributors for Shuguang; they have an unusual affordable high end philosophy and marketing strategy, the latter including a presence at the Calgary Beer and Lilac Festival. Any audio busines that makes friends at a beer festival knows how to feed more than one sense. I have seen no bad-service blog-posts about their company and they affirm a strong customer service policy in their literature and claim to test all tubes before dispatch. Indeed, my only gripe with anything on their site is their suggestion to try 8ohm taps with 4 or 8 ohm speakers to find which sounds better; in my experience the 4ohm taps invariably sound better even with 8ohm speakers (which often drop as low as 3ohms just above bass resonance or in the notch between the twin peaks of wierder reflex alignments) despite typical losses of 3dB headroom on the 4ohm tap.

Now Shuguang is a long established Chinese maker of valves. The company was born in 1958 a year before me, making them established for more than half a century. Starting later than Mullard, Western Electric, British-American (Brimar), Tesla, they have been in the valve making business since the last gasp of the valve golden age. They claim to be the largest valve manufacturer in China and have been selling in the West throughout their history, their designers and engineers having been schooled in the USA in the thirties and forties; even world wars cannot keep international audiophiles away from their benches. Shuguang claim their own brands and OEM supplies accounted for 40% of the worlds new valve production in 2008. They make most popular audio valves, in several ranges, from classic clear glass GT envelopes to the Treasure Series, which have carbon compound coated inside the glass envelopes resembling black bottle types of yore. The Treasure Series was released in December 2008 to celebrate Shuguang's 50th anniversary just in time for mine in 2009. Grant Fidelity noted that their Shuguang Treasure series CV181Z is a 6SN7 substitute, although NOT a direct equivalent or replacement.

The Shuguang Treasure series CV181Z review samples

There were long delays at customs at Heathrow airport, a London airport I detest almost as much as London Gatwick, as the customs decided import duty and VAT were due. This is a perenial problem for UK customers ordering small value items from outside the EU; I have faced customs officials who insist on guessing a UK retail price when I am showing them a receipt for the price paid elsewhere, even when I'd paid local taxes and used the item for a month in the country of purchase. Be warned when ordering items into Britain. Eventually the argument was accepted that the items would be destroyed by testing and therefore of no further value.

Unpacking the Shuguang CV181Z from the box revealed a much larger glass envelope than the usual octal small signal double triode. Octal based double triodes generally suffer less from crosstalk between the two triode sections than their b9a miniature electrical equivalents. In the case of the 6SN7, the b9a cousin is the ECC82, which seldom catches the imagination of audiophile designers and has a reptutaion among audiophool tubeheads for blandness. Conversely, the octal equivalents of the more popular b9a types suffer a reputation for microphony, so you just can't win; there are no generalisations between octal and b9a equivalents. But then you get presented with the CV181Z. Now a 181 is a famous Vincent Bach cornet of exceptional quality, a brass instrument capable of rising above an ensemble with clarity and returning to the brass band ranks to blend seemlessly; I recall a freind in my school brass band lusting after a 181. Will the Shuagang CV181 demonstrate that same versatility and lustfulness?

The glass bottle is a bulging waisted shape like the bastard offspring of a big recifier (like the 5AR4) and a power tube of the Kinkless Tetrode family (KT66 or KT88), but harking back to a 6SN7 predecessor, the 6N7, a 4W dissipation triode for class B circuits that's great sounding and cheap as chips. Standing taller than many common octal double triodes. The Shuagang CV181Z is black, just like my Brimars, and they claim the "High Polymer Compound Carbon (HPCC) coating is for the first time used on vacuum tube production. Compared to traditional carbon spray technology, HPCC coating will allow greater concentration of electrons, enhance and maintain stability of the electron flow, maximize "secondary electron emission surpression" from the glass enclosure, in turn reduce effects of electron swirl which generates internal noises from the vacuum tube itself." So there you are, who am I to argue? So I immediately replace the Black Brimar 6SN7 input valve of my modified Assemblage SET300B Signature Edition and switch it on. I lay my hand on the mains transformer after about 5 minutes to check the temperature.

This is why the CV181Z might not be a drop in replacement for the 6SN7. Typical 6SN7 heaters draw about 0.6A and the CV181z draws about 0.9A (900mA), the old Mullard CV181 drew 950mA. This 50% increase in heater current amounts to 0.6A for a pair of CV181 in a circuit, which as you can see represents the load of a whole extra double triode.

"So what?!?" demand plebs, stage left, "600mA is hardly big, what effect would that have?"

Mains transformers are really big things. They cost a lot and the highest current draw is for the heaters, not the B+ or bias secondary windings, so the heater secondary is a major proportion of the overall cost. When total current demand is closer to its maximum capacity, a transformer's regulation deteriorates. Just as the single winding of a choke (inductor) might be 20H at 30mA but sags to 10H at 60mA (real figures from a really expensive top quality power supply choke on my bench), so the four or five windings of a big transformer all mutually interact (like a big mutual inductor, obviously). Hence, regulation might become poorer in response to the fluctuating demands (as the power supply capacitors cycle) caused by any massive increase in heater demand. Taking heater current demand up to the transformer design limit might result in poor regulation of the critical B+ and bias widings if the transformer has not been specified with plenty of extra headroom. This is why cost-no-object designers always derate 30%+ (hence all those 630v capacitors in your 270v valve circuits). Poor transformer regulation leads to sagging plate voltage, which leads to sagging music and fluctuating bias leads to horrible unharmonic distortion. At worst, if the extra 0.6A (600mA) demand is above the heater winding's rated maximum, it could overheat and eventually cause insulation failure. The mains transformer (mtx) of the Assemblage SET300B already runs warmer than my other SET mtx so I keep my hand on for long enough for the temperature to stabilise. The laminated core is no warmer than usual so I do not dig out my thermocouple for a more accurate check. I put a cd on repeat and leave the room, returning every few minutes to check the mtx core temperature, but trying to avoid listening to the sound of a brand new valve that has not yet settled. Using a separate transformer for the heater supply would be a foolproof upgrade as would a separate mtx for each channel, but that's another story of obsession.

The CV181Z is electrically similar to the Mullard ECC32, with a maximum plate V of 300. Run at 250V (fairly typical) the 6SN7 could be biased around -8v (although typical applications are often less than this) while the CV181 probably benefits at half this for best linearity. So If your 6SN7 run on higher anode V or that high -8V grid bias, installing the CV181 will require some adjustment. Audio applications generally benefit from lower plate voltages and higher currents so the wilder applications of which the 6SN7 is capable (like plates at 450V and grid bias at -20V) are very unlikely. At 200V either the 6SN7 or the CV181 will happily do the business with -2V bias. The CV181Z mu is 32 rather than 20 of the 6SN7, but the level difference is not as great as that looks and unlikely to overdrive the next stage.

Sound Quality

After dinner I return to the system and put on an LP. I am in the middle of conducting various plinth experiments with the Garrard 401 so that's where the lp gets played. The 401 with Series I SME3012 is an obvious partner for a zero feedback all-triode (apart from the rectifier diodes of course, although some folk use TV deflector triodes as rectifiers) amplification chain and Hammer Dynamics based speakers. The LP is Little Feat's Last Record Album; it is out because it is a killer for sorting out front-end weakness. The bass guitar easily overwhelms the vocals on many turntable/arm/cartridge combinations (it's the perfect test to check whether a pre-Cirkus Linn is out of tune) and I have been using it, among others, to help fine-tune an experimental arrangement on the 401. Many LPs have a particular track, like the crux move on a single pitch rock climb that earns the climb its grade, and the crux track on the Last Record Album is Long Distance Love.

Through the CV181Z the vocal "Hello, give me missing persons..." rises above Kenny Gradney's uber-bass with more clarity and precision than the Black Brimar last night. I suspect my aural memory or a variation in mains quality. I play through to the end of the LP (it sounds that good), switch the amplifier off, change the CV181Z for the Brimar 6SN7 and wait 10 minutes after repowering up then play the LP again. I love this album, so repeated exposure has never been a hardship. Sure enough, the black Brimar 6SN7, my previous secret reference, is less well articulated. The Shuagang CV181Z has better defined vocals and the bass is also better defined on the CV181Z. Macro-dynamics are also bolder, but is the top end quite as sweet? Do I detect a hint of hardness from the brand new barely warm CV181Z.

I try Flora Purim's Welcome Back '95 and conduct the same A-B experiments. This is a practice I normally abhor as it exaggerates difference rather than foregrounding quality and creates a delusion that component changes are more significant than long term musical qualities. The CV181Z top end, cymbals and snare brush work are crisper, as is soprano voice. I run the same tests through the Sylvania JAN CHS 6SN7WGT which runs a close second to the black Brimar (having neither its midrange warmth nor fluidity) but the Sylvania is superior to the Brimar right at the top of human hearing. The Sylvania loses to the Brimar on that luscious midband but is more akin to the CV181Z for detail and rhythm, although the CV181Z still edges ahead, and stays in that gold-plated socket.

I play LP after LP until way past bedtime. I play Dave Clark's Archive One and discover the pace and rhythm to be the best I've heard from this amplifier with the TJ meshplate 300B installed. This prompts me to replace the TJ meshplate 300B output valves with my Western Electric pair. The WE300B is superior to the TJ meshplate, especially in rhythm and bass. The TJ also being very microphonic tends to be used as my everyday 300B, still infinitely superior to the Valve Art 300B, the latter a dreadful device IMHO undeserving of the recommendations and reviews that encouraged me to stick with it for too long.

I have never heard the Assemblage SET300B Signature Edition sounding so good. The soundstage is wider and deeper, suggesting both lower crosstalk between the two triode sections and better matching especialy at the highest frequencies. A quick blast of the LEDR test confirms this, despite it being designed to test for early reflections, it can be useful as a quick indicator of why the soundstage is presented in a particular way. While the cd player (Avondale Audio AAA5) is hooked up, I run some cds through too. The results confirm the vinyl experience. This is rearranging audiophool folklore; a new valve being superior to NOS classics!

There may be other reasons this valve is superior as a 6SN7 sub. It's not just heater current that's different. The CV181Z is designed to run in a narrower envelope of plate (anode) voltages (B+) than the 6SN7 (some of the latter can theoretically survive 450v), but typically runs higher currents. Effectively, if driopped into a circuit designed for a 6SN7 the CV181 is operating at a point some might consider juiced up. It would probably make an ideal driver for output valves. The different electrical parameters mean that you must check circuit suitability for the Shuagang CV181Z before inserting it into your amplifier. It would not suit my high plate voltage stretched to within an inch of its life 6SN7 experimental pnono stage location (anodes at 300V+, low Z bypassed cathode) but it is the new reference in my power amp input stage. I also tried it in a breadboard line stage where again the CV181 beat all comers including the Sylvania JAN CHB 6SN7WGT (that I've seen advertised on the web at over $100) and the black Brimar.

"So how does the old fool justify all this enthusiasm?" demand plebs, stage left, "Has he been got at? Slipped a few freebies? Or just finally lost the cynicism we know and loathe?"
My enthusiasm for this valve pales into insignificance compared to that of its parents at Shuagang. They state the CV181Z features "Super Alloy technology" used for the first time used in valve construction. "After 3 years of testing under extreme conditions, Shuguang has successfully adopted such technology in vacuum tube construction to achieve directional solidification and single crystallization, greatly improving the ability of electron emission and electron current stability."

The original 6SN7GT does have some disadvantages, like may octal small signal double triodes it is very microphonic (much more so than its ECC82/12AU7 cousins) and some designs have excessive interaction or crosstalk between the two triode sections (although not as bad as their miniature b9a cousins). The black glass of the Brimar is a graphite coating like that of the grey interior coating of the legendary Osram KT66 (as loved in Quad II), designed to mop up excess accumulations of electrons in the envelope and I am pleased to see the Shuguang Treasure series CV181 shares this feature, which they describe in the excitable terms above. The Shuguang also features the internal top dome silvering similar to the Sylvania JAN 6SN7GT and I can see twin getters at the top of the glass envelope through the gap between carbon and silver (the Pinnacle features a single bottom getter, and I am sure with time I can think of a tabloid pun for that one), but design features alone do not a great valve make; the feeble Pinnacle shares the box anode design of the great Sylvania, yet their audio performance could not be more different. The Shuguang Treasure series CV181 almost achieves the best of all possible worlds.

After extended listening, repeatedly having to remind myself that I am listening to my SET300B rather than my slammin powerhouse solid state behemoth, I begin to take the extra bass definition and finely etched treble for granted; I stop noticing the quicker pace of this valve; the superior portrayal of rhythm becomes the norm and the wider deeper higher soundstage becomes part of the furniture. I begin to wonder, am I missing something? Time to try the old black glass Brimar again and this time use my judging system. In case it is the magic midrange fluidity I first noticed on the Brimar when group testing, I play Muddy Waters Folk Singer and Lightening Hopkins Going Away, two of the greatest male vocal slabs of vinyl ever cut. The excelent Shuguang CV181 highlights a problem with the Brimar sample I've been using recently as the CV181 is better matched between sections than the Brimar, which I notice while gain matching to compensate for the CV181 higher mu. The net difference betwen the valves using my judging system is +15 to the Shuguang, but this is purely on this simple Muddy Waters vocal and guitar lp. The Brimar gains a point on vocal instrument/pitch/tunefulness and vocal inteligibility/clarity but the CV181 gains 2 points over the Brimar for soundstage realism/stabily; for bass speed; for bass clarity; and above all for macro dynamics (getting close to making +3, the maximum difference describable). The CV181 also gains a point over the Brimar in each of 4 dimensions of PRaT; and for soundstage depth; and for soundstage height; and for bass depth/extension; and for bass pitch/tunefullness; and for micro-dynamics. Most telling was the clear impression of Muddy's mouth proximity to the microphone was clearer on the CV181 than the Brimar, despite this being one of the Brimar's noted strengths compared to other brands.

Playing the Lightenin' Hopkins all analogue all valve, from microphones, through mixer, studio tape recorder, cutting head, HQ-180 pressed recording through valve playback pre-amp and power amplifier enables this this spectaular performance to demonstrate the subtleties of NOS and brand new valve choices. The Sylvania JAN CHB 6SN7WGT made no gains in any performance area over the CV181. The difference overall was 18 points to the Shuguang CV181Z, again in the areas of macro dynamics and micro dynamics, but also a giant leap forward over the Sylvania for what I call virtuality (realism, not pinpoint imagery). The Shuguang CV181 improves on the Sylvania for all midrange and vocal qualities, improves in the 4 dimensions of PRaT, and gains extra 'tingle-factor'.

Conclusion

The new valve (tube for those who skipped to the conclusion) on the block is the Shuguang Treasure series CV181.

When I am listening to sand (solid-state amplification) I often yearn for the virtues of valves, the midrange lucidity, the unforced realistic instrumental timbre, the natural soundstage etc. When I am listening to glass I often yearn for those solid-state qualities of solid bass, pace, rhythm, finely etched treble, grandiose macro dynamics etc. Plugging the Shuguang Treasure series CV181Z into 6SN7GT locations takes the traditional valve qualities and adds some of the silicon tautness.

The glass is only there to drive the iron
So the glass driving that glass is one step removed...
and the glass input feeding that glass is two steps away from bass, pace and dynamics...(?)
So the input valve's contribution to the traditional flat earth qualities of PRaT are stymied by context but the Shuguang Treasure series CV181Z excells in this location in these dimensions.

"The old fool's a reggae and rock steady fan," Argue plebs, stage left, "So any name that resembles Treasure Isle gets an unfair advantage...."

This Shuguang Treasure Series CV181Z is a new valve brings solid-state qualities of PRaT to the ubiquitous 6SN7GT family. It succeeds in beating the best NOS samples at their own games with the added bonus of new stock reliability and dependability. Grant Fidelity's added value of testing, selecting and matching at no extra cost takes this even further from the realm of the tube-rollers' box of assorted NOS, 3 days work and an ageing Avo CT-160 valve tester.

This is an expensive valve. A matched pair will cost you $320 which is $100 more than a pair of desirable NOS (although I have seen NOS 6SN7GT at $150 each), but you might have to buy more than one pair of NOS to get a pair matched as close as the sample pair of Shuguang Treasure series CV181Z; the differences were smaller than my test gear could measure reliably. I have spent many long hours burning-in and auditioning NOS 6SN7 (there's so many to choose from) and the Shuguang CV181Z beats them all with maybe 50 hours on them and they claim to need 300 hours. Anyone wanna buy a load of black Brimars?