Columns Retired Columns & Blogs |
I totally get why Stereophile might not want to go there , but I sure would like a review of select tube amps from China currently (*) selling for a song (*) on Amazon.
*Puns unintended , of course .
He reminds me of my least favorite teacher, Professor Grausamkeit, who was just like that and said similar things. Every time I smarted back, "Accurate to what?" he'd whack me with a wooden yardstick.
Grausamkeit would twist my ear painfully while scolding me but would never explain what "accurate" sounded like or why it didn't sound pleasant. Best I could figure, he meant gray and stern, like him, or bland and flavorless like our school lunches. Professor G was constantly reminding us, "Pleasure-seeking is a congenital defect that defines the ignorant classes."
Because of G,I grew up to be an insubordinate audio flâneur, a world-traveling, drug-imbibing Anthony Bourdain type who explores parts unknown seeking full-flavored gourmet audio and musical experiences (footnote 1). I've devoted my life to mocking sanctimony while exploring the back roads of audio design, befriending engineers with open minds who question all forms of received orthodoxy, searching constantly for the wisdom of wizards and the secret keys that open hidden doors to the purest forms of audiophile pleasure.
I'm old now and happy to report: I've been a lucky traveler. I've befriended more than a few audio wizards. In fact, I just "discovered" another one. His name is Steve Deckert. He's been building and selling tube amplifiers since 1993. I've been aware of his company (Decware High Fidelity Engineering, footnote 2), and I knew about his beautiful "Zen Triode Amplifier" because the birds in the tube forest told me. Embarrassingly, I've only now got 'round to knocking on his door, auditioning some of his creations, and grasping his accomplishments. I'm sorry I waited, because this wizard makes ingeniously simple amplifiers that make recorded music sound pure and beautiful.
The Zen Triode Amplifier
Decware founder-in-chief Steve Deckert named his first product The Zen Triode Amplifier because it was incredibly simple and delightfully unpretentious. Against all odds, Deckert's 2.3Wpc integrated amplifier has been wildly successful, selling more than 6000 units and evolving through seven numbered revisions. According to the Decware website, the latest revision, model SE84UFO, is a class-A, zero-loop-feedback, single-ended, stereo tube amplifier using just two resistors and one Jupiter Beeswax film capacitor in its signal path. It costs $995.
The Zen Triode Amplifier is hand-assembled and point-to-point wired in East Peoria, Illinois. It comes in an impressively thick, powder-coated steel chassis that measures 8" H (with tubes) × 6.125" W × 12" D and weighs 17lb.
Each channel employs a single triode-wired, self-biasing, self-balancing 6N15N output tube, which is the electrical equivalent of the American 6BQ5/EL84. Each Zen power tube is driven by one section of a high-transconductance, medium µ, 6H1N/6N1P dual-triode voltage amplifier tube, which is the equivalent of the popular 6922/6DJ8. This is an interesting design decision because both of these tubes are widely esteemed for their ability to retrieve greater-than-average quantities of low-level, small-signal information. In my experience, the 6922/6DJ8 is a high-revving 9-pin tube that recovers so much microdetail, bas-relief texture, and subtle tonal nuance that it makes 12AX7 or 12AU7 tubes seem blunt and generalized. If you look at the photo of the Zen Triode Amplifier, the biggest tube you'll see is the ST-shaped (Russian) 5U4C full-wave rectifier, which may be tube-rolled to great effect or substituted with 5AR4 or 5Y3GT tubes.
On the back of the amplifier, in front of the right- and left-channel speaker-wire binding posts, are little switches. They select between output transformer taps for either High (616 ohms) or Low (26 ohms) impedance loudspeakers. Decware's website claims the Zen can handle impedance dips down to 1 ohm.
In the front are two more little switches. The one on the left chooses between two line-level (RCA) inputs. The one on the right switches between two cathode-bias resistors, raising or lowering the standing current on the input tube. According to the owner's manual, higher current (harder class-A) delivers a punchier sound and is louder. According to Deckert in an email, "The other is more laidback and dimensional, better suited for using preamps."
On the chassis front is a single chickenhead knob that controls stereo volume.
When I asked Steve who designed and built Decware's "highly interleaved" "Ultra Fidelity" output transformers, he replied, "I did. We have always made our own output transformers. They are sneaky good, which is a large part of the reason this amplifier design has been so successful for 25 years."
The Zen Triode Amplifier is sold direct with a 30-day money-back guarantee and is warrantied for life to the original owner. For a fee, second and later owners may get their Zen amplifiers "recertified" by Decware, extending the warranty for their own lifetimes.
The sound of one tube
The minute I heard the Decware Zen playing Benoît Menut: Les Îles (24/96 FLAC Harmonia Mundi/Qobuz), I knew this would be a fun article to write. The Decware Zen's conspicuous purity of sound turned French cellist Emmanuelle Bertrand, soprano Maya Villanueva, and members of the Syntonia Piano Quintetplaying compositions by Benoît Menutinto one of my most thought- and pleasure-filled musical moments of 2020.
The system was simple and low cost: the $899 Denafrips Ares II DAC sourcing the Decware Zen Triode amplifier powering my beloved $549/pair Klipsch RP-600M loudspeakers, which claim a sensitivity of 96dB/2.83V/m and a nominal impedance of approximately 6 ohms. I used Sound Anchor Reference loudspeaker stands with interconnect and speaker cables by Triode Wire Labs. Total system cost: less than $4k.
My Zen-Klipsch listening sessions were dominated by a general sense of radiant, saturated tone plus a feeling of denser-than-usual physical presence. With the Zen's bias switch on High, music from the RP-600Ms was reproduced with a level of overt vividosity that I had not experienced previously. With the bias switch on Low, the music got quieter, more subdued and plain-spoken.
I have always said that the system that reproduces sopranos and pianos with the greatest verity is the most accurate. If that is true, then this little system is one of the more accurate I've encountered. These humble components put Maya Villanueva's seraphic voice in my room and in my heart. (I have a history of swooning for sopranos.) The middle registers of Romain David's piano were conveyed in rich, complex tones. The piano's low registers were tautly drawn. Surprisingly, the Zen Triode yanked more tight bass from the bass-light Klipsch than any other amp I've tried. All that separated this system from bigger-money systems was the Klipsch RP-600M's inescapable lightness below 100Hz and its lack of refinement above 1kHz.
The above system played voice, solo instruments, and small orchestras with incredible grace and substantial verity, but only at modest volumes in my smallish room. Time for bigger, more sensitive speakers.
The Zu-Zen experience
Like Steve Deckert, Zu Audio's Sean Casey is one of those wise wizards who snicker at orthodoxy. I reviewed his $4499/pair, 16 ohm, 97dB/W/m Zu Audio Soul Supreme loudspeakers in June 2016 and have been using them with low-powered amps ever since.
As you may have noticed, the Soul Supremes' sensitivity is rated in watts, not at 2.83 volts as is the fashion these days. The two units are equivalent only for 8 ohm speakers; at the Soul Supreme's 16 ohms, you get more volts per watt. So, the Zen's 2.3W should play comfortably and loudly with the Zu speakers.
Footnote 2: Decware/High Fidelity Engineering Co. 75 S. Riverview Drive, East Peoria, IL 61611. Tel: (309) 822-5255. Web: decware.com.
I totally get why Stereophile might not want to go there , but I sure would like a review of select tube amps from China currently (*) selling for a song (*) on Amazon.
*Puns unintended , of course .
The audio-pond is swimming in these babies, I wish we had an easy way to compare.
If Stereophile made a Go Fund Me page to buy several of these cheap amps in order to review or measure, I'd chip in!
There are 2 justifications for the 2.83V figure. First, the acoustic equivalent of voltage is sound pressure. If you specify sensitivity as spl per watt, you are both confusing sensitivity with efficiency and committing what my physics professor used to call a "dimensional mismatch," ie, the terms on both sides of the statement are not equivalent.
Second, the advantage of specifying sensitivity rather than efficiency is that it remains unchanged no matter what the impedance of the loudspeaker, as it is assumed that the amplifier will always be able to provide the necessary current to maintain the 2.83V. (See www.stereophile.com/content/measuring-loudspeakers-part-one-page-3.)
So why not specify efficiency: the radiated sound power per electrical watt? While it is trivially easy to measure spl, it is extremely difficult to measure sound power. A loudspeaker outputs sound over the entire solid angle into which it radiates. You therefore need to integrate the spls at hundreds or even thousands of points in a sphere around the speaker, or, place the speaker into what is called a maximally reverberant room. Both are impractical.
So the technically appropriate specification must be sensitivity, the spl at a given distance (1m) for 2.83V input. When a speaker has a impedance that is higher or lower than 8 ohms, Stereophile's reviews include that information, thus addressing John DeVore's point. In addition, we now publish the EPDR, the Equivalent Peak Dissipation Resistance, which reveals how demanding a loudspeaker is for current.
John Atkinson
Technical Editor Stereophile
After lurking and reading for a while now I might actually be able to understand some of the amplifier measurements. I know you don’t normally measure equipment featured in these columns, but I was looking forward to seeing measurements of the Zen amp so I could correlate what I hear with what the measurements show.
Hi John.
I appreciate the desire here for perfect technical specification for a given product, but how does this help consumers make decisions, especially with a standard like 2.83V per watt that assumes 8ohm impedance from the speaker? I'll admit that I'm not an engineer, and you're going so deep into jargon that I have no idea what you're talking about. Call me crazy, but I'll bet most folks find this kind of learning curve, just to pick out some sweet new hi-fi, incredibly intimidating. Why can't I, the customer, just be told something like- "Hi! This speaker should be connected to the Xohm plugs on your amp, and that amp should have at least X watts to sound great. Sincerely, Speaker Manufacturer."
I appreciate the desire here for perfect technical specification for a given product, but how does this help consumers make decisions, especially with a standard like 2.83V per watt that assumes 8ohm impedance from the speaker?
Unfortunately, there isn't one specification that allows 2 speakers' suitability for a system to be compared. However, let's say that you are looking at 2 speakers that both have a higher-than-average sensitivity of say, 90dB/2.83V/m. You then look at the impedance specifications for both. Speaker A has an average impedance of 10 ohms, Speaker B an average of 2 ohms. Speaker A will therefore play to loud levels, even with relatively low-powered tube amplifiers. Speaker B will work okay with modern solid-state amplifiers but will suck the life out of many tube amplifiers, even when they are used from their 4 ohm output transformer taps.
John Atkinson
Technical Editor Stereophile
Thanks for the reply, but I'm afraid your answer only results in more obfuscation for me. If the point of a spec based on 2.83V is for evaluation, shouldn't it strive to create a clear, universal standard? Put in another way, if the 2.83V spec still results in "it depends" when in relation to a given amplifier, is the spec really successful, or is it a standard for measurement only inside a vacuum?
Based on your explanation, and as a typical music-loving consumer, I think I'd be lucky to get any dealer to say more than "maybe, maybe not" when asking if a given speaker is a good fit for my amplifier. (In my case, the first generation Air Tight ATM-1.) Any more digging into the 2.83V specs on the back of the speaker for clarification could possibly drive me straight out of the shop. And if I'm a music-living hi-fi newbie, maybe even straight out of bothering with hi-fi. "Sonos, save me" right?
I'm not trying to be negative or cause fights of any kind, I promise. And I absolutely get why measurements are incredibly important for creating baselines for benchmarking and standards. I just find quite a bit of the audio industry to be so steeped in engineering nuances and jargon that a chasm of specialization has developed between them and the rest of us who are just diy nerds and customers, not audio engineers. And ultimately, perhaps to my folly, I've always believed that specs are supposed to be tools of communication not just to state a measured performance under a given circumstance, but to also to help customers make a purchase decision. Yet, how helpful are the specs when the answer they generate is "it depends"?
Thanks for the reply, but I'm afraid your answer only results in more obfuscation for me. If the point of a spec based on 2.83V is for evaluation, shouldn't it strive to create a clear, universal standard? Put in another way, if the 2.83V spec still results in "it depends" when in relation to a given amplifier, is the spec really successful, or is it a standard for measurement only inside a vacuum?
The 2.83V sensitivity specification has been the standard for many decades. But as I wrote, it is not the only answer to your question about how does it help consumers make decisions. It must be considered in conjunction with the loudspeaker’s impedance specification. Each cannot be considered in isolation from the other.
Based on your explanation, and as a typical music-loving consumer, I think I'd be lucky to get any dealer to say more than "maybe, maybe not" when asking if a given speaker is a good fit for my amplifier.
Putting to one side the dealer’s experience in what loudspeakers will work best with your tubed AirTight amplifier, consider the question about what specification best defines your needs in another heavily engineering-based field: automobiles. Is it fuel efficiency? Acceleration? Maximum speed? Road handling? Cargo capacity? Comfort? Will it fit your garage? Not one of these alone answers your question and as with loudspeaker sensitivity and impedance maximizing some of these compromises others.
Yet, how helpful are the specs when the answer they generate is "it depends"?
Taken as a whole, the specifications do answer your question, which is why this magazine's loudspeaker reviews include a full set of measurements that examine them. But if you just look at one specification, as you have been doing regarding sensitivity, the answer can only be "it depends." it can't be otherwise given how loudspeakers work.
John Atkinson
Technical Editor, Stereophile
I think the automobile/speaker analogy is perhaps reaching a bit too much. At best, speakers are analogous to tires for vehicles. And the tires can come in all kinds of tread types, widths etc., for various uses and performance traits, but they fit the wheels or they don't. And one way to know if the tires we may be considering fit both the wheel and the wheel well of a given auto/truck is the code of three big numbers printed on the sidewall.
Thanks Herb. I’m sure confirmation bias is at work, but this review lends credence to my belief that I’ve compiled the best possible system I can for the money (~$3k) and perhaps punching far, far above its weight (well, price class): Bluesound Node 2i > Decware Zen Triode > 10” full-range (crossover-less) Audio Nirvana drivers in home-built cabinets with a 12” SVS sub to fill out the bottom octave. I opted for the copper beeswax and beeswax bypass capacitor mods which are based on the work he did for the 25th Anniversary version. I’ve got some bonus money to put towards upgrades and at this point I think the biggest bang for my buck is curtains for the wall of windows.
Unless you can suggest a turntable and phono preamp combo for $1,500 or less that’s up to the task? Of course, then I have to start buying records.
For some reason just the fact that amplifiers like this exist makes me smile. Little pieces of audio artwork to remind us that listening should be about the pleasure.
Hi
Great minds think alike ??
Obviously low output-power tube amps MADE IN USA selling for below a grand is a hard-found bargain for many many HiFi beginners worldwide with limited budget. I believe tons of this amps shippped off shore in the past years.
Again it depends on the circuit design & equipment matching.
FYI, the tube power amp I design/built 4 years back, now being used as my main power amp, is incidentally pretty similar to Decware Zen:
only 2 stages= triode driver stage + triode-strapped pentode.
Very simple "musically correct" circuitry design.
While the Zen was designed in very centry-old traditional 'old school'
way, mine is sort much more 'modern' circuit topology that no commercial amps ever used: battery bias to the tubes throughout, & UNIQUE triode-strapping topology applied to the output power pentode.
It crucial to undertand traditonal old-school way of triode-wired power output pentode used by many many brandname amp makers, is very simple but drags down the original pentode performance big bigtime. This is not desirable of course. The defect of such simple traditonal basic triode-strapping of pentodes has been widely discussed in many audio electronic journals.
My home-brew "all" triode power amp only get 4.5Wrms per channel yet it sounds so LOUD. It rocks my 700sq-ft basement audio den no sweat, on playing heavy symphony works, like Tchaikovsky 1812 Overtures & church organ musics.
Mind you, I got a US brandname 35W+35W ultra-linear push-pull Class AB power amp, already fully upgraded by me, still can't touch the fast punchy transients of this 9W home-brew Little David !
So output power alone does not tell you the actual performance of any amp, my friend.
Listening is believing
Jack L
Dear Mr.thatguy,
These types of electronics have always been part of the DIY community, even from the post World War II ham radio days.
Stereo 70s and 35s in kits were and remain wonderful gear. ( even old EICO & Heathkit stuff )
These types of gear don't make attractive Covers of Glossys because beautiful $50,000 +++ Gear sells magazines and buys needed advertising.
Stereophile does cover Decware & LTA & Felikes as well as other DIY types of information.
The Audiophiliac specializes in discovering these nuggets.
The Heaphone folks cover all manner of under $5,000 Audiophile grade electronics and transducer systems.
Being an Audiophile means DIY! Checkbook Audiophiles can't get it right because it takes skill not $$$$.
Tony in Venice Florida
ps. eBay doesn't have one single Decware piece for sale at the moment, a couple of Decware pieces sold over the last 90 days. You'd think that after all these decades we'd find eBay having quite a few amps changing hands. Schiit has scads of eBay sales on the regular ( I'm hunting for an Asgard 2 in Black )
I’m listening to a se tube amp I built from a fairly popular schematic right now. I have a mono PP amp my dad built roughly 60 years ago that I listen to regularly too.
Herb,
What you describe hearing with the TU-8600R is what I've heard in other amplifying devices using solid state rectifiers when the transformer rectifier combination is not properly damped to minimize the reverse recovery effects of the diodes turning off. From the '8600R schematic I found, it appears that there is no damping applied to any of the rectifier circuits.
I know that Stereophile reviewers are not in the business of modifying products, but since you're a curious fellow, you may wish to read about the subject:
http://www.hagtech.com/pdf/snubber.pdf
I believe you are 100% correct and I remember Hagerman's "snubber" paper from when it came out.
I imagine the chattering side-effects of solid state rectifiers are the main reason solid state sounds so solid state.
herb
Hi
Ringing occurs in AC-DC power supplies when the diode turns off momentarily due to the diode's nonlinearity. The natural ringing frequency will be 560KHz+++ with high orders of harmonics up to MHz is is modulated into the AC power line frequency.
What worse is such ringing will be magnetically & capacitively coupled to the audio amp section. Noooo goood!
Ideally, the AC-DC power supple for the amp should be installed in its own separated housing away from the amp section.
That's why I installed the HV power supply using fast recovery silicon diode (for tube plates) & LV DC supply (for tube heaters) in 2 separated boxes & placed 2 feet away from my phone preamp box. Take a look at my signature logo top of my post.
Besides, snubber is a MUST to damp off the HF ringing, to be installed immediately before the rectifier diode(s).
Say for 1N4007 diode in 60Hz powerline, a simple snubber using only 3 small parts: (0.068uF+110R//0.01uF) will effectively snubbed off 560KHz natural ringing.
Jack L
Hi
Yes, snubbers is NEEDED to damp out then natural ringing caused by the non-linearity of the diodes every time when it turns off momentarily. This is crucial when the power supply shares the same housing of the amplifier due to magnetic & capacitive coupling.
Jack L
Well, it's one of the reasons.
The side-effects that are "signal errors" can be fixed.
The side-effects that are because solid state can't emulate some of the "signal errors" inherent to some tube designs are harder to fix.
The oddity is that many of the flaws in tube amplifier design happen to sound appealing. Often, very appealing. Not so much with solid-state. Sound reproduction is an illusion, after all.
Hi
Tube (=vaccum state) is not a solid state device. They are totally different constructed hardwares! Can't "emulate" with each other.
Just like apple can't "emulate" orange though both are excellnet vitamin C sources for our health.
Too bad, my friend
Jack L
Hi
Tube (=vaccum state) is not a solid state device. They are totally different constructed hardwares! Can't "emulate" with each other.
Just like apple can't "emulate" orange though both are excellnet vitamin C sources for our health.
Too bad, my friend
Jack L
Definitely biased. This review would carry real weight if this amp was given to someone who is not a tube dilettante. It is like asking what Elon Musk thinks of his cars.
I heard this reviewer can't stand anything Nelson Pass ever designed or made.
I am a tube dilettante myself, or maybe something north of that given I could build an amp given a design. Anyway, I have two rooms. One's all tube and the other is Pass and... wait for it... Wilson. Still love tubes and sensitive speakers, though.
I get what you mean. Look at how he was completely disgusted by the XA-25:
https://www.stereophile.com/content/pass-laboratories-xa25-power-amplifier
Then, the INT-25 brought his listening experience to a new low:
https://www.stereophile.com/content/pass-labs-int-25-integrated-amplifier
(Full disclosure - I have no connection with Pass anything. I only ever saw or heard a single Pass product - the Aleph 3 - back in the last century.)
Hi
Apparently, you attacked Herb R at the wrong place.
Why don't you launch yr allegation in the "solid state power amp review" forum.
Jack L
If this was directed at me, I was just responding to the comment "this reviewer can't stand anything Nelson Pass ever designed or made".
If the comment was sarcastic, I was simply adding to the sarcasm. Hardly an attack on Herb on my part.
If the comment was serious, it's factually wrong, at least based on the published evidence I cited. Hardly an attack on Herb on my part. (Perhaps the poster has inside information the rest of us don't have access to.)
I was definitely being sarcastic ;)
Thought so. As was I.
Probably a bad idea, as it turns out. It seems that home audio reproduction isn't a life or death proposition - it's far more serious than that. Fun? Ha!
Hi
So hearsay again.
You believe what you just heard without finding out why?
Like hearsay said you are silly to make such comment here. I should believe it, correct ?
Jack L
This entire Magazine is Opinion Based with the exception of Mr.JA1's measurements, isn't it?
For the most part, Mr.HR & Mr.Micallef are the Keynote writers for this Magazine, their contributions are more like adventure journalism than opinion reviewing.
Tony in Venice Florida
Hi
Elon Musk design/build electrical cars.
Herb R only reviews the Zen amp not design/built by him. This is his job here to review his subjective opinion. Take it or leave it - choice of his readers.
Likewise, I take yr allegation against Herb is "definitely biased".
Where is yr logic ?
Jack L
Elon Musk neither designs or builds his cars - his emloyees do.
Hi
He is the CEO & he MASTER-MINDED 100% it. He is God-made genius,
from a South African immigrant to the richest man on this planet !!
Jack L
searched for the meaning of " adventure journalism ". None that I found come close to what tonykaz is trying to allude to. Going somewhere new and finding something unexpected or unique is the description I found. Herb's opinion piece is definitely not in that category
Reviewer puts ears on a device. Reports result. Random comments thread guy finds this unacceptable, *and simply because it's a tube device*, alleges bias against reviewer.
Wouldn't that be intentionalism, mememe2, of the sort that presumes to climb into a guy's head and rearrange his goings-on for him?
I've said it for years: As definitions go, the dynamic involving audio objectivism and subjectivism is exactly reversed.
Thanks
Yes,
I used the Rogue RP-7 hybrid preamp
hr
Dear Herb, thanks for the reply. I ask this as a remote would be essential to me and I don't see a solution a la John Devore possible with this amp.
Regards, Lutry
A DIY remote would be possible just by swapping the knob for the volume control with one amenable to the DIY solution Devore uses. I considered it, but just use the Bluesound remote for my Node 2i and set the amp volume to the 3 o’clock position.
Hi
"3 o'clock position" volume setting is for yr preamp or power amp?
Jack L
On my power amp, the Decware Zen Triode subject of this article. Steve Deckert included a volume control so customers wouldn’t have the amps’ virtues sullied by a sub-par preamp. I have no preamp; my Node 2i is connected directly to the amp and I use it for source selection and volume control.
Hi
Bingo !
That's the best musical way to go - directly to the power amp !!!!!
Why let a "sub-par" preamp screw up the music for you ???
When J. Gordon Holt recommended straight-line BYPASS (music goes direct to the poower amp without any preamp) as the absolute reference, listen to him !!
I've been using my design/built active linestage with straight-line bypass mode ON on all the time for CD/DAC & most time for LPs for years. My precision crazed ears accept it- gorgeous airly 3-D livelike music with microdetails !!!
Don't worry about the setting of the volume control as long as you got enough loudness at when you sit. You may be surprised without any preamp electronics in between to slow down the music, the music will sound much much live & LOUD !
Technically & sonically, best sound will be NO volume control - volume set to maximum so the music can go direct into the power amp without passing through any resistance along the signal path. IMO, the best volume control still affect the music quality. My critical ears can detect it though standard bench measurement can't.
Keep it up this musically perfect way. We are on the same page, my friend
Listening is believing
Jack L
........it makes 12AX7 or 12AU7 tubes seem blunt and generalized. " quoted Herb R
Hi Herb
It all depends, IMO. Not all tubes of the same models made equal, depending almost decidedly where they are made.
The problem is: many brandname tube makers tend to use tubes of so-so brands to save COST, compromising the amps sound quality.
My ear-on experience tells me cheapie made tubes sound lousy! Money talks like any other commodities.
FYI, I chose costly 50-year-young vintage Telefunken ECC83s (=12AX7) & UK Mullard ECC82 (=12AU7) in my design/built phono-preamp. It "recovers so much microdetails & subtle tonal nuance" & 3-D soundstage presentation, which fulfills my tonal precision craze. I can't telerate anything less ! It depends on where the tube are made !!!!!! No free lunch.
Just a few months last year, my friend who is sorta a tube amps collector, dropped off at my door, a heavy-duty single-ended class A power amp of some unknown brandname, using the most-fonded 2A3 direct-heated triode output power tubes. He wanted me to upgrade the so-so factory coupling caps of the amp for him.
I replaced all the coupling caps therein with costly caps of metallized polypropylene film no sweat. The sound improved.
2A3 & 300B are the power tubes I fancy a lot sonically. I finally got the chance to audition one after persuading its owner to let me keep it for a few weeks. Why? I wanted to hear the pretty uniquely complex circuitry design pretty unlike too many conventional straight-forward 2A3 amps designs I read before.
Yes, what impressed me is the very airy well defined 3-D soundstage, famous for 2A3. But what let me down is it lacked the unique tonal signature of a direct heated triode tube besides being a bit too soft, like slowly leaking bike tires ! Ouch! The amp was made in
China, therefore using Chinese made 2A3s ! Today online sale for US$55 a Chinese made 2A3 vs $165 Westinghouse made !
Please don't get me wrong as my personal bias as I already read a lot about the quality control inconsistency of many China made tubes.
Well, like it or not, tube is a money-talk game !
Listening is believing
Jack L
Hi Herb.
I am a long time Stereophile subscriber, and when I picked up the most recent recommended components issue, I saw that the Decware Zen was in it, and that I had somehow missed your review of it last month. I’m pretty much a cover to cover Stereophile reader, so I can’t imagine how that happened. I found last month’s issue, and read your review immediately as I have long been a fan of the Decware Zen amps. Your review is very much in accordance with my experiences and feelings about the amp, and as a person whose opinions on equipment I have long respected and paid particular attention to, it was very nice to read your positive impressions.
I owned an original Zen amp (the SE84CS) years back, and have built several clones over the years, so I am very familiar with the circuit. The Zen that I have in rotation now is an SE84CS clone made with very high quality parts, and it sounds great. I use it with Klipsch La Scala and Heresy speakers, and also as a tweeter amp in a pair of horn two ways. The Decware Zen is responsible for putting me on the SET and high sensitivity speaker path. The first time I heard a Decware Zen, years ago, I was completely blown away. At its price point, it is an unbeatable bargain. Thanks for the review!
Hi.
Glad you are also a DIYer like me.
You can build a clone pretty likely sounding better than Zen by improving the old school circuitry design of Zen like what I have been doing so years back. Seriousy !
(1) spend more money on better brand-name tubes if you care. Like it or not, tube is money-talk game.
(2) do not use the extremely popular conventional old school topoloy of so called "automatic bias" for the drive tube (6922 etc) & the power tube (6BQ5)
We seasoned tube DIYers call it "Miss Piggy" = a resistor bridged btween the cathode of the tube & the signal ground, to provide the negative bias for the tube. This bias resistor is often shunted with a cap for mid/high frequency bypass.
Sonicaly I don't like Miss Piggy bias at all as it somewhat taints the music passing thru it, IMO. I beat 'her' by replacing it with battery instead. Rechargeable batteries, of course, so no need of replacing for years.
(3) The most significant sonic improvement will be replacing the very popular simple way of triode wiring the output power tube: by strapping the screen grid G2 to the plate of power tube directly or thru a low-value resistor. Nearly all brandname makes use it since day one.
Yes, simple but compromising the orginal pentode sonics big time: softer, slower & lacking punch as per many reports published.
So I beat it by replacing it with my unique plate-screen grid SPLIT POTENTIAL topology, never seen yet used in any commercial makes.
It brings back the vibrant, punching & powerful sonics of the original pentode, yet reserving the seductive sonic of a triode !
No kidding.
Listening is believing
Jack L
PS: Today is my day-off. I am now watching Hauser's seductive cello performance: "Borodin's Nocturne", streaming YouTube onto my 4K UHD TV while I am typing my post to you. Double entertainment, alright !
JackL: “So I beat it by replacing it with my unique plate-screen grid SPLIT POTENTIAL topology, never seen yet used in any commercial makes”. Very interesting. What does this look like?
In terms of improvements, the one Decware Zen that I am currently using is a DIY SE84CS clone that is basically the exact circuit as factory, with improved parts and a larger than stock output transformer. Incidentally, Decware always used Edcor output transformers up to the UFO. I am not sure about the UFO itself, but prior to the UFO, output transformers were Edcor for sure. This is a good thing, as Edcor makes very nice, Reasonably priced stuff. I built the Zen that is in my rotation years ago. At the time I wanted to stick to the circuit as I didn’t want to lose the mojo of the original. In retrospect, there are two areas that might be improved. One is that the power supply could have either a choke or at minimum a little more capacitance in it. My speakers are very sensitive, and the current supply has a trace of hum right at the speaker. That said, I did try adding a chunk of capacitance to the power supply, and it did make the circuit dead quiet, but it also didn’t sound as good as it did before. The other obvious are for improvement would be giving each output tube its own bias resistor rather than sharing one. resistor for bias. However, at the end of the day, the proof is in the listening, and the original circuit has a certain magic that seems to transcend the sum of its parts. The choice of tubes is likely a big part of that. The 6P15P-EV really is a standout in my opinion.
........a little more capacitance in it." quoted Tizman.
Choke(s) is always better than resistors for tube HV supply. It inductance acts as sorta a cushion against the AC line voltage fluctuation. Besides, it sounds much better than resistors.
I installed chokes for my brandname & DIYed tube phono-preamps & tube power amps.
That said, IF you want to add additional filter caps onto the existing FIRST filter cap (the cap immediately after the rectifier tube), make very sure the total first filter cap capacitance can not be over the rectifier tube manufacturer's max specified value.
For 5U4G/ 5U4GB, max first cap = 32uF. Otherwise the rectifier tube will be either damaged quick or shortened lifespan !!
Mind you, not all filter caps sound equal ! The best sounding filter caps as per many many DIYers' experience, is motor-RUN oil filled caps !!
I would never use any electrolytic caps as the FIRST filter cap: sound so-so !! Always oil-filled caps, period.
The for the second/third filter caps, then you may use electrolytic caps of larger values. I am very skeptical in choosing HV electrolytic filter caps: only fast speed ones. I always use electrolytic caps of computer grade used for DC-DC converters - super fast speed.
Don't forget to shunt the electrolytic HV filter caps, each with a film cap to bypass the HF music signals. Very important !
Jack L
JackL: “So I beat it by replacing it with my unique plate-screen grid SPLIT POTENTIAL topology, never seen yet used in any commercial makes”. Very interesting. What does this look like?
In terms of improvements, the one Decware Zen that I am currently using is a DIY SE84CS clone that is basically the exact circuit as factory, with improved parts and a larger than stock output transformer. Incidentally, Decware always used Edcor output transformers up to the UFO. I am not sure about the UFO itself, but prior to the UFO, output transformers were Edcor for sure. This is a good thing, as Edcor makes very nice, Reasonably priced stuff. I built the Zen that is in my rotation years ago. At the time I wanted to stick to the circuit as I didn’t want to lose the mojo of the original. In retrospect, there are two areas that might be improved. One is that the power supply could have either a choke or at minimum a little more capacitance in it. My speakers are very sensitive, and the current supply has a trace of hum right at the speaker. That said, I did try adding a chunk of capacitance to the power supply, and it did make the circuit dead quiet, but it also didn’t sound as good as it did before. The other obvious improvement would be giving each output tube its own bias resistor rather than sharing one. resistor for bias. However, at the end of the day, the proof is in the listening, and the original circuit has a certain magic that seems to transcend the sum of its parts. The choice of tubes is likely a big part of that. The 6P15P-EV is a great sounding tube.
In the mid/late 90s these Russian tubes were marketed and distributed by Svetlana USA as 6N1P and SV83. Svetlana USA published some Technical Bulletins on subjects of applications of these, and the one I would like to bring to your attention was No.27 on the subject of an input and mu stage designed by Cyrus Brenneman using the 6N1P and SV83. Overall gain at the output relative to the input is approximately 30_V/V, 20*log(30)=29.5_dB, while exhibiting low noise and low nonlinear distortion (for a simple vacuum tube audio circuit). And it is single ended, so what nonlinearity does exist does not have the second order harmonics cancelled out.
Again, this was written in the late 20th century, and the comments within are contemporary to that time.
Note that the 6N1P twin triode has both in parallel for a lowered noise floor.
Numerous mu-follower and SRPP designs have been presented to the readers of audiophile magazines in recent years. And yet, in spite of the limitations on present-day receiving tube manufacturing, a large proportion of these designs have used tube types which are no longer in production. Such designs are not useful to the OEM of audio equipment, as he cannot be guaranteed of future supplies of the tubes for his product line.
The mu-follower presented here uses only Russian-made tubes of current manufacture. Its performance and sound are the equal of nearly any other mu-follower yet presented to the public or used in a commercial design. Thanks to the high transconductance and high linearity of the SV83 and the very high linearity of the 6N1P, distortion is vanishingly low at line levels, making this circuit excellent for no-feedback preamp line stages. And the large cathodes of the 6N1P offer superior performance and lifetime to 6922-type dual triodes.
This circuit, when operated on the full 500-volt plate supply shown, can drive a single SV572-3 effectively. If the SV572-3 is operated on its maximum 1000-volt supply, the resulting power output is in excess of 25 watts in Class A2. Other power tubes, such as the 300B, 211 or 845, are also suitable for use in SE designs using this mu-follower as a driver. In any case, it is recommended that the resistors and capacitors in this circuit be audiophile grade. Coupling capacitors should be polypropylene or PTFE dielectric. The 100k input resistor may be replaced with a 100k volume control if desired. The heaters of the SV83 and 6N1P may be operated on AC or DC, the latter being recommended for best noise performance in a line stage. Because the cathode of the SV83 is about 250v dc above ground, two methods of powering the heaters are recommended. One way would be to use separate supplies, one at ground potential for the 6N1P. The SV83 supply may then be stood above ground with a resistive divider attached to the 500v plate supply and then to either end of the heater, or to a center-tap in the filament winding of the power transformer. A simpler and well-tested method would be to use a single supply to power both the 6N1P and the SV83, floating completely free from the rest of the circuit except for a capacitor to suppress induced noise. This capacitor may have to be selected experimentally in a breadboard, for best noise suppression. Start with 0.22 uF, 600v, polyester dielectric. For best lifetime of the tubes in this circuit, a delay in the plate supply is recommended. A standby switch or a slow-warmup rectifier tube (such as the Svetlana 6D22S) are suitable methods of obtaining this delay. An electronic delay relay may also serve this purpose, with the delay set for at least 15 seconds to allow proper warmup of the 6N1P and SV83.
OBSERVED PERFORMANCE (measured on breadboarded prototype)
---------------------------------
Voltage gain at 1000 Hz:
about 30
Frequency response:
<10 Hz to >50 kHz, -1 dB
Maximum output signal into 47k-ohm load, at 1000 Hz, onset of clipping:
125v RMS (354 v p-p).
Distortion into 47k-ohm load at 1000 Hz:
0.08% at 10 v RMS out
0.45% at 100v RMS out
Hi
Ouch! Since when this forum twisted to a "Tech Square" ? I don't mind though.
Before I comment on the above circuit design which can be upgraded bigtime, let me ask technically why these two Russian tubes were chosen in SRPP topology? We got a Russian festival here or what ?
(1) 6N1P
6922 is a much better substitute of 6N1P which draw double as much
heater current of 6922.
(2) SV83
EL84 is a much much better substitute of SV83 but cost a few times
more to buy.
More costly because of better specs: G2 of EL84 max 300V 2W power
dissipation vs G2 of SV83: max 200V 1.5W. This alone render SV83
pretty risky when used in Ultra-Linear output topology.
Jack L
6922 is a much better substitute of 6N1P which draw double as much heater current of 6922.
Both tubes are designed to operate on 6.3+/-0.3 Volts, AC or DC. At that voltage, the 6922 draws 300 milliamperes while dissipating 1.9 Watts and the 6N1P draws 600 milliamperes while dissipating 3.8 Watts. The difference in power dissipation is trivial in a well designed amplifier, and the increased current is only a problem when upgrading an existing circuit from the 6922 to the 6N1P/6N1P-EV in the specific cases where that existing circuit cannot accomodate the marginal increase in heater current.
EL84 is a much much better substitute of SV83 but cost a few times
more to buy. More costly because of better specs: G2 of EL84 max 300V 2W power
dissipation vs G2 of SV83: max 200V 1.5W. This alone render SV83 pretty risky when used in Ultra-Linear output topology.
There was no ultralinear amplifier in this. That is a completely different application.
Choose horses for courses. The SV83 works well in that mu stage. That specific circuit has been optimized for those tubes, and the design would have to be reworked if substituting significantly different tubes, and there is no need to substitute another as it works well as is.
The amplifier in the article uses an output tube that is much more similar to the SV83 than EL84, and is sold as a direct substitute for the SV83. The EL84 has different pinout and, as you mention, can tolerate a higher 300V on the screen grid. The limitation is really in the power dissipated by the screen grid as a result of the voltage.
The amplifier that is the subject of this article configures that pentode to operate much like a triode in a single ended output design, and not as a push pull ultralinear.
And back to the point about screen grid voltages in ultralinear configurations, that is a bigger problem in cheap output transformers where one shared center tapped primary winding connects to the push pull anodes (at 100%) and also to the screen grids at the ultralinear taps (sometimes 40%), lacking a separate center tapped winding for the screen grids, with the anodes and screen grids sharing the same B+ supply voltage attached at the shared center tap. With that, the design selection among supply voltages is compromised, with the anodes operating at voltages below optimal, and screen grids operating at voltages above optimal. Those pentodes would deliver higher output and lower nonliearity with longer tube life with the screen voltage decreased and the anode voltage increased. That is why the better ultralinear ouput transformer designs include a separate center tapped winding for screen grids, to allow a reduced supply voltage to be applied at the center tap of that separate winding and to allow a higher voltage at the anode. Better still are the ultralinear transformers that also include separate cathode tape windings (approximately 10% of the turns of the primary windings) wired in series with the cathodes for reduced nonlinearity and reduced output impedance. The cheaper ones do not include any of those extra cost features, and that is a compromise in the design.
Hi
Yes, who doesn't want a Mercedes to drive to work ? But not every Joe Blow can afford it ! I can but I don't want to own one due to its bank wrecking huge maintenance costs.
Likewise for UL output transformers. Maybe only very few hi-end power amps makers, like McIntosh, still builds UL power amps with UL output transformers with such complex intra-windings. Not every DIYer can afford such expensive irons assuming such hi-end irons available in the DIY supply market !!
As a die-hard DIYer like yours truly, it makes a lot sense to choose pentodes with high G2 voltage, like EL84 (G2=300V) but ones with low low G2 voltage, like SV83 (G2=200V). Right?
Why ask for trouble ??
Jack L
Menno van der Veen offers a good transformer with the windings that I mentioned. He used to be associated with Plitron, and that transformer (or one very similar) used to carry a Plitron part number, and it is now available directly from Menno van der Veen. Note that it has a separate center tapped winding for the screen grids and two cathode tapes (one for each side of the push-pull output).
https://www.mennovanderveen.nl/cms/index.php/nl/producten/specialist/vdv-2100-cfb-sscr-pps-detail
His price is €246 tax exclusive plus shipping. A replacement A470 output transformer for your Dynaco Mark IV would cost you in the neighborhood of $150 plus shipping, and that A470 does not have the added windings, and is not a toroidal. One of the advantages of DIY is the ability to include higher quality components where those matter in the resulting performance without breaking the bank.
To your point about your MB commuter car, the pricey Audio Research VT-150SE likewise has a separate winding for the screen grids, and has four separate cathode tapes. The Dynaco does not have any of that.
Audio Research VT-150SE:
Dynaco Mark IV:
Hi
I think you missed my point bigtime. When I stated in my previous post:
"SV83: max 200V 1.5W. This alone render SV83
pretty risky when used in Ultra-Linear output topology."
For ultra-linear output power stage, why should we choose SV83 with max G2 voltage=200V so low low, risking damage when using standard UL power transformers.
So you want to spend money (246 euro = USD293) to order a more expensive toroidal transformer with separate windings for just for suit SV83's low G2 ? Does it make sense to any DIYers, except to patriotic Russians. Not to me, anyway.
Besides toroidal irons are prone to make noise in the presence of very small trace in DC while EI irons are more forgiving !
This is common sense logic, my friend.
Jack L
Nowhere did I suggest any ultralinear amplifier using either the SV83 or EL84.
The transformer that I mentioned, one of the transformers currently offered by Menno van der Veen, is an output transformer suitable for use in a 100W ultralinear amplifier utizing a quad of SV-6550C or KT88 thru KT150.
Recall that I initiated this by pointing out a late 20th century tech bulletin describing a mu stage using 6N1P and SV83.
The single ended Decware Zen Triode amplifier in Herb's article uses a similar complement of tubes in different application, the same 6N1P twin triode at the input and a 6N15N pentode configured to operate as a triode at the ouput. The 6N15N is a close sister to the 6N15P which Svetlana USA marketed as an SV83. The SV83 was not the same as an EL84, but some designs using the EL84 could be modified to make use of the SV83.
.........configured to operate as a triode at the ouput." quoted JRT
Hi
That Zen amp is a very old-school design which I already recommended to upgrade with my unique 'modern' topology re my previous post here
to Thatguy dated Feb 26.
Such 'modern' topology has already been employed in my home-brew phonostage, linestage & triode-wired pentode single-ended Class A power amp. They all work like a chime for many years now. Musically edged out many commercial brandnames no sweat, IMO.
Read my above post & welcome to comment !
Listening is believing
Jack L
........ either the SV83 or EL84." quoted JRT.
Hi
No you did NOT suggest that.
I did suggest that SV83 should NOT be used in any UL power output amp. Its low G2 = 200V will get the tube damaged if using standard UL output transformer. Then EL-84 would be more suitable in such UL application with its G2 = 300V.
Then you responded by suggesting using Menno UL toroidal transformers come with separated windings for G2 connction. Yes, good suggestion. But it will be 73% (+ shipping) more costly than standard UL iron like Dynaco's.
Want to spend 73% more on the Menno output irons or not is up to DIY customers, right ?
The question is: why you should have brought up Menno UL irons when I mentioned the problematic issue of SV-83 using in standard UL application ? Confusion to our readers, right ?
Jack L
....... which are no longer in production." quoted JRT
Sorry to be too upfront, this Russian SRPP design is a huge 'White Elephant", IMO !
Maybe I've missed a lot, first time I read SRPP design employing a power tube like SV83 which requires very high HV up to 500V !
There are many smarter ways of SRPP designs, most commonly used tubes are 12AU7, 6SN7, 6DJ8 & 6922 etc which are readily available anywhere !
Jack L
Maybe I've missed a lot, first time I read SRPP design employing a power tube like SV83 which requires very high HV up to 500V ! There are many smarter ways of SRPP designs, most commonly used tubes are 12AU7, 6SN7, 6DJ8 & 6922 etc...
I think you might find the mu stage interesting if you investigate it further, and give it a closer look and listen.
A pair of suitable triodes in a conventional symmetric SRPP totem pole with a bypass capacitor on the input tube's cathode resistor (as shown in the graphic) delivers an overall mu (amplification factor, output relative to input), that is approximately 60% of the mu of the input triode, and that is reduced to approximately 40% of the mu of the input triode if you remove that bypass capacitor.
The mu stage described up thread in that tech bulletin delivers nearly all of the mu of the input triode. While you could use an input triode with higher mu, that will likely also increase nonlinearity and will also increase the parasitic capacitance at the input, Miller capacitance of the input, equal to the product of the plate to grid capacitance and a factor which is mu plus one.
So the mu stage doesn't waste any of the mu provided by the input triode while conventional symmetric SRPP does waste plenty, and the mu stage also provides a lower output impedance and better linearity.
.......nearly all of the mu of the input triode. " quoted JRT
Yes, this is a good mu-follower design. When optimumly designed, it will deliver load current = the triodes idle current.
FYI, it can easily be upgraded to a SRPP-PLUS delivering 2xidle current of the triodes into its external load with much higher PSRR (power supply rejection ratio).
Forget the above white elephant like Russian mu-follower schematic.
Instead replace the 2 Russian tubes with one single, say, 6922.
By adding only ONE resistor at the bottom of & in serie of the existing cathode bias resistor and the output coupling cap reconnected to the newly formed junction of these 2 resistors. Done ! SRPP=Plus !!
By only adding one such resistor & relocate the output coupling cap connection, the upper triode is now converted into an impedance multiplier for the bottom triode which acts as a normal voltage-gain amplifier. HV for this RSPP-plus only needs 250V HV !!!!
What is the advantage besides simpler circuit using only half the HV? The bottom triode will appear to have an much larger load impedance & therefore higher gain.
Needless to say, the value of this newly added cathode resistor got to be the right value, basing on the triode's plate resistance Rp, mu value, desigated external load resistance & the existing cathode bias resistor value.
This RSPP-Plus topology is based on the mu-follower design you showed me above. There are always other design variation of it.
Better RSPP+++ designs will be Super-Triode RSPP by adding a transistor to each half of the triode, thus reducing the HV to 150 V !!!!!!
Let compare notes, my friend!
Jack L
...discussion of topologies and distortion: http://nutshellhifi.com/library/FindingCG.html
See also: http://nutshellhifi.com/triode4.html
Hi
Please cut that "Nutshellhifi.com/triode..." long story short. What do you want to tell me specifically ?
FYI, I am a pretty busy working class up to 40 hours a week.
Jack L
...let me ask technically why these two Russian tubes were chosen... We got a Russian festival here or what ?
You seemed to have missed the fact that the text and circuit graphic were from a late 20th century technical bulletin from Svetlana USA (now defunct) intended to promote sales and use of the Russian Svetlana tubes that they were marketing here.
More to the point, it was the similar choice of tubes in dissimilar topologies that prompted my post of that old tech bulletin. That 6N1P was a good choice for the input tube in both, but the SV83 pentode might have found better application in the mu stage, rather than strapped as a triode in the output of a flea powered single ended amplifier. Since Herb likes to work with tube circuits, I thought that he might find it interesting.
Hi
No wonder Russian tubes were used in the above SRPP design !
Strictly from technical view point, I would not use such 'white elephant' tubes in any Mu follower or SRPP designs.
Jack L
I’m thinking about getting a Decware Zen Triode amp and want to also get some (relatively inexpensive) appropriate hi-efficiency speakers and would welcome some input (pun intended).
So far, I am looking at these (all ~ $1K per pair:
Klipsch Reference Premiere RP-8000F
Sensitivity = 98db
Impedance = 12 ohm
Zu Omen Dirty Weekend
Sensitivity = 97db
Impedance = 12 ohm
Decware model DM945
Sensitivity = 94DB
Impedance = 6 OHM
thx
A relatively impecunious and electrically challenged audiophile, I've had my Zen SE84B for about twenty years, returned once for service; they improved its sound with some updates, charged me nothing for doing so. The current model, improved in many ways, is about twice what I paid for mine (in last-millennium dollars)—a big bargain.
Grateful as I am to Steve Deckert, I disagree with him on the power tubes. Almost from the beginning I've preferred EL84's. For the past few months I've been using the Preferred Series 7189 from The Tube Store, and it's the best sound I've had from the little beast. I remember corresponding with the owner/designer of Parker Audio, who made the Parker 95's to mate with the Zen; he also felt he got the best sound with EL84's.
For years I've used The Little Dot MK III as preamp (a transformerless tube headphone amp); it adds warmth and body with no loss of detail.
Re: speakers for the Zen. Check out Caintuck Audio (My LII 15, with Chinese paper drivers matches wonderfully well with the Zen: 98db efficient, easy to place, and disappears when the music starts).
Haven't heard the American Betsy drivers, but I have a lot of faith in proprietor Randy's judgment. I picked the LII 15's for their beauty, and because they reached lower into bass frequencies—in my small room in an apartment, plenty low enough, and clean.
I got mine before shipping damaged so many (heavy drivers) that Randy of Caintuck decided to ship the "baffles" as a separate purchase, and have the customer order the drivers directly from LII. But because one of my drivers arrived damaged, Randy sent me a replacement. Replacing the old drivers was pretty simple, eight screws out and in.
Betsy Baffles come complete. Good luck.