January 2004

Audiolics Anonymous Chapter 52
Products of the Year: Vacuum State DPA Amplifier
Article By Bill Gaw

  Happy holidays everyone. Hopefully, Steve has gotten this edition out on Jan. 1, on a day when all of us can spend some time thinking back on the previous year's system improvements, and make a New Year's resolution to spend more time listening than tweaking this annum. Of course it will be one of many you'll probably break. Thinking back over the past year I have made several great audio and video improvements in my system, which have been discussed in previous articles, that have brought it up to a nearly unimaginable level of reproduction. Unhappily, also at a significant cost to my savings. Isn't that always the way?

Last January I started off with the Electraprint DRD SET Amplifier by Jack Eliano of Electraprint Audio. Voted as the product of the year, this new circuit design produced the cleanest single-ended amplifier sound I have heard. I still use them on three of my horns, but tweaking does go on and they have just been replaced on my left and right front mid-tweeters by a push-pull amplifier that gives the best all around tube sound I've heard to date. More on that below.

February and March's articles were reports on the 2003 CES, the attendance of which has hopefully cured me from repeating over the next several years. So to my fellow reporters I hope you survive this year's show. April saw my review of the Marchand XM-44 solid-state active speaker crossover that still has an honored place in my system driving my center channel horns on one channel, and my subwoofers and overhead speakers on the other. May's gem was the Velocitor Power Line Enhancer from Lloyd Walker of Walker Audio of which I now have four units scrubbing the AC before it enters my system.

June brought a review of the Elrod EPS-3 Signature and the Silent Source AC Power Cords, with my system now using two Elrod's powering my Plinius woofer amplifiers and the Silent Source's running from my media room's junction box to the Velocitors. July's and November's articles covered the Denon 2900 Universal Disc player with its Underwood Hi Fi modifications, which I am using as my DVD-A and SACD player. August brought the best low amperage cable I've ever heard to my system, the Omega Micro Planar AC cords, which are now being used on all of my source equipment. September's review piece, the Loricraft Record Cleaning Machine from England, has replaced my 22-year-old VPI unit. October brought the Audio Monument A La Carte computer Audio Server. December's ended the year with evaluation of several mods and tweaks.

As you can see from the above, it has been a very expensive year for the Gaw Media Room as I put my money where my mouth is and buy the products I recommend. Some I pay for in advance at full list price, others, I get to evaluate before having to pay up and some I get to buy at somewhat less than list. I always complete and send in the review before asking price if I want to keep the equipment, as I don't want to be influenced in my review by the possible cost break.

I often get letters from readers wondering why I give such high marks to the majority of equipment I write up. Some even ask if there was ever something I've reviewed that I didn't like, or, even worse, ask if I'm being paid off by the equipment sellers to give only great ratings.

I do evaluate many pieces of equipment that do not make it to this column. Why is it that most equipment written up gets a favorable rating? For two reasons. First, I usually go by word of mouth from other members of the high-end audio community or with equipment from builders I am familiar with, so most of the chaff is removed before I evaluate. Second, I would prefer to not place in writing a negative review as this country is too full of lawyers willing to grab my hard earned savings for a disgruntled manufacturer. I do not get paid enough by my esteemed editor to place my pension plan in danger. Thus, any piece that would get a completely negative review is sent back with a polite "Thank You" but "No Thanks". Even the high-end magazines with big bucks and their own lawyers behind them are careful for the same reasons.

The one review this year in which I gave a less than positive finding was written because I thought the concept was valid but the implementation was less than I'd hoped for, for the price. Hopefully the designer took some of my comments to heart as I really felt that his idea would revolutionize high-end digital audio if done properly.

Finally to the Products of the Year
This has been the most difficult year for deciding on one or two best products because of the great luck I have had in reviewing superb equipment and recordings. Several of the above would have been Product of the Year in the past, but one piece of equipment and one SACD that I got in October and would have reviewed for December, but had to put on the back burner, stand out as by far the best I have ever heard in their respective fields.

The first is a multi-channel DSD SACD from Philips of the Budapest Festival Orchestra under Ivan Fischer doing Dvorak's Symphonies 8 & 9. That's right, two major symphonies on one disc, and both playable as CD, SACD 2 track and SACD multi-track, done by an orchestra brought up to play Dvorak.

I actually have had two copies. The first one I listened to I was extremely impressed with both its sound and musicality. Then I did my normal disc preparation routine and made a little mistake. With all digital discs, I first clean them with Walker Audio Vivid, then run them through my Audiodesk System lathe, which removes the edge of the disc, then use a Bedini demagnetizer. Unhappily, I did not notice that there was so much information on the disc that it went to within 1 mm. of the edge. Thus, the last minute of the Symphony #8 got lathed away. A learning experience for me. Now I look at each disc first to see where the digits end, then lathe down appropriately. The disc's last minute may have been destroyed, but the remainder of the performances was superb, so much so that I bought a second copy.

First the sound. (This is a high-end audio journal and not a musicology review, you know.) This is the most natural sounding multi-track recording I have heard. I've never been in the recording venue, at the Italian Institute in Budapest, but from the sound it must be a marvelous listening space on par with the best halls, at least for recording acoustics. The electricity must have been perfect and the recording engineer must have used fairly simple mic'ing for a multi-track recording, and not done any major post-session messing around with the digits.

Next the performance. Both of these symphonies are favorites of mine, especially the 8th, and I have heard each several times live and by different orchestras, and have also heard and have had most of the major recordings. To me, this is the best single disc combination of the two symphonies. The playing is electric, especially the last movement of the 8th. The letdown when I found out that I had destroyed the last minute made me so mad that I actually turned off my system for the night, it was that traumatic. I am not a musicologist, just a musicophile, so I won't go into the intricacies of the playing or interpretation. I'll just say that this is the finest all- around recording I've ever heard and for $17.99 list, a steal considering you get two symphonies in surround sound.

Now for the grand prize for the best piece of equipment to come into my possession this year...

I first heard of Allen Wright, the designer, about six years ago when I was evaluating cables, and found his Cable Cookbook, which espoused the use of very thin silver foil and minimal dielectric to allow minimal skin effect and maximize signal transfer purity. While purchasing some of his foil and accessories for building interconnects and speaker wire, (AA Chapter 5 & 6) I also purchased his Preamplifier Cookbook, which turned me on to his theories on electronics. He came to the USA and I was introduced to him by Clark Johnsen, and after discussion, commissioned him to build a pair of combination monoblock phono-preamplifier, -24dB active crossover units for my main loudspeaker channels based on his differential RTP balanced design. One year and several thousand dollars later, he presented me with the units which to this day I use in my system for the left and right front speakers, and which have yet to be beaten by any preamplifier, phono stage or active crossover I've had here. He even came over from Munich, Germany and installed them in my system, made sure the wires that I had made from his foil and design were up to his standards, and then mentioned he was working on a new design for a fully balanced push-pull tube amplifier based on his differential circuit.

Being a SET freak for the mid-high end and solid state for woofers, I was a little reluctant to go for a push-pull design, most of which mask that first milliwatt of sound that holds the space information, and round off the bass, both so important to audio reproduction. But Allen had been correct in his theories on cabling and had done a superb job with my preamplifiers, so, during my last trip to Austria, three years ago, I took a side trip to Munich and brought him a down payment of $5,000 to realize his design and build me two units. I was somewhat nervous about giving $5,000 for an unproven amplifier, but figured his preamplifiers were fantastic, and I was possibly furthering research in high-end audio.

Then, I waited... and waited... and waited. Three years I waited, sending various emails, some of which were answered and some not. Each answer said the design was progressing but not finalized. If nothing else, Allen is a type A personality, this was his penultimate design, and he wanted it perfected. Also, in the mean time, he had designed a mod for the Sony top of the line stereo SACD player which was selling well and taking up his time. Finally I received an email in September that the design was completed, he was building my amplifiers, would be bringing the amplifiers to the VSAC show in Seattle for demonstration, and would drop them off on his way back to Munich.

You can see the circuit design and the theory behind it by clicking here (Adobe Acrobat file) so I won't go into that. What he has done is to build a simple circuit that turns the push-pull output stage into a so-called differential amplifier that makes sure the total current flowing through the two sides of the circuit is perfectly equal down to the microampere level. This in theory should allow the differential amplifier to have the macrodynamics and wattage of a push-pull design with the microdynamics and low level resolution of an SET amplifier. He then used the 6H30pi "Super Tube" from Sovtek as the driver, and his self-designed "Super Reg" power supply.

The unit has three inputs and outputs. The inputs are RCA, XLR, and his favorite Redel plugs. The unit is balanced, and if the RCA is used a special Redel shorting plug in included that must be plugged into its slot for proper unbalanced operation. The outputs are two sets of banana plugs for bi-wiring and a special Centronics 25 pin plug for his silver foil speaker wire. The unit is self biasing, can be used with just about any output tube, triode or pentode, with the possibility of paralleling a pair of tubes per side for more power, and he has already designed units for 300B, and single or doubled KT-88 tubes. I had him build mine using VV-32 tubes as I had several sets left over from my first Electraprint Amplifiers and wanted triodes with a little more oomph than 300B's to drive my horn woofers. The VV-32's can do 18 watts each, or 36 watts total, but they were conservatively set for about 18 watts output, which should allow them to work for years. He used his silver foil for signal transmission, and Lundahl transformers.

As I bi-amplify all of my loudspeakers using active crossovers and loved my DRD amplifiers driving the mid-tweeters, the amplifiers were placed into my system to drive the dual Electrovoice12L woofers on my main horns as my Plinius SA50 stereo amplifiers were at the factory being updated. After balancing the outputs of the separate amplifiers, we sat back to listen, and of course the first piece I played was the first movement of the above mentioned Dvorak Symphony 9, but only in stereo at first. The bass was tight and strong, at least to the 50Hz range of the horn, and was the tightest I had ever heard from my system. Remember, I am not measuring the bass reproduction of this amplifier to another tubed one, but my Distech Monoblock 140 watt per channel solid-state amplifiers. Then the tympani kicked in and I heard the most authentic reproduction of tympani I'd ever witnessed. So they did do the bass better than any amplifier here.

I had no way at that point of hearing them on the mid-tweeters, so I called up a local dealer and took Allen over there for an evaluation. They were using the $35,000 Kondo Nero 2A3 push-pull amplifiers on the Beauhorns with a Kondo IO-J cartridge and transformer to a Kondo pre-amplifier. Probably better than $100,000 worth of equipment, and the sound was excellent with a liquid midrange to die for, and very good bass down to the Beauhorn's 40Hz. limit where his subwoofers took over. Then we listened without the subs.

With the Wright Differentials in the system, and the subs off, the Beauhorn bass tightened up giving more chest compression. There was as much micro-information hall space cues coming in as with the Nero's. Where the Nero's slightly beat out the Wright amplifiers was in the midrange that was somewhat more liquid and natural. Whether that was due to the amplifier circuitry difference or the Nero's 2A3 tubes verses the Wright's VV-32's I have no idea, but it was very close. All in all I was ecstatic that the Wright's took on the mighty Kondo and, to my ears, won. There was even an apparent increase in the low frequency information such that we did not feel compelled to use the subwoofers for the rest of the evening even though Steve threw some very bass heavy albums at them.

The next week saw a further break-in in the amplifiers, with the bass tightening up even more. Then, my Plinius SA-50 stereo amplifiers came back from their update and I had them running the woofers in Class A one channel per woofer, and the Wright units the mid-tweeters. Believe it or not, my first impression, after balancing the system, was a loss of bass slam. The Wright push-pull two-tube amplifier putting out 18 watts had actually done tighter and deeper bass than four channels of 100 watts of "Class A" solid-state power. This amazed me.

The second amazing thing was how beautiful they sounded on the mid-tweeters. You will remember that I had been using the Electraprint DRD 300 B amplifiers since I gave them "Best Product of the Year" award last year. They had the clearest cleanest mid-range production that I had heard, but the Wright's beat them out. The imaging became more 3-dimensional with greater depth and width and fullness to the soundstage, filling out even more the space between the instruments. The Wright's were actually letting more information through. While the VV-32 based Wright's didn't have the 300B romance of the Electraprint's, I do think this was a tube thing, so I'd love to hear Mr. Wright's circuit with 300B's or even 2A3's.

Two weekends ago Lloyd Walker brought back my Proscenium turntable updated to his latest Gold Standard, which I'll report on next month, so I invited Steve Klein, Clark Johnsen and Kwami Ofori-Asanti over for a listening session. While they had not been here in a few months (and we did not change amplifiers out), they all agreed that the system in two channel using the Wright amplifiers was the best they had ever heard it sound, at least until we finally turned it off a 3 A.M.

This past weekend I took the amplifiers down to Kwami's house to see how they compared to his Border Patrol 300B amplifiers, and the new Tom Evans Soul 300B amplifiers from England. All three agreed that Allen's amplifier beat the others in information retrieval and deep bass tightness, with the Evan's coming close, but with the Border Patrol having that 300B lushness that some tubaholics prefer. Whether it is the VV-32's sound that does tend toward the sterile, or the amplifier's, I can't tell yet. For me, their retrieval of microdynamics and very low volume soundspace information make up for the slight loss of SET lushness.

The only problem with the setup now is that I miss the bass that the Wright amplifiers produced; so much so, that I am ordering two more from Allen using 300B tubes, which I will use on the mid tweeters and use the VV-32 iteration on the woofers. That is how much I like them! Plus, then I will be able to see if the difference in the sound we heard was due to the amplifiers or the tubes. Since I'm not into electronics engineering, will let Allen Wright step in now and describe the design.

Allen Wright (Munich, Nov 2003)

I’d like to thank Bill for his kind words and, as invited, expand a little on this new technology. These Vacuum State dpa300B amps are unique in that they are neither Single Ended (SE) or Push–Pull (P–P) but use a third topology: Differential. I had thought we were the first to ever build a fully differential tube power amplifier and still have not found one elsewhere in audio—but recently learned of some built back in the 60’s to control some critical operation in German (not Russian…) Atomic Power Stations!

I’m sure some techies will jump up and claim that Differential is just a fancy word for P–P, and while they are partly correct, the truth is that Differential is P–P done to an engineering optimum, and I believe, solves all the problems of conventional P–P design. In my experience it equals or betters SE in it’s area of strengths, and comprehensively betters it in it’s areas of weakness!

To illustrate these three design topologies, I’d like to offer this three part thought experiment:

1) Consider a thin, perfectly even bar, say 2 meters long (a little over 6 feet) and 2.5cm (1 inch) in diameter. This bar, if balanced at it’s exact center on a precise pivot, will freely swing up and down just like a very good tone arm with zero tracking force applied. This center pivoted balanced bar is the basic concept behind all P–P amps.

But let’s use a large diameter, soft & ill defined pivot such as a roll of kitchen wipe-up paper. The bar will still pivot (at least with larger inputs) but the preciseness is gone. There now can be some sideways movement, and with small inputs it may do so in preference to rotating. The actual pivot point is no longer constant as it will move around the paper roll during rotation, and perhaps even up & down, both causing lopsided motions which are distortions to the movement. Like it or not — this scenario describes a good Class A1 P–P amp’s operation — and there are very few this good out there in commercial amp land...

Next, squash the paper roll a little so the bar sits on a flatish section. This creates two pivot points, one each end of the flat, neither well defined, giving us a dead zone in the center of the rotation as it rocks from one edge of the flat section to the other. This is technically (and sonically) very nasty, but is unfortunately representative of most commercial P–P amps that are and/or have been used in high-end audio.

This dead zone is representative of Cross-over Distortion, present to a greater or lesser degree in all P–P amps operated in other than pure Class A1, i.e. in Class AB1, AB2, etc.

2) Now let’s see if we can visualize the operation of a Single-Ended (SE) amp: remove the pivot and have a friend hold one end of the bar in his hands, as still as possible. With you moving the other end, the bar will respond to you inputs, large, small or microscopic, but (ignoring any spring functions) will have two major peculiarities:

a. the force needed to move it up will be different (greater) than that needed to move it down—due to the effects of gravity, and:

b. there will be unwanted and non linear distortions in the movement as the far end of the bar unavoidably wriggles around in your friend’s hands. This is similar to what occurs in ‘classic’ SE amps, but of course can be solved by making that end’s “termination” stable—such as anchoring it to a solid wall with a rigid clamp. This is what is done in a few of the very best SE amps by using a very stable power supply—but even the very best power supply cannot correct the asymmetrical force problem of "a".

I allow that this SE visualization is perhaps a less accurate description than the P–P examples above, but never the less it demonstrates the key plus/minus points of the Single Ended topology: very accurate with small inputs (microdynamics) but becoming increasingly non linear with larger inputs. This non-linearity shows up as added distortion, mostly 2nd harmonic, and some people like the added ‘warmth’ this brings. But the point missed by most SE fans is what’s described in a/ above, which translates to different output impedances (damping factor) on positive and negative signal swings!

Simply put, it means that all SE amps exert a different amount of control on a woofer cone when it’s moving inwards to what it exerts on the cone when it’s moving outwards. This is unique to SE amps, and is the prime cause of their loose/inaccurate/odd bass tonality on certain speaker systems.

3) Now let’s unclamp our bar from the wall and pivot it at it’s exact center on a very sharp hardened steel knife edge: now the pivoting of the bar will be absolutely precise and extremely sensitive to all influences, even minute air currents. The bar will appear to float in space—no motion lost or added, and it’ll respond accurately to all inputs. It’s movements will be totally symmetrical and as it’s perfectly balanced, will need exactly the same amount of force to move it up as to move it down. This describes the operation of a correctly implemented differential amp.

Bill ordered his first pair of amps after being pleased with his Vacuum State RTP preamp, but while they were still at a conceptual stage. So he had to wait quite some time to have them delivered as considerable development was needed (mostly to do with providing a really consistently accurate and razor sharp ‘pivot point’) before they were ready for prime time—but with prototypes now running successfully for over three years with a variety of output tubes, they are in production and (I believe) are a paradigm shift in tube power amp technology.

Production amps do not have the front panel meters or knobs shown in the photos of Bill’s units. These are to set the vital DC balance of the output tubes, but in production units this is done automatically (and even more accurately) by an electronic servo circuit. And those trick WW2 Siemans meters are anyway unobtainium...

I set Bill’s initial pair of amps to specifically suit his chosen VV32B output tubes, but production amps can use any (5V heater) 300B variant from the cheapest Sovtek through real Western Electrics and right up to the monster 52B tubes from the Prague consortium. An unusual feature is that while the power output with classic (WE) 300Bs is 18 watts, it can be set to over 40 watts with the 52B tube (or anywhere between) by changing just one resistor in each amp. The power supply and input/driver stage handles this more than 2:1 power ratio without sonic compromise!

I thank Bill for his beautiful review, and even more for his patience and trust in my skills, and end off to say that delivery time on the dpa300B is now around one month — demand and production capability (they are 100% hand built) allowing. Expected US retail will be circa $15,000 — but until a distributor is finalized they can be obtained directly from Vacuum State at the ex-factory price of 8,800 euros/pair!

Allen Wright (Munich, Dec 2003)