A U D I O... P A P E R

APRIL 2001
by Steve Deckert

Diary of a phono stage - Designer's Log


As the day wound down and 5 o'clock approached I found myself looking forward to the quiet time that immediately follows once everyone has gone home for the day. I finally finished my phono stage and wanted to make sure it would drive a Zen amp by itself without a preamp. I've been using my Signature Monoblocks and ZTPRE for most of the evaluations up till now. I was fairly certain it would and thought if that be the case I would have some closure on what has been slightly over a 6 month ordeal of working every night and some weekends on the design.


Diary of a phono stage
Click for larger image


The past 6 weeks have been a flurry of circuits, and I've made so many phono stages it's become hard to keep track of them all. It has been at least a month since I drove a Zen amp with one of these phono stages and while I remember it worked fine, I don't remember which one I used, so tonight will be another hopefully final test.

I have employed the Zen Design technique along the way on most all of the design variations that I've done. All the while confirming that it is indeed the superior method by wasting tons of time building and listening to everyone else's circuits. In any case, while using this technique I made an original discovery that no one has ever done that adds a level of real magic to the sound so I've been pretty excited about it. It's very much a Zen thing and will have most engineer's rolling their eyes but ask me if I care?

So I plunked a stock Zen amp off the shelf (SE84C) and plugged the phono stage directly into it. I'm using a high output moving coil cartridge (Dynavector 10x) on a Rega 3 table as a source because it's an affordable combination that can (I emphasize 'can') sound very good. I want to make sure this specific combination will work with our stock Zen Triode amp.

Well as Murphy's law would have it, the record I grabbed was a little on the quiet side, and the bias switch on the front of the Zen was in the lower gain position so the results were a little iffy. I used my 8 ohm 90dB speakers for this test. More efficient speakers and it would have been fine right off the bat but not everyone will have more efficient speakers. Besides, I'm trying to achieve a 2 volt output with this style of cartridge so I have the same volume as a CD player. That is what my customers will expect.

Almost doing it, but not quite... I decided to take a duplicate (yes, I built two of every design I tried) and add a stage of gain so I could get the dynamics and voltage up a bit more. That went on for a couple hours running back and forth comparing to the original until I finally figured out that the original was just fine, perfect in fact and that it was the particular record I'd chosen that had the reduced level. I should have had more faith. My punishment was the two hours I wasted and the capacitor I discharged across my hand by accident.

Then it happened.... I grabbed an older Windam Hill record of average quality (not a master) and plopped it on the turntable, turned the amp up all the way and sat down for a listen. It was going very good. You know the feeling - when things sound a little too good and you feel like you're getting away with something! By the 3rd track I was so overcome by the sound that I felt a deep emotional rush building up. It was so good I was just floored. Soon tears were running down my face and I just kept shaking my head in disbelief.

Then as if it somehow knew it was gaining control of my inner spirit, It just kept turning on the screws until I had no choice but to surrender to it. It reached right inside my soul and purged all the residual stress I had which resulted in my breaking down and sobbing. In my adult life I can still count the number of times I've had a good cry on my fingers. I was very impressed considering I have even better amplifiers. And I must admit that during this balls to the wall listening session the amp did clip just slightly on a couple parts. I remember dropping the level just slightly and that was all it took to loose the edge or the magic on these 90 dB speakers.

At this point I received my closure - the phono stage was perfect. But, now I wanted to hear my favorite track on the LP again without clipping so I hooked up the Signature Monoblocks and the ZTPRE and went for it again. I also after being so moved by the stock Zen amp wondered if the Monoblocks would be as good. Sonic signatures are delicate fragile things and while almost identical between the two amps, there are subtle differences. Notice how conservatively I worded that?

Well I hooked up the Mono's and ZTPRE and gave it a listen. At first I wasn't sure which was better. There were no signs of stress or hints of clipping but something was different. It was good, but I didn't feel like I was going to get teary eyed again. I realized while listening that what I was hearing was probably the better stereo separation of the mono blocks. Less perfect separation always enhances the center image (vocals) and makes them glow by super imposing the surrounding ambiance over the center image. That changes the timbre and presence of center image a bit.

Then I engaged the dual core transformer on the Signature Zens - something that adds bloom and size to the presentation. Well, holy cow... I guess. Now it was surpassing the stock Zen without question. In fact I didn't realize how good that option on the signature mono's really was. I did realize that you really can't add bloom, only reveal it. On CD's it's easy to tell when the switch is on, but I often go back and forth between the two modes never totally sure which is better. Now for the first time I was hearing it's magic, and there was no question that it was better. No question that there is far more bloom hidden in the grooves of a record than the bits of a DAC.

Naturally this was somewhat of a relief because these amps certainly have more potential than our stock Zen, wonderful as it may be. I listened to another LP and then decided just for kicks to run the phono stage directly into the amps without the preamp. I knew the level would be just about perfect with these 90dB speakers so there was no fear of blasting myself out of the room.

I hooked them and listened to my reference track on the LP that caused me to loose it a while back. Instantly I was stunned by what I was hearing. The quality and perfection of it rose to a completely unexpected height. I simply couldn't believe it. Frankly I was flabbergasted. It didn't take long for me to realize that I primarily voiced my preamp for CD's. If ever there was a reason to push the design a bit farther this was it. All it would take is some 6922 tubes in place of the 6N1P's and the preamp will be up to par with this source.

Anyway, this experience sets what I suspect is the final rung in the ladder. I've never heard anything this good before, and I can't imagine that it could get much better. I had no idea I would reach this level, no idea this level even was possible. So with great confidence I will start building and selling this phono stage and put to rest once and for all the ongoing argument of which can sound better, CD or LP.

Now that I've told you the end result, here is what I've tried:

I spent the first 45 days auditioning the best solid state units I could find. I researched them, and went with designs that were highly acclaimed and received class A ratings in Stereophile. My hope was to find something I liked because I knew solid state could be stone quiet, have no hum, and be compatible with most every cartridge.

It didn't take long to go to batteries. A pure DC supply sounded superior in every way to the best regulated supplies I could build. This meant that to make it user friendly it would have to be rechargeable. By the time I had designed a charging circuit that was foolproof, the battery supply was more expensive than the phono stage itself. Add to this the unpleasant discovery that rechargeable batteries having a different chemical makeup do not sound as good as non-rechargeable alkaline's.

I could get between 5 and 10 hours out of a set of 8 - 9 Volt batteries, and when I reached a point where I'd chosen the best phono stage and best DC power supply the end result was no better than the best DAC's I've heard. In fact it made everything sound like really good CD's.

I then put it aside and did a Curcio designed passive RIAA stage using 6DJ8's in a cascode configuration. I knew this would be pretty decent, and felt confident it would surpass my reference solid state design. It did.

In fact, once you compare the bodacious and huge soundstage of the tube design to the solid state side by side you can't help but come to the conclusion that any more work with the solid state designs would be a complete waist of time.

The problem now was noise and hum. Again, when you A/B the two units, you quickly conclude that the noise and hum are nothing compared to the huge gain in fidelity so you say who cares? Well, reality check. Unless I send a $4500.00 solid state phono stage with every unit so you can compare the two and draw the same conclusion, I'm screwed.

So the new objective was now to get a tube stage quiet enough and stable enough to work in adverse conditions without it's sound or performance faltering in any way. I spent about three weeks building prototypes of this cascode design and ended up with 5 different chassis/layouts to compare noise and hum issues between them. Some were better that others, but overall there was a problem. During testing under adverse conditions the design was unstable, had a tendency to be microphonic, and just undermined my confidence, despite it's good sound.

I then decided to build a couple Dynaco Pass designs so I would have a stable unit to compare the cascode design with. If the cascode design blew it away I would peruse that path. If not, I would possibly come up with an exceptional variation of the last Pass. Well, the cascode design sounded better than the 3 different versions of the Dynaco circuit that I built so I turned back to the cascode circuit. All the while this was confirming my preconceived notion that passive RIAA sounds better than active RIAA.

Another few weeks past and a couple more prototypes materialized around the Curcio design and even though I made some progress, the stability just wasn't there.

I then decided to try some other passive designs, some cascode, some more conventional. I built two variations of each, having chosen what I felt were the best looking circuits on paper. I got one of each style to sound pretty good and felt like I was starting to make some real progress. I still preferred the passive design, although it had more noise and far less gain than the active circuit. It did however have a larger soundstage and better bloom, and was more bodacious.

I played with these two opposite designs for almost two months until I couldn't improve them any more and evaluated the results. To get enough gain to drive a Zen amp (2 volts) was making this project doubly difficult. I now had a couple different passive designs that were quiet enough to get by with, but they needed a preamp to get the gain up. Same with most of the active designs.

I experimented with tube head amps built into each and compared that with studio grade silver wound step-up transformers and rotated between all 4 different turn table/cartridge combinations. Getting something to work on all 4 was proving to be impossible.

In the end I had created some real monsters in the gain department. Enough to take the worlds lowest moving coil cartridge to a full 2 volts of output plus some. And with just the right combination of low efficiency speakers (MMG magnapans 86dB and conventional cone speakers at 90dB) I was able to start really enjoying the music. Yes there was hum, and some hiss, but we were making progress.

At this stage, now having 15 different working phono stages to listen to, I stilled preferred the passive models for sound, and the active models for quietness and gain.

The units I ended up with were the final results of 15 different designs.

1) 4 tube passive in a cascode circuit all 6DJ8's

2) 2 tube passive in a conventional circuit using 12AX7's

3) 3 tube passive with a tube head amp, 12AX7's and 6922

4) 4 tube passive with tube head amp and cathode follower, all 6922's

5) 2 tube active using 12AX7's

6) 3 tube active with head amp using 12AT7's and 6DJ8

7) 3 tube active with 12Ax7's followed by a 6922

8) 4 tube passive using 6DJ8's in cascode with transformer head amp

9) 2 tube passive using 12AX7's with transformer head amp

10) 3 tube passive using 6922's in a cathode follower with transformer head amp

11) 3 tube active using 12AX7's with a transformer head amp

12) 3 tube active using 12AX7's with selectable transformers and cathode follower 6N1P

13) Reference solid state battery driven supply (to remain nameless.)

All of the above (except for the solid state) used the ZTPRE dual mono remote power supply and were dual mono circuits.

14 3 tube active using 12AX7's with 12AX7 head amp and dual mono power supply that was NOT remote.

15) 2 tube active using 12AX7's with dual mono power supply that was NOT remote.

The general idea I had at this point, and I wrote some articles to that effect at the time, was that the best of my passive designs would be my signature phono stage, and the best of the active designs would be my more affordable unit. I spent a lot of time on number 14 and 15 since the cost to manufacture was half of all the rest. In fact I spent 5 almost 6 weeks working with only those two. The end result was a sound I was very happy with, one that had come a long way with all the tweaking, but under adverse conditions and with high efficiency speakers (100dB) the level of hum was just unacceptable. Perhaps this is where some of that residual stress came from :)

Determined to make it happen, I started modifying the power supply with every different form of regulation I could find. This journey took another 4 weeks of nightly work and I had the ripple down to 1/10th of 1 millivolt and there was still FULISDJFLKSDJFinG hum. AHHHHHH

Since I had some other units that were almost identical except for the fact that they used a remote power supply, I moded one, number #5 to be an exact match and compared it. I did not however modify the power supply with regulation. Without regulation btw, the ripple was at 0.7 millivolts.

Guess what, unit number #5 had no hum compared to this, and for the sake of comparison none of the previous units has as much hum either. At his stage of the game I would rather have eaten buffalo shit than hear any more hum. Even though when the music played you couldn't detect it, bla bla bla. My brilliant chassis design for number #15 was a failure, and the on board power supply was inducing magnetic field hum into the circuit. Number #14 was even worse having 10 times the gain, and from time to time the damn things turned into tuners and started playing a radio station.

Long break.

A couple weeks later I was ready to tally up the results and see where I really was. Certainly just saying the hell with it was a clear option. The thought of having customers complain about microphonic tubes, hum and hiss scares the hell out of me.

This brings us to the recent past, where a couple weeks ago I went in the room and lined up all 15 units on the floor. I took them one at a time and listened to each and weeded out anything that was unacceptable from a noise standpoint. It would have been amusing to watch, as I in less than 1 hour had created a reject shelf with 14 units on it.

Sad is that during my diligent efforts with unit #14 and #15 I made a magic discovery that took the performance of my active prototype to a level that surpassed the best passive units. Unfortunately it was at the cost of added noise and hum and I had to abandon it.

I took the one remaining unit that was pleasantly quiet and had enough gain in my hands and cradled it like it was my last chance. It really did sound good, and I'm sure my customers would probably love it, but deep down I knew it wasn't as good as some of the noisier units I'd made. Then it hit me.. the magic discovery that didn't work because of added noise may have been a side effect of the inducted hum from the on board power supply, so I tried it again in unit number #5. There was no added hum or noise, and the magic rushed right back. Now that it was quiet, there was no question this was superior to the passive units. It had better clarity, better focus, better depth, and superior dynamics.

And that brings us to the present time. Personally I have never worked so hard on anything, but then I've never had such a good reputation to protect before either.

The final circuit in #5 obsoleted the passive designs in both sound and gain so out the window was my idea of having two models. But that's okay because #5 will be less expensive to build that my previous designs for the signature model, and will drive a Zen without a preamp making it most cost effective.

Some key features that make this unit special are as follows:

1) I designed it as well as the RIAA from scratch.

2) I designed the RIAA by ear rather than by spec. (Don't worry it's very close - I did check it afterwards)

3) Parts count is at a bare minimum with .022 coupling caps allowing for incredible speed

4) The input circuit has a variable impedance infinitely adjustable from 3.5 ohms to 50K in 20 steps. That means any high output moving coil or moving magnet will bring a Zen amp to full power, and even regular lower output MC cartridges will work if the phono stage is used with a preamp.(or a step-up transformer)

5) The input grid is decoupled with a capacitor that terminates on the plate of a 3rd tube that has the heater series wired with the other two tubes. The cathode and grid allowed to float. This puts a small potential between the heater and the plate in relation to the swing of the input signal. A fascinating result of this seems to be a more relax and believable sound. I played with the screen and cathode possibilities and got different results with each thing I tried, like grounding the cathode for example. I preferred letting it all float.

6) I have found some 12AX7's that sound so good I'm no longer sure 6DJ8's are a better tube in this application. Especially given the fact that with 3 times the gain- a single tube per channel can do the job offering the simplest circuit possibilities.

7) No hum, no hiss, no RF, no microphics and highly stable.

8) Dual mono - noiseless remote power supply with high reliability

9) Can easily be built under the $1000.00 goal

10) Perfectly drives a Zen amp without needing a preamp.

Specifications and circuit details will be available when the unit is released for sale on this web site.

Anyone considering asking me to build a DAC can save their breath. I'm taking a vacation, and the only thing that could possibly motivate me to do such a thing now would be someone proving to me that CD's (even superCDs) can sound better than LPs. When you hear this phono stage behind a Zen amp, you'll understand how ridiculous that really is.


phono stage

The infamous number 14 under construction - one of 74 photos shot of different variations of the 15 units.




Decware is a trademark of High Fidelity Engineering Co.
Copyright 1996 1997 1998 1999 2000 2001 2002 2003 2004  2005 2006 2007 2008 by Steve Deckert