OK, now I recognize my S2 driver, thank God.

I was give to my new PS for MF channel one more change today to show itself off. I relayed the Hartmann ‘s first movement of the Shostakovich First Violin Concerto. As I said above the progress was very nice and the majority of the problems that I was experiencing at beginning were not there or were at much-much less degree presented. Still, there was an effect that I did not like, and since the bigger problem got burned out this “effect” become to annoy me quite annoyingly. I am talking about some strange micro-details and micro-dynamics eliminating noise. Those who use S2 driver properly knows that this driver has an ability to track all engravings of musical texture with a level of discrimination that no other compression driver can approach. So, my MF does not do “tracking” but rather sliding across the texture of event, filing the dynamic indentations with some kind of generic noise. Furthermore, when a tonal event decays into ambiance or is cared juts above ambiance then the event do not decay into “ambianic nothingness” but rather it dives into a “foggy noise-like substance”. Do you know what happen next? What the tonal event ended then the “foggy noise-like substance” is ended.

I really did not know what it is. Bad electricity gives it but left channel have no such an effect. So, I decided to take actions. I got rid the last 100uF Ansar polypropylene cap in MF power supply and replaced it with my well tested and well beloved 250V/3000uF  Nichicon electrolytic and just “in case” bypassed it with 100V/1uF MultiCap RTX.

When I turned the amp back (again the MF and upper bass only) it was with the very first none obvious that ALL problems are gone. You remember the bass section opening of the Shostakovich First Violin Concerto? Before, playing it juts with MF it was not sound from channel but a spray of radioactive noise.  Now it was… nothing “radioactive”. With the very first second of the play I know that the problem was not there anymore. The rest of listing was nothing new to me and I did not detect any issues of any kind that would make me to be concerned. So the journey with power supply for one-stage amps is over. One more notch on the project. Now I might covert the left amp to dedicated PS for one-stage amps and the entire project of 6-ways Super Melquiades will be over

Does the new dedicated PS for MF one-stage amps sounds better than in the situation when I sors3ed voltage from B+ of 6C33C? I do not know. Perhaps before I convert the left amp I would need to make this observation for myself… Perhaps I should… anyhow, now I am happy. Well, happier…

I bet if you take that RTX off it don't sound as good.

jh

Here is a semi-final draft of 6-chennals Melquiades Power Supply. Since I use my site pretty much as a notebook would like to have the diagraph posted in this thread for my personal references.

http://www.RomyTheCat.com/PDF/6way_Melquiades_PS.pdf

At this point I will pu on hold my interests with use the gas tubes and with 50W zener 1N3350 diode instead of capacitors in singlestage amps as what I have now does incredibly well. I have a bank of 5  2uF film caps that run AS signal over 500Hz. Enough...

I pack up the new amp today and plug it in with my right Macondo channel. The electricity was not bad I was VERY surprised with the fact that the new amp sounded MUCH more advanced then I anticipated from a chance of MF channels power supplies. The lesson learned: never ever ever ever try to save money and efforts for MF channel. I will talk more about the Macondo-Melquiades New Sound but for now I would like to mention one monumental achievement that the Macondo-Melquiades tandem has accomplish.

I has NOT been talking about it for years  as this subject is absolutely unknown to audio people and frankly I never head heard from ANY playback  that encourage to think about it in context of playback. I am talking about an ability of playback to create mechanical sound. I am talking about Mechanical Sound reproduced by the acoustic system of playback it never happened...

Never say never and I have heard it before. When I was 10-11 years I played my grandma 78s. I still very distinctly remember that sound and I do remember that grandma’s 78’s did have the sights of mechanical sound but they had not fidelity of Sound.  A few years ago I experimented with intentionally wrong cartridge mounting that mimic the Mechanical Sound, again the fidelity was not there. Now, completely without my anticipation the Macondo-Melquiades throw a very surprising ability to attack the Mechanical Sound only it dose it at very serious level of fidelity. The sound is wonderful but all this none-musical events of recording got a new level of realism, kind of spooky realism.  Do you know that feeling what you drive a car with very high level audio installation and NPR play  some kind of ad which has siren or beeping in it and you for a second do not know it is your radio or it is somebody signal you from the traffic outside?  The very same with the new Macondo-Melquiades. The individual sounds get a new crazy level or realism, so real that I was confused at beginning. The slips of bows, the air coming from valves, the plucking of strings upon fingerboard, the sticking of the strings to the abrasiveness of the skin suddenly go not the stupid prominences (courtesy to the stupid OTL amps) but a very distinctive sense of realism – they just do not some like recorded. I was initially thinking that my S2 drivers was scratching gap but it was not.  The sound that the MF channel is as close to be mechanical-like as I ever experiences – very very welcoming sign.

Sure, music is not the artifacts of recording but do not worry the musical information itself is presented at the very same level of Mechanical Sound the level of Mechanical disambiguation and sensible reality. Interesting that the Mechanical Soundnes of music is very useful but the Mechanical Soundnes of the recording’s artifacts sound… a bit strange as now.  Each time a cello develops a body crack  or the player move the hard stopped finder across the bass strings then I really feel a temptation to walk to the speaker and  to realign the diaphragm…

Good concern, I had and have it as well. I do not have an answer at this point. It should be stressed that my description of artifact was just my specific description of THE artifact, the MF texture  music had also a “face lift” -  I just did not speak about it. 

Still, a “better” sounding musical instrument does not create a Pavlovian Reflex to Mechanical Sound, the artifacts do. I am family very well with the hyper-resolution results and I am concern do not be hooked on this, again and again. I do not know as now how the Intensification of Reality with new Milq MF channel will work out for me on a long run. Even if it is the hyper-resolution then it is a hyper-resolution at another level and I have with this DSET some means to moderate the harmonics that supports that hyper-resolution. I never was able to do it with any other situation without screwing anything else. Nevertheless, I do not feel, at least at this point that the “Mechanization” of Milq’s MF Sound is some kind of Surrogate Realism. You see, the motor reaction to some things that the MF does now are very nature and very not contrived. It is like a dog that hump a picture of a dog that he sees in a book… it is kind of different experience then to hear a hyper-resolution layback, it is hard to explain and it should be heard.

Sure, I need to live with it for a while to see what happens next. I am wondering what the reason was.  Perhaps the people who were bitching about the electrolytic caps were right after all. In the MF channel in a past I got a major advance when I got read of cap in line-level filter replacing it with a coil. Now my MF signal has no capacitors in signal path (including PS) beside 10uF of very high quality Cube 950. Perhaps that made the trick? What else I change: a new transformer, new choke and new Schottky diodes… I think still it was primary die to the new way of use capacitance - in DSET codification I might go away with it.

How much useful value it has? Well, I will wait for a final conclusion in a week or so when the PS with its caps will burn-in but so far the result is more than pleasing and very promising. BTW, what also is very interning is that now HF can tolerate +6 dB without being too hot.

N-set wrote:

just a minor tech remark (feel free to ignore): the big-ass elco is theoretically still well in the picture. 10uF will have 320Ohm@500Hz, which is a fraction of 20Ohm R+3mF elco.

As I understand the AC signal flows from anode to PS and see 10uF cap and 25R resistor. It will pass the 25R resistor and go to electrolytic when the filter what the AC frequencies sees is above the 25R resistor. With 10uF cap we have 25R resistance at 630Hz so the higher frequencies (above 630Hz) flow to ground via the film 10uF capacitor and the lover cross frequencies the 25R resistor and get shorted in the big electrolytic cap. Well, of course the “the big-ass elco is theoretically still well in the picture” as it is just a first order filter and not a break wall switch and the higher frequencies are juts divided between the paths back proportionally to impedances. Still, at 630Hz we have 50% go to film and 50% to electrolytic, however the MF is filters at 3200hz that makes juts fifth of AC signal go to electrolytic cap.  BTW, I made another tweak in there and just before the electrolytic cap I put a poster for electrons saying:  “If you are HF then turn your ass back to the film capacitor”. It looks like since I pit that poster the sound become “much better”. Now, the film cap is made in Mexico, so wait until I pit the poster in Spanish….

Here are the basics of the 6-Chennal Melquiades DSET Amplifier complete, without secondary features.

http://www.RomyTheCat.com/PDF/6-Chennal_Melquiades_DSET_Amplifier.pdf

N-set wrote:

… I think that there is something more interesting going on here: what you have here is rather a rare instance of an intelligent bypassing…

Yes, it is highly intelligent and looks like it is highly effective sonically. It is not my idea and the intelligence came from my advisers, one more time: thanks to anybody who have played along.

What again and again I would like to stress is that the opportunity for that intelligence actually have organically derived from the nature of DSET topology and I was able to go away with it because the MF channels operate in a narrow bandwidth and have dedicated power supplies. In case of a full-range oration we would defiantly would like do not have a frequency-reactive power supply but in DSET world each supply might be explicitly optimized for a narrow operation range. In fact, I am so pleased (so far) with what I got that I am not violently considering a thought of converting my upperbass channel to the same pattern. I can raise PS impedance for upperbass (as it is high-passed) and if I found a very good quality 100-150uF film capacitors (not the Ansar/Solen crap that I have) and if I found a space in my amp then I might go for it.

Paul S wrote:

So, separate wipers only for injection and HF, or are there other means to adjust relative levels amongst/between all channels?

op.9 wrote:

Could you help me understand your biasing scheme with this version? I thought I understood the original Milq circuit - but the only channel that looks like that to me is the D channel. I don't get A and B at all! what is the 50 ohm resistor doing? is it just a grid stopper? I assume you can adjust bias and 0v at input in each stage - but don't the two red legs (on channel B for example) work against each other?

james,

I was tiring to do it before in:

 http://www.GoodSoundClub.com/TreeItem.aspx?PostID=5940

….but here is even a better opportunity. In the recent circuit I drew (BTW, I very proud about it) I made the positive and negative bias lines with different colors that makes the entire circuitry extremely easy the understand.

http://www.RomyTheCat.com/PDF/6-Chennal_Melquiades_DSET_Amplifier.pdf

Here is a stupid need of DIY audio level. As many of you know am big lover of Nichicon electrolytic capacitors, even their lowers grade sounds right to me and never made to think about anything better. Still the Nichicon never stopped to surprise me.

In Super Milq I have a first caps in the B+ of the first stage supply. It is 100uF, nothing special. I use usually regular LK line, which is base commercial grade. However, I make a stupid mistake and instead of 450V put there 250V/100uF cap. The voltage after the bridge is 440V with 115V on primary. Now I have much high voltage in mains that make on secondary 475V… and it is all loaded after a choke with 73 DCR to a …250V capacitor? So, what do you think happed when I connected it? Well, it is not only nothing happened but…. I realized that it was running like this for a week!!!! Wow!!! The heat off to the Nichicon Corporation and their capacitors!

I am kind of at lost, as what is doing on with my playback is surprising to me. With a pair of new 6-ch Milqs, more or less acceptably calibrated (but not perfectly), the system begin to throw some very astonishing effects, so astonishing that I am very convinced that had I never heard Sound like this nether from my playback nor from any other playbacks . The “new Sound” is not “goo electricity” related as I know what comes from electricity – what I am getting now is not that.  The new 12 channels of dedicated amps, direct driving Macondo’s channels made some extremely profound “strategic” advancements within Second Level of listening perception:

 http://www.GoodSoundClub.com/TreeItem.aspx?PostID=59 

I do not know what responsible for it – lowering of intermodulation distortions or anything else but the fact of recent huge positive changes is undeniable. I will post in future some thoughts about some specifies of that “new sound” but for now I would like to chew upon what I am getting from the playback.

Something had changed very positively with multi-amping. I am not quite sure if it is because the full multi-amping of the specific of the implementation but it is what it is. The entire dynamic relationship between volume of a given tone and the tone’s vibrant properties become much more remarkable. There are many other inversing observations about the new Milq sound, and they relay keep me in unexpectedly-delighted state, but at this point I would like to accent this one: Melquiades got smarter.

If you read the few comments below (let try my new search engine):

http://www.romythecat.com/GetPost.aspx?PostID=4247

http://www.romythecat.com/GetPost.aspx?PostID=3207

http://www.romythecat.com/GetPost.aspx?PostID=6048

http://www.romythecat.com/Forums/ShowPost.aspx?postID=672

http://www.romythecat.com/GetPost.aspx?PostID=2091

http://www.romythecat.com/GetPost.aspx?PostID=1816

..and many other comments from the period when I was discovering Milq’s sound then you remember that one of the most remarkable and distinctive characteristic of the Melquiades amplifier I consider something the I called Milq’s “Dynamic Viscosity” and the “Volume Intelligence”. It means that Melquiades has own intellect that knows how to react to soft and to loud passages differently. It is all comes from the bias of 6E5P/6E6P but the technicalities are not subject of this post. So, Milq’s driver stage kind of self-adjusts itself and reacts differently to loud and soft moments of music and the driver stage somehow implements the velocity of changes of loudness VERY differently than any other amplifier I am familiar.

I have to make a comment here. I certainly can say any BS about Melquiades. Very few people ever heard driving properly loaded and practically no one build Milq out there. So, I can go away with any of my drooling, even foolish, about actual or imaginary advances of Melquiades Sound. Well, whatever it is. The Milq is publicly available, feel free to accuse me in laying. However, do not approach Milq from the position of senseless stupid engendering of from a position of DIYAudio counselor – I will send you to fuck yourself and it will be the only activities you will be qualify to do in audio. Many of you have been doing it for years, despite of “30 year of experience” and own sense of audio pretentiousness…

So, the “Dynamic Viscosity” and the “Volume Intelligence” is the very kinky keys of the Melquiades sound but what have happened with it in the 6-Chennal Melquiades DSET?

Well, the amplitude of the Milq’s Viscose Intelligence got wider and the Volume Intelligence have increased it’s range. Before Milq was able to cared loud passages with force and aggressiveness if the music called upon the aggressiveness but at the same time Milq was able to be very gentle and very considering when music called for placidness and softness. Now, with 6-Chennal Milq DSET the forcefulness of the higher dynamic range got multiplied by a very large number. That is itself is a phenomenal think but along with advances in vigor at upper dynamic range the amp go absolutely stunning sense of effortless of getting there. The musical crashes got cleaner, less confusing but at the same time much more intense and more passionate. Perhaps it relates to loosing intermediations via multi-amping of perhaps it is because the way “cleaner” sound from upper channels… I do not know. It is kind of interesting, the upper channels during loud passages sound almost “thin” but as soon I inject more second harmonic into them I realized that it is not “thinness” but a very new level of cleanness. Balancing that new “clean” MF/HF sound with proper lower MF and Fundamentals super I got so slick sound at orchestral crashes that I make me audio-ecstatic…

OK, from another side there are low-level soft signals… I have no idea why but the low level discrimination of details went way over the roof. Even at 16 bit if I go for -50-60dB signals they suddenly became to sound with unseen character and unseen temperament. It looks like Milq, even at the level where  16 bit begin to bare itself into nose grabs the information and pamper it with some kind of special attention. I found myself during the last few days to drive my high voltage output DAC to drive wide opened New Milq and playing music recorded at very low level (opening of Pathétique, opening of the Saint-Saëns last symphony, some Mahler, etc…). I was surprised again and again how Milq could dive into depth of dynamic range and search there for even of its attention…

However, the most stunning thing is HOW Milq moves across the dynamic range – it does it in a way how it wants and I love each second of it. It is hard to explain and it needs to be experienced. It does not do it too fast (typical for OTLs and another crap) or too slow (typical for DHT with overloaded driver stage). Milq does it with own pace, the pace that Milq chose for itself at given moment… according to the current musical volume and the dominating frequency range. Absolutely amassing feeling, particularly taking in account the Macondo/Milq tonal pyrotechnics…

Let me in the end give you an association about the Melquiades “Dynamic Viscosity” and “Volume Intelligence”. Pretended that you are at the bottom of a tall building and somebody throws a very fragile glass vase to you. The mass of the vase is 1Kg, you need to catch and your task is to ketch it at minimum distance from ground. If you catch it too high then the vase will be stopped to high ground and it you catch it too low than you will have too little space to apply the necessary contra-inertia movement and the vase will be cracking in your hands. You have eventually figure out how to catch the vases and how slow down their fall in order to keep them from cracking and in order to stop them as close to ground as possible. Now it is not just 1Kg vase is thrown but many different vases with different masses. This time you need to moderate dynamically the elevation where you catch the vases and the strength you need to apply in your hands in order to compensate the mass inertia, still keeping the slowing down of the vases with the acceptable G- acceleration             that the fragile glass would be able to survive. Now, If you got what I meant and understood all my misspellings, then you have a basic understanding how Melquiades “Dynamic Viscosity” and “Volume Intelligence” might sound.

Well, as I understood it the strategy from the beginning was to use a number of specialized/optimized "channels" working together as opposed to the usual single amp or two amps for FR, and from your descriptions above it seems like you're getting what you sought, along with some bonuses.  No doubt the specialization has factored large, but in looking at the schemos for both the amps themselves and their power supplies, they are pretty sophisticated, as well.  Nor did you stinge on transformers...

Not to rush your exposition, but how are LF and HF now, and how is channel/channels integration?

What are you using for pre-amp chores?

Have you tried phono sources yet?

BTW, Congratulations!

Best regards,
Paul S