Rerurn to Romy the Cat's Site


In the Forum: Melquiades Amplifier
In the Thread: Single-stage Melquiades vs. DHT amps
Post Subject: My DHT experiment: ObservationsPosted by Romy the Cat on: 11/29/2008
fiogf49gjkf0d

 Romy the Cat wrote:
There were some educational for myself observations made, in some way saying something about this specific DHT, in some way saying something about Melquiades’ single stage version and the Melquiades’ full-range two-stages version. Differences were very clean. The heads off to this DHT amp, I do not know if it was a good DHT of bad one but it was superbly instrumental in what I was looking. I will pass my observations and interpretation of the results in the second section “My DHT experiment: Observations”, I will right it most like later on this week …”

I will not make any concussions in this section; I will lea it for the last section of my “DHT experiment” journal.  In this section I would like to describe my listening impression during my Milq vs. DHT contest.

Remind you that I used two types of Milq: single stage and full range two-stages. The single-stage Milq was driving a single 15R Vitavox S2 with own RL-line level filter at 3200Hz. The full range Milq was driving Tannoy RED in custom enclosure with original RED’s speaker-level filters. The DHT amp was driving  Vitavox with a speaker-level 3uF filter and then driving the REDs with the same Tannoy’s speaker-level filters. The DHT was well-burned in before listening; otherwise it was out of the scope of any acceptance. The Milq was hot for hours, it was on weekend and my amps are mostly always on during weekends, if it is not summer…

I was running Milq single-stage into S2. I know this sound; it was my “default” “In Macondo” sound, the one that I know very well.  I set the reference volume level on my RTA using my TU-1X’s internal generator. Then I switch to DHT amps to drive my S2 driver and matched the volume my RTA. I wanted to keep the “competition” as clean and possible, at least what I thought…

The sound is different from Milq and DHT. The Milq is more dynamic. OK, not just more dynamic but rather hugely more dynamic. Milq has more contrast, more rhythm, and had more LF articulation if the LF were called upon (and it is in context of 3200Hz high-pass filter). He sound from the Milq was way cleaner and Milq was running across dynamic range with much faster speed and willingness.  Milq had way more extended and way more sophisticated HF. The transient characteristics were not even contestable as the DHT had almost lethargic transients with quite bleary articulation of details.  I would say that the immediate switching between the amps left very little change for DHT, at least at the first level of listening perception.

There was in DHT some positive moments. The Milq was a high-quality performer but it was musically-balanced ONLY in context of the rest Macondo’s channels. The single-stage Milq itself (without the support of other channels) was rather “quality pushing devise” and was in a way over-hygienic. It might be said that it was what Joe Roberts called “painful accuracy” - it had a phenomenal sounds quality, in away hi-fi quality, but it not so smoothly “compiled” along with the flow of Sound, it rather pushed the sound forward. It did not diced and sliced the sound as it was harmonically balanced but right from proper harmonic flow it has a tendency to jump with a bite quite devastatingly. The DHT amp in this over MF only configuration was more balanced, even though less radical and less far-reaching.  I have to admit that with my Vitavox S2 driver from 3.5K and up and driven without other Macondo’s supporting channels sounded more interesting wit this DHT 300B amp. (The last time I heard 300B driving S2 it was with Cary 300B amp and it was a nightmare). 

Considering what I said in the above paragraph and what I will say in the next paragraph (about the respectful loading) would it make sense to presume that the DHT driving my MF was better then a single-stage Milq. Well, it is a bit more complicated then this. The negative moment of Milq driving S2 and the DHT driving the S2 I described but they are binary differences without any regards to anything else. They are just differences but they are not Sound yet. The DHT and Milq did not have differences per se but rather those amps were concentrated upon different way to treat phonetics of sonic language.  Melquiades was about clarity of vowels and tension of consonants. It made Italian soft and fluent, Arabic - harsh, German - barking, Russian - broken up, Spanish – meowing, English – overly logical. The DHT was about murky vowels and smooth consonants. It made Italian oily, Arabic-smooth, German- singable and fluent, Russian- palatable, Spanish- royal, the English’s of Allen Ginsberg it made almost like William Shakespeare. So, the amps did different service and had even different objective. I have to also add that putting the MF channel driven by Milq or by DHT into the play with the rest of Macondo (were the MF is injected with REDs) made the assessment VERY difficult. I did not have time to pay more attention listening the drivers fully incorporated with Macondo. Frankly I did not have much intention to do it. Read up the next 2 paragraphs why.

Listening those things back and forth it was very clearly that my Milq and DTH were respectfully differently loaded. It sounded like DHT has more second harmonics. Milq was loaded in context of the sound that I need to get from Vitavox S2 in respect to the rest of Macondo (1kOh of 6E6P plate, 15:1 transformer and 15R driver). This is a quite heavy load and S2 is this configuration is quite heavily damped:  this how I deal with S2’s intrinsic resonances about 9kHz (I wrote about this in the there what I was fighting with S2). The one-stage Milq is sufficiently damped, however this Milq’s MF channel is “built for speed”.  One stage uses RL filter, very fast RF tube and low inductance transformer with very fast amorphous core.  It has a feeling that it fast and contrasty, but NOT harmonically deficient Sound. Vitavox is not Altec and I do not to need to unload the anodes all the way in order to make my MF channel to sound “exiting”. It might have higher order of 3rd harmonics then DHT but I do not measure harmonics, instead I rather try to learn how to hear and to understand the things. So, I subjectively do not feel that single-stage Milq and S2 driver are sound inappropriately together.  The Milq is DSET and it works in own, much-optimized for HF environment. The DTH is not DSET and it is even not near in the same HF-optimized environment – it sounds more liturgical, sleepy and very much NOT fast. Was it loading due to loading? If I had time I would remap the sections on my Milq transformer and change the load to have harmonics to match DHT. The DHT had only one tap, so there is no ways for me to do anything (it might have internal taps but I did not open it up). I shunted my S2 driver with 12R resistor, effectively loading my 6E6P harder. This method is not the same as the properly loading a plate with active load but I know how this “half-ass trick” sounds and might to a degree to extrapolate the result. I also played with plate currant on the Milq’s single-stage. It insignificantly changed some degrees of differences but the basic still was there – the DTH was just a bit too sluggish with some lack of something that I would call cleanliness, precession, articulation, granularity and exactness …and it is with very fast and high transients-compliance driver of high impedance, the impedance that shell idle the 300B quite nicely.

OK, was what I did a methodologically-kosher way to facilitate the DHT vs. IDHT observation? Well, it was not.

In my view the amps did not work in the same conditions. The DHT and IDHT amps SHELL BE IDENTICALLY LOADED IN SUBJECTIVE EQUIVALENT OF RESPECTFUL TO OWN TOPOLOGY HARMONIC TEXTURE.  (This is VERY loaded statement as the recalculation of ratio and plate impedances very much might not be enough to harmonically match DHT and IDHT amps).  It means that the DHT vs. IDHT shell to output into the same speaker, measured by the same microphone, plotted by the same analyzer and then gauged approximately the same harmonic texture. If we “compare” the differently-loaded amp then do we really compare DH cathodes vs. IDH cathodes or we just compare a zillion other reasons that impact sound? Talking about the reasons: the RL line-level Milq filter vs. a cap at speaker level, the Dominis cable vs. the 2M litz cable, the single-stage IDHT with grounded cathode vs. a large 3-stages DHT intentionally designed to have specific “pre-sold”  character of sound…. I mean: we do not compare the DHT vs. IDHT rather we compare one amp in random conditions to another amp in random conditions – now intelligent and how indicative is it? In this situation one amp might demonstrate itself better or worse but it says nothing about the rivalry between DHT and IDHT amps…

I need also to add that with more neutral speakers then S2 the result was predicable: when I was running my full-range Milq and full-range DHT against the very much not neutral Tannoy Red then DHT produced a lot of very poisoned colorations.

OK, from what I said it is clear that the DHT worse and out of the competition? Well not really, in fact with all negativity that I recognize belongs this amp in context of my S2 driver I need to say that I actually liked this DHT amp and if I did not have my DSET expectations then for a full-range operation I found the DHT was more reasonable with my MF driver, even despite of the above-mentioned shortcomings. To use it would require idling the 300B 2-4 times more that would leave juts 2-3 usable watts. Still, I do not think that I would live with those shortcomings, I would rather to DEAL WITH THEM in order to get the best from each amp or topology; I will talk about the Conclusions later on, in the Third Section of my DHT exploration….

Looking at my comments it is important is to understand the scope of my inquiry. When I said that this DHT was “superbly instrumental in what I was looking” I meant that I was not looking for “better amp” but rather for an opportunity for a DHT to show me something else that I do not get from my Milq. Did this DHT amp show anything “interning” to me? Oh, yes! It was a few things that made to think.

A very interesting characteristic of this DHT was not the amount of second harmonics but rather their colorations.  Loading anodes and increase of second harmonics makes the second harmonic more domains but does not change the color saturation of the harmonics. With Tannoy Red this DHT had purely syrupy sound but with my MF S2 driver it was very colorful and the colorfulness was not as much annoying. The S2 is very fast, very contrasty itself, superbly articulate and color-wire is it very neutral - it worked very well with this DHT amp’s colors saturation. The S2 driver with Milq was colors-wise on a “save side”, neutral and impartial, but with DHI it was although slower and less dynamic but at the same time more colorfully-flamboyant. This flamboyantness had some colorizing common denominator that I did not liked but it was still educational, practically knowing that S2 is the most color-articulate compression driver out there. I listened it very short time (which usually is perfectly enough for me) and I got feeling that the wildness of this DHT coloration (in positive way of using the word “coloration”) was pleasantly-high but at the same time it appeared to me that this amplifier was not able play gray, black and white colors. The absolute color was perfect but the relative colors were … colored.  However, the fact that a “clean” driver might be colored by DHT (or whatever coloring method the amp used inside) was very worth paying attention and it is one of the moments upon which I will capitalize in my further Conclusions. Paying tribute to wonderful DHT color potency I am not sure that this would please me on the long run. I like if music is gray then it to be played gray via playback, if the music cold then to the playback shell have it cold and if an  instrumentalist or his/her playing are crap then I would like to have it to sound like crap via my playback. I did felt the DHT color grandeur but I a bit afraid if DHT is able to play colorfully without taking sides. Also, running the S2 “injected” buy Reds along with the rest Macondo completely removed any DHT’s color advantage. I have some futher idea/directions about it that I will propose in the “conclusions”

There was however a characteristic of this DHT’s Sound that I loved without any concerns, second thoughts and that completely eclipsed my amp: this DHT amp has more random pace and more random texture accents then Melquiades. The DHT had less articulated pace and less defined texture then Milq but the UNSYSTEMATIC AND CHAOTIC NATURE of its timing was truly phenomenal. It is hard to explain in audio terms – you need to look for it in order to see it. What kind association I can give to you to illustrate what I mean? Let look at two great paints:  Byron Janis and Josef Hofmann. Janis, in is best time, was wonderful and in many ways his micro-timing of notes opening was exemplary. Everything Janis did was in great respect to reasons, taste and intentions - it was truly perfect. Then there was Hofmann.  Hofmann’s timing of notes opening (when he was in his mid 20s-30s) was against all lows of physics, against all measure of time and … space. Hofmann hit the keys in his own manner and very frequently a few milliseconds after of before it was expected from a perspective of the “rules” and “logic” for a given piece of music. It was not even milliseconds but rather the Hofmann’s own curvature of time. These micro-timing presumably (very loaded word) random differences between Hofmann’s timing and listener’s recomposing consciousness, that seems like a randomness of micro-events deeply imbedded into the hugely organized flow of musical ideas  did swing listener’s mind with a force of good size hurricane. Here is where Janis was one of the greatest pianists on his generation but Hofmann was a pinnacle upon which all greatest are measured.  So, Melquiades has its perfect timing and perfect pace. Milq is not a damn machine but it is a sensitive, smart and responsive mechanism that is perfection itself.  The level of “timing” perfection with Milq is unquestionable, it is the Horowitz plays Schumman’s Seventh Toccata, - it is what the right sensitive timing and right susceptible balance is about. This DHT take it a step further – it introduced own sense of randomness about that timing perfection. It is where Hofmann with his expressed “unpredictability” reigned supreme and it is where this direct-heated amplifier in my view destroyed my Melquiades. If I was able to fight with some DHT tonal saturation with my Injection channel then I had no tool in my disposal to make Milq to have more RANDOM pace

The next last section of my little “DHT field trip” I will try to share some my conclusion about all of it and will indicate where my thinking is in respect to the DHT subject.

Rgs, Romy the Cat

Rerurn to Romy the Cat's Site