Steve Deckert
|
The adventure continues...
UFO transformers for this amp are getting responses out to 100K with lightning speed which naturally make the amplifier somewhat unforgiving both sonically and electrically. Electrically the penalty is parasitic oscillations of which I have not been able to measure any evidence of... but since this amp reproduces a laboratory reference square wave at 722Hz and uses 100% F&T German electrolytic (The Decware standard in all our amps) the speed is uncanny.
It would seem that that digital streams that have aliasing in the sine waves, is actually able to be heard with this insane microscope on the music. The amplifier is reproducing the aliasing causing it to start and stop dozens of times during the reproduction of the digital approximation of each note in the sine wave. With no global negative feedback to time smear (a rude but effective form of anti-aliasing) the waveform, you can actually hear variations in digital streams as they happen. At least this is the conclusion I came to after hearing strange distortions during audio streams over the internet vs. analog.
After a couple months of listening to this phenomenon, I realized that on digital sources time may have to be stretched to fill the gaps, a sort of mechanical anti-aliasing if you will that can occur naturally in the right type and value capacitor. So right now I'm experimenting with dual inputs... analog inputs and digital inputs where the amplifier is being voiced differently for each.
Also I am experimenting with input tube plate voltages which can be varied by up to 75 volts by simply tube swapping the VR tube feeding the input stage. I started with the VR75 (oA3) and am now on the other other end of the window with a VR150 (OD3) and as predicted the sound is much more relaxed an natural sounding if you're not pushing the amplifier by pairing it with non-efficient speakers.
6N1P's as predicted still largely do not work in this amp, needing more voltage to keep distortions away. It some cases where the user has high efficiency speakers and does not push the amp, the power limiting aspect of the 6N1P may not be an issue, and the warmer sound may prove to be a benefit with lower end digital sources. 6922, 6N23P, 7DJ8, 6DJ8 are the tubes currently being tested.
I guess the reason behind this post is mainly to point out that this is by far the fastest TORII ever produced, and that brings some new challenges to the voicing. Since this amp is Ultra Linear a treble control is not being considered. Regardless of the type of speaker if the speaker is not peaked in the treble, the amplifier has flat treble response.
So far all testing is direct from the source, which is ZDSD, ZCD240, VINYL RIG, 3 OTARI REEL TO REEL machines. Also some bench testing is done using a Gen 2 IPOD touch. This is to say that no preamp is being used, and clearly a preamp like our CSP3 would be a blessing with this amp and almost any DAC.
Basically I'm thick into the things customers will try, and pretty much on the tail end of circuit modifications.
At this stage I can already see the reason for getting a TORII mk4 vs. a TORII JR, is a more forgiving, more bodacious sound... but - the hard core audiophile, especially analog-philes, will have the holy grail of resolution and exquisite timing at a much lower price in Torii Jr.
Soon I'll have two Torii Jr's on an AB switch so I can figure out the best tube compliments and power transformer designs in real time while listening to music. That should be all I need to write the web page.
Steve
|