Listening impressions of DSD and Tube amplifiers.

[YashN]

A natural pairing.


A few years back, while I was researching the DSD format I decided that since it is really analogue in nature and happens to be stored in digital format, I would like to pair it with a native DSD DAC or chip-less DSD DAC as well as a Tube amplifier.

It is a natural pairing. I already had my native DSD DAC in the trusty little iFi iDSD Nano, which thanks to those excellent people at AMR/iFi can be upgrade for free to support DSD256 even on Mac OS X through DoP now.

It's an affordable alternative to the magnificent Lampizators although the architecture isn't exactly the same since the Lampis have the DAC I-V and Filter sections all tubed, but it still sounds super sweet! So here is some overview and listening impressions. What are yours?

Troubleshooting

I recently (it was this Sunday!) repaired my DIY Single-Ended Triode Tube amp as well. It is based on a Tubelab design by George and his PCB. Tubes are Electro-Harmonix KT-88, Sovtek 5AR4 and Electro-Harmonix 12AT7EH. Power supply and Output Transformers are all big Edcors. The enjoyment of it a couple of years back was short-lived: a mechanical issue caused a short and one channel was down for a good while until I got the opportunity to work on it and got the help necessary to sold(i)er on.

It turns out the only casualty in the damaged was probably a constant-current source, easily replaced, and there remain a couple of very evil issues with an input connector taken from my salvage box.

Now, work on the amp isn't finished because I still need to finalise some of the Mains switch arrangement as well as allow all its other configurations for better SQ to be available at the flick of a switch: ultra-linear mode, Cathode-Feedback. Perhaps in the future I will also be able to add the following enhancements which also improve SQ: Motor Run Capacitor and Choke. I could also pair the amplifier with more efficient loudspeakers than the Mites as is apparently warranted by SET Tube amps.

Listening Impressions

The first impression I had was of power, especially in the initial sounds. To give an example I would take that of 'the first kick drum' in songs. Compared to my MOSFET SS Amp, the SET Amp just straight away sounded more powerful on those kick drums and you can hear that immediately on the first kick.

This was despite recently getting better sound with the SS Amp because in March, we obtained a couple of dedicated lines for audio.

Bass is clear and not bloated and suprisingly I don't miss the connector of a sub-woofer based on what we listened to. The Piano on Diana Krall music as well as the bass elements on some tracks on a Cassandra Wilson album (Love is Blindness) sound great and deep and not muddied.

That solid basis makes for much better grooves. Toe-tapping, head-nodding or getting up to dance is of the order of the day.

We also appreciated the much more detailed presentation for guitar articulations and vocal expressions. Listening to Al Di Meola's Mediterranean Sundance brought out everything that truly makes this a superlative performance. Listening to offline up-converted tracks by Ella Fitzgerald, we could discern the vibrato on the ending of her lines much more clearly. Lip and voice articulations have become clearer.

DSD + Tubes makes for a stunningly natural and fast set of attack transients and timbre. There appears to be a great separation on instruments.

All percussions are better, and we also get great results with reverb tails. For percussions today, I chose Marvin Gaye's What's Going On in its Detroit Mix version. Just sublime.

One thing I noticed about soundstage though is that it appears more recessed than with the SS amp. Considering that the SS Amp is much more powerful in Watt terms than the SS amp, that could be a normal result.

So if the SS amp is more powerful and the SET Tube amp is less powerful, what is it that is giving me that 'power' impression with the tube amp?

I think it is that 'first kick' solidity and the subsequent solidity in power it displays. It doesn't get congested or 'sags' like the SS appears to be.

DSD256

I have also been through my very small collection of DSD256 tracks, some of which were sourced from nativedsd.com, including the Vanish album (vol. 1). Here again, the presence, naturalness and, especially, the rapidity of the attack were exemplary.

Now, I thought that maybe my first 12AT7 by Electro-Harmonix was blown by the initial issue and so what we had in as driver tube was a generic Chinese tube (no brand indication anywhere).

However, when I tried it, it worked! So I had my first taste of tube rolling (of all things the driver tube only) as well. What I noticed is that there is some more high-frequency sizzle in the upper frequencies with the 12AT7EH than with the Chinese 12AT7. On the other hand, I am also told that tubes can take a good 100 hours to sound their best, so I expect some additional changes over the following two weeks.

So here it is: DSD with Tube amplification. Have you tried it? If so, what did you think?

Here, we're rocking!

 

[YashN]

Two significant changes to my DSD + Single-Ended Triode Tube Amp configuration.

Over this holiday, I took the time to deconstruct and re-construct my DIY USB Cable: I must have reached beyond the 10th version of it, but now I am rather satisfied of it.

It required some investigation and experimenting but at least now I have:

New USB Connection:

- Isolation from the computer's USB power supply (both VBus and GND, not full galvanic isolation because of the need to support the large rates of higher formats and signalling)

- No shield-induced current noise

- Injection of clean power to the DAC

Low Noise, Linear Regulated Power Supply

For Clean Power to the DAC, I have built a Low-Noise, Regulated Linear Power Supply, part of which is based on some clever ideas by John Swenson, and part from the adjustable side of the Doodlebug USB isolator power supply.

The first iteration wowed us quite hard. I was somewhat graced by the choice of an Elna as main output capacitor in the regulated side, having read much good about that brand, because much of the sense of 'presence' comes from it.

Because the bass-to-mid was excellent but there was high-frequency content and rapidity of the upper registers lacking, I experimented more with additional capacitors post-regulator. Initial tests with smaller caps did provide the high-frequency content, but with a lot of 'hardness' to it - not quite the "glare" of the original computer-powered config, but something different in nature. Also tried a solid polymer cap, quite detailed and clean, not clinical, just precise, lacking the presence and roundness of the Elna though (maybe I should try combining them).

Tried some of the Vishay foil caps instead, and there was the high-frequency content back again, in a very delicate manner.

Still missing was some dynamism and rapidity in the overall presentation.

The final piece of the puzzle was found in the addition of a Rubycon capacitor today, so it is now very balanced, without losing the 'presence' that is so thrilling.

Several days of experimentation and work, but nevertheless for really, solid results in the end.

Some friends will bear a triple listening test tonight:

(a) Default USB cable - computer power

(b) My custom USB Connector with split at the DAC-end for external power + iPhone charger (SMPS)

(c) My custom USB Connector but this time with my custom, tweaked Low-Noise, Linear Regulated Power Supply Unit.

 

[jfrech]

very interesting project you have going on...I think you're getting to the essence of the music...DSD, to me, allows a greater connection, than other digital formats...assuming the recording is well done...

 

[YashN]

very interesting project you have going on...I think you're getting to the essence of the music...DSD, to me, allows a greater connection, than other digital formats...assuming the recording is well done...

Thanks @jfrech and Happy New Year! Indeed, there is a fluidity to high-rate DSD, and then into a Single-Ended Triode Tube Amp that I find provides deep insight into performances.

Over the transition of 2016 into 2017, I worked again on my AC Filter box, testing various permutations of the iMac and the DIY Low-Noise Linear, Regulated Power Supply with it, either on Filter Lines, or straight into the main power strip.

It turns out as I had found before that the iMac definitely needs a filter line, but the LRP must not be put on one or else we get losses in transient rapidity, and the TV also needs to be put on a filter line as it is connected to the iMac via cable for screen mirroring.

I managed to get a big power increase with the iMac on a filter line, as well as presence, bass expansion and slam and fast transients with the iMac+TV on filter lines and the LRP on the same dedicated outlet as the SET Tube Amp.

The big power increase was quite interesting - we already had it to a smaller extent with the custom USB connector, and I had already heard it before on its own, but I hadn't expected the change to be that big.

The LRP is a 2-prong floating design, so I had the idea of rapidly testing changing the plug orientation, and we had a large bass expansion (not just extension to the lower end).