"My vacuum tube output stage perfected almost 20 years ago. Applying it was like opening doors on a new room full of candies. Since Lektor III, all my players used tube outputs. First I applied the famous E88CC/6DJ8 twin triode, a nice choice for this application. The most famous became the Russian 6H30 initially promoted by BAT. This has high current and low output impedance. In my application it gave more kick, space and sonic colours. Over the years I improved power supply parts. You reviewed Lektor Prime with a balanced tube stage. This circuit remained unchanged for years. I just applied better capacitors, finally V-Cap CuTF copper/teflon on the tube inputs, oil-impregnated ODAM as output capacitors. Volume regulated by analog resistor switch inside Burr Brown's PGA 2310 chips. I then tested and applied this combo of PGA volume controller + 6H30 tubes +V-Caps in my preamplifiers with very good results. So when in February last year I started to work on Lektor Joy, this solution was the main pretender to the throne."
Lektor Air precursor with granite plinth, Philips CD-Pro2 transport, tube output stage, analog input and remote-controlled analog volume.
"However, the prototype PCB included a few options. I could quickly compare different circuits with jumpers. The next trial eliminated the analog volume control chip and used digital volume instead. Earlier this hadn't sounded right because it always 'muted' the signal by decreasing resolution. But for Lektor Joy I picked the best current DAC of ESS 9038Pro. With its 32-bit volume, 16-bit CD retains sufficient resolution even at low settings. The next trials compared tubes to IC and discrete transistors. Because I'm an older man with time limits, ever since my first Fram project I now always design an entire family of new products at once. It saves time. I thought that with a simpler version than Joy I might use Burr Brown op-amp outputs."
"Then everything changed within a week. Russia attacked Ukraine. This created instant supply issues of Russian parts not only for current but future inventories. And buying anything Russian would merely finance aggression. For many people this war could seem far away but for Poland it's right at the border. The lands invaded were once Polish. My dad was born in Lviv, Polish before the war, now Ukraine. Their people are our relatives despite our painful past history. So I began exploring new ideas. First I compared analog vs. digital volume again. In theory analog would be superior since it works the DAC at always full signal and resolution. But it does require an extra circuit, here the PGA 2310 which digital volume eliminates. And the shorter signal path did produce more direct better focused soundstaging filled with more venue reverb. The tube/transistor comparison was more complex.
"In truth, the difficulties associated with Russian tubes gave me extra incentive to develop a competitive solid-stage solution. Fortunately I had already done a lot of work on my Silver Stage phono stage, a set'n'forget device of low power consumption and high reliability impossible with tubes. My phono preamp used discrete field-effect transistors. It reminded me of the most difficult such amplifiers to design, vertical oscilloscope amps which need very high gain, DC-GHz bandwidth, linear amplitude and phase impossible to achieve with op-amps. My phono preamp had almost 30 discrete hand-selected transistors across five gain stages. Just so it delivered massive, warm, emotional and smooth sonics. Now I applied the same concept with the necessary modifications to my CD player circuit. A bank of relays allowed instantaneous remote-controlled switching between transistor and tube output stage from the seat. First impressions were promising. The op-amps came dead last but Jfets and tubes were close."
"For a second opinion I approached Hifiphilosophy's editor Piotr Ryka. His understanding of music is impressive, his opinions are independent and I've asked his feedback many times before. Over a weekend he compared both outputs by switch. He knew what he compared, just not which switch position was which. His impressions confirmed my doubts. For extra feedback I asked two more audiophiles, Tadeusz and Dark. Without fail my 'golden-ear' acquaintances came to the same conclusion despite working blind. When I told them that it was the transistor output which was softer, warmer yet more energetic than the tubes, they didn't want to believe it. So for final production I went with the transistor output stage. It's simple and pure without global negative feedback and only adds a single stage to the post-DAC signal path. The topology remains similar to tubes—in fact Fets are voltage controlled like tubes—even duplicates their V-Cap ODAM output caps. Yet output impedance is three times lower and gain higher. The benefits are enormous because there's no tube aging for zero maintenance, no heater power where the 6H30 twin triodes needed 12 watts just for cathode heating… and no Russian parts. Of course there's a disadvantage too: how to convince audiophiles that a similar Fet solution outperforms the classic famous 6H30? This will be up to reviewers."
