How did you choose the right tubes for this project?
The choice of tubes was critical for many reasons, not all of them musical or technical. The current market is very complicated. Not long ago the world's biggest tube factory in China's Shuguang was damaged. That overloaded other smaller suppliers with orders so they raised prices or changed distributors. This disappearance of the world's biggest tube player caused a catastrophe. Another key factor was Russia's invasion of the Ukraine combined with the fact that after China, Russia is the second largest top tube manufacturer. Most Russian tube factories not working for the Russian army were taken over by foreign investors or investors who owned the rights to Western tube brands. These two facts together make the complicated situation of tube brands even worse. And then we need wholesale quantities. We can't buy small quantities on eBay. A few NOS tubes from the 1950s are no solution. We had to pick from 10 perhaps 12 available options not involving Russia. We tested them all and chose the best. The process was simple, logical and about stable repeatable parameters. You must recall that tubes whose parameters vary by a few percent may be acceptable for an amplifier but when it comes to phono with amplification factors of x 1'000 to 3'000, initially tiny differences multiply to result in amplitude and channel imbalances. Hence tube selection is critical. We need large quantities to select from a big pool since we reject so many. We need a few hundred pairs, measure them, break them in, pair them then test them again in-circuit to insure that RIAA equalization remains within tight tolerances. I can honestly say that the tubes we chose are best because they deliver excellent performance. We haven't found better and ours are available and reliable.
The most critical element due to its place in the signal path is actually right at the input so the SUT including its primary winding. We tested several and built numerous prototypes. Another challenge was to ensure enough gain to work properly with any cartridge. A lot of top pickups only produce 0.2mV and we couldn't afford not to support them when they cost $5'000 and their owners should be our customers. For 0.2mV we need another 10dB gain atop the global 70dB. Yet another consideration were the load settings. We decided to stop at five and offer some standard and personalized values. We can't go beyond 2kΩ though.
But there is also an MM input that requires 47kΩ loading?
Yes and it was another challenge to offer an MM input that bypasses the SUT. Here we had to compromise. Standard settings for MM cartridges are 47kΩ and 100pF but true aficionados expect more. We couldn't offer that in the VP4 but perhaps one day we will design a phono stage exclusively for MM cartridges to add more adjustments. With the VP4, the MM input is not meant exclusively for MM cartridges. Why would anyone pay 30K PLN for a phono stage with a 300 PLN pickup? It makes no sense. The high-end guys use MC cartridges. Yet a lot of them also expect an MM input because they want to use their own SUT. If you have a Silver Kondo SUT, all you need is the last gain stage or a top-quality MM input for your $20K cartridge. That's why we had to treat our MM input as seriously as the MC. Customers who require an MM input are often twice as demanding because they use it with a cartridge and SUT that cost as much as a good car. And this is an important reason to take proper care of this input's loading because it sees a SUT and the SUT sees the cartridge and finally the cartridge via the SUT sees our input.
What's more, the MM input may be used with mono cartridges. A lot of customers who listen to mono records usually have a second tonearm with a mono cartridge, often an MM. There are also people who have such an expensive MC cartridge that playing a record may cost them more than the record itself. Playing a record means that a stylus wears. Since you can't simply replace the stylus of an MC pickup, you send it for re-tipping which usually costs around 80% of the price of a new cartridge. So if you use a $10K pickup and assume that it can play 2'000 records, playing just one costs you $5. So people with very expensive cartridges often have a second tonearm with a decent MM pickup just to check out a record they bought at some flea market. These guys want an inexpensive input on an expensive phono stage.
Do the two MC inputs differ?
No, they're identical. We can't really offer two different inputs because it would mean two different input stages and two sets of SUT. It would be very expensive to implement and probably not really be used. It'd be easier to just buy two phono stages. In terms of features, there is also the question of output/s. Some phono stages are balanced throughout but there are only a handful. They are very expensive and difficult to make. We do not belong to those rare manufacturers but in this one regard rather follow the other 99%. Right after the signal enters our device, the circuit is single-ended. Because there are customers with fully balanced systems, we can offer a balanced output. The rest of the design remains single-ended, however.
So the signal is symmetrized right before the outputs?
Yes. It is not difficult to build a symmetrical circuit but almost impossible to build a balanced RIAA equalization circuit. There is so much gain between input and output that building four independent RIAA channels then summing them expecting flat or zero deviation is practically unachievable. Even if you consider a just 1% or even 0.1% tolerance between resistors, it is already too much for the required precision. Additionally solder points may start to drift after two years and shift the value by a few ohms which will be hugely amplified in this circuit. It is easier to imagine such a solution using solid-state components because it would be easier to control all their electrical parameters. With tubes it seems very difficult. I am not sure if anybody actually does it.
We worked on the RIAA equalization for a long time because it determines not just the sound but precise channel balance. It's not just about tonality, frequency range, detail and so on but soundstaging and imaging. The RIAA reconstruction in both channels ultimately decides the final sonic effect. You need not only precise parts but parts paired with the utmost precision to mandate a big selection pool. Even the quality/precision of your measuring gear contributes to the final effect. Earlier we spoke about the size and weight of our machines. Let's give you an example. At some junction in the circuit we use signal-path capacitors. From a purely technical standpoint these could be very small, say the size of three grains of salt. If you decide to use somewhat higher quality, size increases to a nickel. Even better ones will be the size of a dime. Then we compare those to capacitors in our most expensive DAC and the sound improves. But these are as big as soda cans. What to do about it? We need eight of them and there isn't enough space in our chassis. Now we need new PCB because we can't just squeeze these caps using force. We need proper spacing. So we do all that, put these capacitors onto the newly designed PCB and where before there was no hum, now there is some. Now we start all over again to design and make a new PCB. To fit these huge caps we had to eliminate some other less critical component(s). Even though it is so much work, time consuming and money draining, we still do it because once we hear the sound we truly like, we don't want to go back to less. That's provided we manage to keep a proper signal-to-noise ratio because the bigger parts grow, the more interference-prone they get. There's a limit and you have to watch for S/NR so you produce better sound without more noise. One of the capacitors we use in the signal path is custom-made. It has low capacitance but must withstand high voltage and remain small.
Would you like to download the current image?
Would you like to download the current image?