Q: What is I/V conversion? What role does it play?A: I/V (or more accurately, I/U) conversion stage is the process of converting signal current into voltage signals, situated between the DAC current output and the first gain (amplification) stage. The I/U (current to voltage conversion stage) is a key element in the overall sound quality of the DAC, as important as the choice of the DAC chip itself. To put it metaphorically, if the DAC is the engine, then the I/U conversion stage is like the transmission of a car.All R2R (also known as resistor-to-resistor ladder) type DACs convert digital data into a continuous analog signal in the form of current. Therefore, the output of these DACs is a weak current, essentially high impedance. However, we need a relatively large voltage signal. This means that the current must be converted to voltage to transform the signal into voltage form. In most cases, the resulting voltage signal is still not large enough and needs to be amplified by a voltage amplifier (such as a vacuum tube or transistor) to drive the subsequent gain stage (amplifier). The conversion process itself can be done passively (through a shunt resistor) or actively (through a transistor). If someone tells you that the current output of an R2R DAC does not require resistors or transistors, they either do not understand at all or are misleading you.Q: Audiophiles prefer vacuum tubes over transistors. You use vacuum tubes in the output stage, but why do you use transistors (active) in I/V conversion instead of the commonly used resistors (passive I/V)?A: The reasons are as follows:1. Active I/U is a more precise method of converting current to voltage. If designed correctly, it provides higher resolution and greater dynamic range than passive I/U conversion. In this regard, it can compete with any modern DAC chip, which is based on voltage output and relies on complex transistor circuits inside operational amplifiers (please read point 3), where operational amplifiers integrate modern Delta Sigma type DAC chips (yes, almost all DACs that support DSD format use this circuitry).2. All DACs, including the finest R2R DACs (like TDA1541), have transistors built-in. Otherwise, how could all the logical conversion processes that convert digital encoded data streams into pure analog data be controlled? Transistors, like vacuum tubes, act as gate switches, controlling current flow by allowing or blocking it. Since these DACs have a large number of transistors internally, how could a pair of additional, selected high-quality transistors be detrimental? They do not actually amplify voltage but perform a crucial conversion task by matching the impedance between the DAC output and the gate of the vacuum tube. Resistors cannot match impedance at all because they are passive components. Their role is merely to impede current flow to some extent, ultimately irreversibly eliminating typical dynamic characteristics.3. Not all active I/U stages are designed the same. Unlike the vast majority of products on the market, we use a custom-designed discrete I/U converter based on a selected transistor and powered by our custom discrete power supply. In contrast, most products’ active I/U conversion is implemented through integrated circuit (IC) chips (also known as operational amplifiers). If these chips are not integrated into the DAC chip (almost all Delta Sigma DACs integrate operational amplifiers), then these chips are usually small, with only 8 pins, but very complex, using up to 30-40 low-quality silicon transistors and resistors to amplify the DAC output voltage. Since they are essentially nonlinear voltage amplifiers, they require closed-loop feedback designs, which necessitate applying a lot of negative feedback (this can correct nonlinearity but also degrade sound quality). Therefore, using Delta Sigma DAC voltage output means using built-in operational amplifiers. Otherwise, how could the DAC chip output 2 volts RMS voltage?Q: As far as I know, almost all high-end DACs use passive I/V conversion. Since passive (shunt resistor) I/V conversion can be used, why bother with active I/V conversion?A: Of course, using resistors instead of transistors is also possible. The resulting sound quality is decent, but from a sound quality and technical perspective, it is not ideal because there is impedance mismatch between the active amplification stages, especially when connecting to vacuum tubes after I/V conversion (transistor gain stages have fewer issues, but there are still other problems). Vacuum tubes are inherently high input impedance devices, while the output current of R2R DACs is very small and weak (essentially high output impedance, ideally hoping for a lower input impedance in the subsequent amplification stage). Therefore, the high output impedance of R2R DACs and the high input impedance of vacuum tubes lead to impedance mismatch.Q: Passive I/V is simpler to implement, sounds good, and is a relatively accurate current-to-voltage conversion method. Why do you choose active I/U conversion?A: Passive I/U conversion technology based on shunt resistors may seem to have a lower error rate since it does not involve active components, but due to impedance mismatch, its accuracy is greatly reduced. The quality of these shunt resistors is crucial to sound. Choosing the wrong resistor can lead to a thin or distorted sound. The type and material of the resistors also present the same issues, but overall, it has a much higher tolerance than the active I/U conversion method we employ.Correct implementation of active I/U conversion is more challenging, but if optimized, the rewards are substantial. Once the active I/U implementation is optimized, it can yield clearer sound quality, higher resolution, and greater dynamic range. However, due to the involvement of active components, active I/U is more prone to errors than passive I/U. Active I/U has a lower tolerance but higher conversion accuracy because the active I/U part relies on the low output impedance of those tiny DAC output currents. Due to the presence of active components, it is more susceptible to errors, and these components are the source of power supply voltage fluctuations (also known as noise)—thus, the quietness and low output impedance of the power supply are crucial. For this reason, we use a custom discrete parallel voltage regulator based on powerful germanium transistors in all low-voltage power supplies. Of course, the quality of materials and components can also adversely affect resistor selection in passive I/U. Therefore, the choice of power supply and transistors is a key factor in achieving precise I/U conversion—it determines the success or failure of the sound. From this perspective, it is very similar to excitation-type speakers.Q: What is the difference in sound between germanium transistors and silicon transistors used in the I/U stage?A: Compared to the silicon-based transistors we currently use, germanium transistors have a softer, warmer sound and a more tube-like quality. The clarity and ability to convey low-level musical information of germanium transistors are also an order of magnitude higher than that of silicon transistors, partly because germanium transistors have a much lower breakdown voltage than silicon transistors. Both types of transistors sound excellent, but their characteristics differ. If we compare these transistors to humans, silicon transistors are like a masculine man, while germanium transistors are like a gentle, warm, yet expressive woman.Q: So, from a cost-performance perspective, using transistors for active I/U conversion is the best solution, but using high-quality I/U interface transformers, which are more expensive and well-designed, can achieve cleaner sound than any active I/U conversion stage. Is my understanding correct?A: This is not entirely correct. We are discussing I/U silver wire transformers that cost thousands of dollars and are of the highest quality, paired with a pair of excellent tantalum resistors, to match our method. While it is possible to use interface transformers costing as little as a few hundred dollars, most of them sound rough and grainy, failing to convey low-level musical information. This is also one reason why some people revert to the simple shunt resistor method. Regardless of the solution (low-cost I/U transformers or resistors), the sound is generally inferior to our selected transistors (transistors from the 60s, 70s, or 80s), and the dynamic sound of I/U transformers is not better than our solution.