As I was not able to find it in the internet I traced the schematic of the infamous Joyo Ultimate Drive v2 (2010-11-24 version).
here it is:
For your convenience I prepared two pictures with values and component naming. I cannot warranty the naming is 100% compliant with the PCB silkscreen, as reading under the resistors was in many case impossible, but at least the naming is coherent with the schematics.
And now some comments about the changes between the first and the second version.
My educated guess is that this redesign was mainly intend to avoid possible issues on the intellectual property of the circuit design, by adding a buffer, rearranging the clipper, and changing one (insignificant) resistor value.
The first difference with respect to the 2009 version is the buffer formed by Q1, C1, C2, R1, R2 and D1. It is one of the most simple and common MOS-FET buffer. The input impedance is well above 1.5M in all the audible range. The Zener Diode (I was not able to read the exact model, so I put a plausible device) should be a 9.1V model intended to protect the FET from electrostatics. It is quite interesting to note that many comments on the internet refer to it as a wrong implementation of the millennium bypass but I respectfully disagree. They simply decided to add a (quite transparent and simple) buffer.
The second difference is that the MOS-FETs and the Ge Diode in the clipping stage have been rearranged. I have to point out that I put the Ge Diode as it should be and not as apparently is from factory. When reversed (as from factory) the clipping stage formed by the D2 and Q2 does not work so only one half-wave is clipped (by Q3). If you want it to sound as it was intended to be, check D2 and swap it.
The last (insignificant) difference is that R8 was changed from 10k (value that is used both in the OCD and in the 2009 of the ultimate drive) to 470K. It can be useful to note that due to the circuit design this resistor is not necessary and should have no effect on the sound. On the contrary the higher value could increase the thermal noise of the pedal, but probably in no audible way (I made no test). Anyhow, moving it back to 10K could be an idea...
What is clear is that this the pedal is no more true-bypass as the buffer is in front of the switch. Anyhow the "old" input stage after the 3TDP switch is unchanged. If someone want to turn the pedal back to true-bypass, it should be pretty strateforward to modify the circuit by removing C1 and C2 (the large capacitors placed in between the jacks) and putting a jumper in between the two holes closer to the PCB border. Obviously different implementations are possible, including providing the pedal with two alternative inputs one buffered and one unbuffered, or modify the pedal in such a way that the switching include the buffer too, making it again true-bypass. But that is another story and shall be told another time.
