That stuff is just too funny. So let's start at the beginning, 1k in series with the output is better than nothing (Cavalli, Woo, Kingsound etc.) but it should be around 5K. How 1K is similar to 5K and offer the same protection is beyond me. Does anybody at Hifiman understand Ohm's law?
There are indeed two 2M resistors on that board. When I initially wrote that I had only looked at the amp board and was still trying to untangle the mess of this build. Still that should be 4.7M but 4M is close enough.
Also, while the amp is very reminiscent of the SRA-10S... it does have the opamp input. You know what amp is based off the SRA-10S and has an opamp inputstage? The SRM-001... aka the 1995 portable amp. Yup there are some similarities there as well.
So the output stage uses FQPF4N90C's as can be seen here:
...and the datasheet is here:
http://www.mouser.com/ds/2/149/FQPF4N90C-195517.pdf
Go down into Dynamic characteristics and there we have output capacitance, typical 65pf. So the headphones are roughly 130pf or so and you have two of these per driver. Yup, stellar job of keeping the Cob low. For reference, the 2SC4686A used in the KGSS and KGSSHV, 2.2pf. So the output capacitance here is just about 30 times higher. Same goes for what Stax have used for years. Also, notice the black cans on the source pins, those are inductors for some reason. Inductance and capacitive loads don't really mix, just look at the train wreck that is the Woo WES.
For any laymen reading this, many know about the 10 times rule when matching sources to amps. The goal there is for the input impedance of the amp to be at least 10 times higher than the output impedance of the source. Same goes for preamps into poweramps etc. Same thing applies here, how do you think the amp likes driving the capacitance of the headphones when it first has to overcome its own output Cob?
Bottom line, if anybody like JimL wants to take a look at the Jade2 amp, I'll send it to you for free. The only things I did to it was simplify the input wiring, as it makes no sense to go from the sockets to the amp PCB, then to the volume control, then back again. I also redid the earthing of the amp as it was very lacking for CE safety requirements.
One more thing that amuses me endlessly with Hifiman is that familiar Chinese trait... victim hood. I still remember it well at Canjam 2010 when somebody gave Tyll a set of HE-5's (I think) to measure and Fang freaked out. It was his own personal set so he could do what ever he wanted with it but Fang did not approve. That speaks volumes about his character. I'm all for right being right but what about addressing the flat out lies in the specs, that a few parts were not properly soldered to the boards or that I'm sure the same crap applies to the 50K$ Shangri-La. They certainly use the same cable...