Still casing it up (it works but I still need to do the cosmetic stuff before I can actually sell it). will post picture this weekend.
Received some PM's about it, I thought I should put some info here to help clarify things a little:
*I will guarranty its functionality within 5 years, provided the buyer is within the US and the use is normal (the bias circuitry is very robust, there is nothing to break; the only way it could fail is if you crank the speaker amp up too high and burn the transformer coil, but that clearly will be considered as abused).
I can even change the voltage once for you, how about that? (say, you want to try a 660v bias just for the heck of it, so I will set the highest to 660v before shipping it out. Later you decide leaving that extra high setting on the knob might be risky, then you send it back to me and I will set it back to a more moderate level such as 615v)
*it will work with 220v. But I won't provide warranty for overseas.
*Bias method: an 12v AC (not DC!) wallwart is plugged into the back of the transformer box. There is a little transformer inside that steps it back up to 100~110v. This voltage is then multiplied 7x, then dropped down in steps using a string of zeners. Different voltage can be taken from different location of the zener string.
The reason for using walwart then step up is to improve safety, I don't like unrestricted wall AC line goes directly into an all steel box. The reason for using zener string is to have better accuracy. Zeners string (koss style) don't sound better, it is just more accurate and controlable. If the biasing output is the same, the biasing circuitry has almost no impact on SQ, whether you use zener at the bias input then multiply (stax style), or multiply the voltage up first then use zeners to provide precision drop (Koss style). That said, the Koss style does provide a lot more freedom in voltage selections. In a stax type circuit the output voltage can only be multiples of 82v, which means 246v for low bias and 574v for pro, if the voltage doesn't collapse on the multiplier ladder, it is difficult to get the "voltages in between", such as 600v or 620v.
*effects of different voltage:
-- When driven with a high power but low detail speaker amp (I have a 100w per channel Kenwood KA89, supposedly class-A), there is not much difference between the pro 584 and 614v. The effect of lower voltage is noticiable, 480~530v gives a more melow sound, good for etchy phones such as Lambda pro.
-- When driven with a fine detailed low power amp (such as a maxxed out Tripath amp or a battery powered Gainclone type), the difference brought out by the higher bias is clearly audiable. 615 volt really gives a much more open sound, better transparency, and more bass energy.
*the transformers are direct wired, you will need a good 4 ohm speaker amp to drive it. The sound is better than stock MK2 or Pro due to the direct wiring.
now if you can DIY and don't need the fancy feature like the 4 step selectable bias, I highly recommand you to just go mod a standard SRD7. It really is not difficult, and using the zener string method you are guaranteed to have accurate bias voltages, no guess work involved. If you can DIY, I think you can figure out what to do from the info above.
---
current bias is 230v for fixed, 433v/584v/599v/615v for selectable. will put on lables once the voltage options are finalized.
:
:
The most recent photos: (note the chipped-off area in pro-bias jack and how it is now plugged with a black pin. It is concealed when the phone cord is plugged in, makes practically no difference).
