If that's true, then why are they recommending the 12VAC 200mA supply. That's the one I got and it lasted all of 8 days
Unfortunately in those few instances when "Voldemort" made a mistake he tended to go into defensive mode rather than fix the problem it seemed.
That WAU12-200 transformer choice was something else I tried to talk him out of at one point last year after he first posted the O2 design. One problem with current capability of the transformer secondary is shown in the simulation below. This is done for a 16VAC transformer but the problem is the same. A half wave rectifier, like the two in the O2, charge up the filter capacitors with a very short high(er) current spike after the rectifier turns on each cycle, as opposed to a lower current spike with full wave or bridge rectifiers. That is a design trade-off in using a half wave rectifier. Note in that LT Spice simulation a 200mA (rms) continuous load produces a 800mA charging spike in each 470uF filter capacitor, and hence through the transformer secondary.
The chips in the O2 pull about 22mA just as they sit (quiescent current consumed by the chips) and the batteries pull about 22mA of charging current during the maximum part of the charge curve if they are completely dead (7Vdc each where the PM circuit cuts off) so a headphone load of more than only 3mA or so would produce a charging spike of more than 200mA in the transformer.
If the transformer secondary is rated less than the amount of the charging current spike the core will saturate. That can cause heating and also some noise generation. In the case of heating, the Triad wall adapters may contain thermal fuses (can't find any reference to that, but pretty good guess to be UL certified) which may eventually open up if the transformer gets hot enough. My suggestion to "Voldemort" was to specify the 1 Amp WAU16-1000s (Mouser #553-WAU16-1000), for the case when the O2 was fully loaded with 200mA per channel, or the 400mA WAU16-400 (Mouser #553-WAU16-400) for more normal loads that would probably be the case 95% of the time. He eventually added those two under the "optional" column on his BOM. The 400mA transformer would increase the maximum permissible headphone current load before transformer core saturation from 3mA to 56mA.
To summarize, if your headphone load is less than only about 3mA or so, or up to around 25mA if you are not using batteries (so are not charging anything) you would probably be OK with the 200mA transformer secondary rating. The 12Vac transformer secondary voltage rating is another issue. That is 12Vac if your line voltage is at least 110VAC or so. If your line voltage sags less than that the transformer output will drop lower and the voltage regulators won't have the 3.5Vdc or so across them they like for maximum ripple and noise rejection from the datasheet. At one point some folks had found "12Vac" transformers made for LED lighting, in reality outputting 11Vac, causing the O2 regulators to actually go into dropout. 14 or 16Vac would provide more headroom against line voltage fluctuations.
If the incoming line voltage from the wall socket is on the high(er) end, 120Vac - 130Vac in the USA, and is known to be fairly stable over the course of the day, then the "12Vac" transformer will be producing closer to 13Vac - 13.5Vac loaded (14.5Vac+ unloaded) and would be OK from a voltage standpoint.
For more tech details, see the second picture down on the left on this rectifier application note for Hammond transformers, the half wave capacitive input situation [can't post the link, just do a Google search on "Hammond rectifier" and it is the first result]. Hammond recommends de-rating the rms secondary current by a factor of 0.28 in the case of half wave rectifiers. So in the case of the 200mA WAU12-200 transformer, that secondary is only good for (200mA)(0.28) = 56mA (rms) continuous, in the case of capacitor input half wave rectification, to prevent transformer core saturation.
EDIT 1 11/22/2012: In fairness to "Voldemort", since he seems to be AWOL these days for some reason, his defense of the first issue a year ago was that the amount of excess heating and/or noise generation in the transformer from small amounts of transformer core saturation would be within specifications. His defense of the second issue was that if low line voltage results in smaller voltage drops across the regulators, or even if they drop out of regulation completely, the power supply rejection ratio of the amplifier op amps is good enough to not matter even if voltage ripple winds up on the power supply rails. True to some extent (many tech details involved), but my view remains why create those marginal operating conditions in the first place when for $1-$2 more USD, the way transformers are priced, you can buy a 14Vac - 16Vac 400mA - 800mA unit and avoid the issues entirely.
EDIT 2 11/25/2012. Added some more information and cleaned up some grammar. Another issue that probably figured into "Voldemort" leaving that WAU12-200 transformer as the primary BOM specification was cost. I can say from some email exchange a year ago that for some reason he was very adamant about keeping the basic parts cost at $30. He considered raising the parts cost even a couple of dollars a big deal. I think his view was along the lines of keeping the O2 very affordable for DIYers.
[ATTACHMENT=267]Capture.PNG (31k. PNG file)[/ATTACHMENT]