A few years back I built the AGDR ODA amp. Its a diy amp that is no longer available for DIY. I was and still am a friend of the designer, he was around when “Rocket Scientist” was out there spreading the word on his O2 amp…
My friend actually communicated with him alot and discussed what a desktop ODA would possibly be like. and then he disappeared…but my friend came up with a possible iteration and this is what I use as a standard amp that drives almost anything well (except electro’s)…
I have a post on this build Desktop ODA?
So I am very much aware of the current vs voltage as it relates to headphones etc…
Here is a discussion I had back then?
The Quesion was: ODA maximum output power into various loads with +/-12.5Vdc power supply rails
Good question! 
A lot of folks looking at the ODA would want to know.
The ODA will output a maximum swing of 7.25 volts rms without the attenuation resistors (or with them jumper-shorted), or half that at 3.75Vrms with the attenuation resistors, unless the output current exceeds about 60mA per output chip. There are 6 chips in parallel, so that would be 60ma * 6 = 360mA maximum per channel. I often write 300mA for the ODA just to be conservative. Sort of my reaction to the commercial stuff where everything is over-rated, I prefer to under-rate a little bit. The output chips can actually do 70mA too, according to NwAvGuy, but i knock that down to 60mA to be conservative.
For output current it is necessary to switch back from rms (for power) to peak to make sure the peak currents don’t exceed what the output chips can do. So at 300R, the maximum current with a 7.25V swing would be (7.25Vrms * 1.414)/300R = 34.2 mA(peak) vs. the 360mA, well within bounds. With the atten resistors gives even less current draw, (3.75Vrms * 1.414)/ 300 = 17.7mA. In fact the biggest problem you may run into with some 300R (and 600R) headphones, if they have lower sensitivity, is that even 7.25Vrms isn’t enough voltage swing. That is why the +/-15Vdc option is available, to give about 2.5Vrms more swing for 300R and up phones that need it.
The maximum power output into 300R is then (7.25Vrms)^2 / 300R = 175mW. With the atten resistors it is (3.75Vrms)^2 / 300R = 47mW.
Using the same math on the rest:
Max current with 120R load = 85.4mA(peak), and with the atten resistors = 44.2mA(peak). Both OK vs. the 360mA available.
The maximum power into 120R is 438mW, or 117mW with the atten resistors.
Max current with 64R load = 160.2mA(peak), and with the atten resistors = 82.9mA(peak). Both OK vs. the 360mA available.
The maximum power into 64R is 821mW, or 220mW with the atten resistors.
Max current with 32R load = 320mA(peak), and with the atten resistors = 166mA(peak). Both OK vs. the 360mA available.
The maximum power into 32R is 1.64W, or 439mW with the atten resistors.
Note that things are starting to get interesting with the current at 32R. The maximum current with the 7.25Vrms output at 320 is getting close to that 360mA maximum. We can expect that with a 16R load the ODA won’t be able to go all the way up to 7.25Vrms without distorting. Sure enough, that is what mgalusha’s dScope tests showed. The ODA went up to round 4.4Vrms at 16R load with THD+N levels 36% less than the O2, but then the distortion rose going up to around 6Vrms, which was the maximum before the distortion rose exponentially.
This set of numbers is for the 4.4V4ms output level, and 2.2Vrms with the atten resistors:
Max current with 16R load = 388mA(peak), and with the atten resistors = 194mA(peak).
The maximum power into 16R is 1.21W, or 302mW with the atten resistors (again, this is with 4.4Vrms and 2.2Vrms levels).
Notice the 388mA peak vs. the 360mA “max”. Remember above that I said I was being slightly conservative and using 60mA max for each output op-am vs. the 70mA that NwAvGuy used. Looks like NwAvGuy is right once again, and Mike’s dScope measurement prove it. Using the full 70mA number the maximum output rating of the ODA goes up to 6 * 70mA = 420mA instead of 360mA. So 388mA actually is within the OK range.
And finally, at the 6.0Vrms level (or 3.0Vrms with atten resistors) where Mike measured the maximum output before distortion zooms up, we get:
Max current with 16R load = 530mA(peak), and with the atten resistors = 265mA(peak).
The maximum power into 16R is 2.25W, or 562mW with the atten resistors (again, this is with 6Vrms and 3Vrms levels).
Notice the 530mA here vs. the “new” 420mA max. This says that even NwAvGuy was a bit conservative with the 70mA per op-amp number. The real maximum for the NJM4556AL chips looks like around 530/6 = 88mA or so per chip. The current capability of the chip is taken off a graph in the datasheet with a rounded curve which is why there is some much variance in interpreting it.
This is the amp that worked the best with the Anandas…
Alex