The SD series Detectors produce the same output power from all models at around 640mA (as Eric said) depending on supply voltage.
The supply voltage is proportionate to the output power produced, hence why there is some benefit from the 12 volt/reduced to 7.45V power supply system currently being marketed around the traps for these units.
The 7.45V power supply isn't recommended for the Extreme. (partly due I beleive, to my thinking that the "Dual Voltage" claims of the extreme are perhaps related to a higher internal Power Supply Voltage to the TX and a separate nominal 5V-6V supply to the processing CCTy.)
The increase in available power using 7.45V in the SD series is nominal at around 15% max (measured with a CRO)
Although the actual improvement in detecting targets is not overly increased in practice from what I have tested.
I think you would need a logarythmic increase in power to get a noticable increase in target response for these units.
In my view, the real key to improving SD's is in modyfying the pulse timing and therefore chasing different shaped targets accordingly. <IMG SRC="/forums/images/wink.gif" BORDER=0 ALT="

">
I feel the only real benefit of an increased supply Voltage would be to maintain a consistant supply over the days detecting, because with the standard 6V system, there is a marginal reduction in performance as the battery Volts drop from a fully charged 6.45 down to around 5.98V.
The majority of current for all SD units would appear to be used in the TX/Coil cct, this is clearly evident if anyone has made a practically tuned probe coil for one of these things... the wiring gets quite warm. <IMG SRC="/forums/images/biggrin.gif" BORDER=0 ALT="

">