I've got great news regarding my Pixie2! I have now reached 330 mW at 7.04 MHz and a little more than 400 mW at 3.579 MHz! First I'd like to say that this is an excellent transceiver project. Very simple and a tremendous learning experience for those who will go all the way and insist on extracting the last bit of performance out of the rig. What I did: Swapped out the Q2 Final 2n3904 transistor with a 2n2222A. It immediately went from 2.X volts on the detector to 5 Volts. I was so excited I immediately called back a CQ-er and got a QRZ? de .... I called 2 more times and got a WE6? de .... Never QSO'd with the Pixie2 but did get a SRI QRN too Weak etc... Hooked up the HW-8 and after biding my time while he qso'd a third party, finally connected and had an excellent rag chew. He was in Tacoma Washington. My QTH Santa Rosa, CA. I think that's about 900 miles north of me. Don't remember the Call so no QRZ database check on the miles yet. I also added a .01 uF cap from the + supply line to ensure adequate RF bypass. Currently the XTAL provides this function but I decided to go the extra mile. Also changed the PI output netowork to 390pf-1uH-560 uF per w1fb mods. (Didn't change the 1uH to 1.5uH yet.) This added another 15 mW or so to my 100 mW of yesterday. The transistor change was afterward. Still Way too much BC coming in so will attempt to solve this problem someday. The BC station is about 4 miles away and you can pick them up with rig off (Almost). On Diode Detector Power Measurement conversions: Unless you've got an exceptionally accurate power meter to calibrate you xtal detector, I've determined that significant non-linearity exists as to make the detector useless. I have computed a 'K' factor for my detector where Voltage * 'K'=Vrms Power=(Vrms)^2/50 The 'k' factor ranges from 2.147 at 100 uW to .812 at 330 mW. A .85 'k' factor gives less than 10 % error at power levels between 10 mW and 300 mW. To compute your detector 'K' factor: K=sqrt(Pwr*50 Ohms)/Volts ; where Pwr is Lab-Quality measured. NOTE:No need to do a Peak*.707 for rms, 'K' factor only needed. I'm putting this note on my pixie2 web page section for reference. Happy QRPp operation to all! -Ed Loranger, mailto:we6w@qsl.net (Unlimited Distribution of unedited document allowed)