So you play around with this stuff long enough, and pretty soon you see stuff on the shelves and in the corners where you say: "..I"ve GOT to do something with that some day!!" Well, that day has come...... at least for some of it. I'm starting a build on an old Motorola Chassis that I've put an 'aluminum deck top' on, and have planned as using an old Channel Master 6600 power transformer, and a pair of Edcor PP 15W 7.6K PI @ 8 ohm output transformers. To be 'tube rectifier, with 12AX7 drivers and EL84 outputs. First, I don't have a schematic for the 6600, so I'm not exactly sure what it's outputs are. I believe the unit used a 5AR4 rectifier.....so I'm thinking it may have a 5V filament current limit at around 2 amps. So, to be on the safe side, I'm sticking with a 5AR4 rectifier. Since I don't know how Channel Master configured it's power supply, I'm going to have to play with that some as well. I believe the 6600 used 4 6V6 output tubes, so I'm only guessing that this will all work out. But, I'm thinking that I should end up with a B+ right around 300 VDC, give or take. My thought is that in using the EL84/6BQ5 output tubes, I should be able to run those into the Edcor output transformers even though they're a 7.6K Primary vs the more desirable 8K. But, because of the PI discrepancy, I'm wondering if anyone has any advice for me on the issue? I'd planned to model the thing somewhat on the same kind of final output section one might see in a Magnavox 9300. In fact, even the front end may look like that to some degree, as the 6EU5s can easibly be replaced with 12AX7. I'd also been looking at an old Fisher Custom Electra power amp schematic..... which also used EL84s, non - UL, and has a 12AX7 front end. The main thing is that the amp has to stay 'cathode biased'.....unless later EFB'd....which will only happen if the results of standard cathode bias prove rewarding enough to merit further effort into the amp. But, just wondering if there might be some 'target' voltages I should shoot for with this power supply section, and if the 7.6K PI should be considered when setting the bias level? Any thoughts or suggestions would be appreciated. I'll post a pic or two shortly!! Many thanks, Tom D.