Originally Posted by 00940
did you make any noise measurements ? From what I've been told from reliable sources, the Dynahi has a significantly higher noise floor, due to his use of more powerfull transistor and the near-impossibility to get a power supply as clean as for the "dynalo"
My Dynahi RMAA Results
show that the unit has a very acceptable noise floor with a 33 ohm load. The area I am battling with is crosstalk. Since I recorded these tests, I have rewired the amp and psu in many, many ways. The only improvement from the graphs you see, was made when I ran the headphone jack ground wire directly back the PSU ground terminal, instead of the amp terminal block ground. After that, the "resistive type crosstalk" figure measured identical, but not any lower, for each channel. I am using the stock group buy boards and dynahi psu board. I have given exceptional care and some experimentation to all of the ground wiring. I'm not sure better performance is possible without design topology and/or layout changes.
If you read Doug Self and Randy Slone, they talk about crosstalk that is independent of frequency as being "resistive crosstalk". At this stage of the game with my experiements and measurements, it's more of an academic challenge to get the crosstalk lower. Perhaps I'll hack the design to attempt to lower it.