i am working and listening with P3A for more than 3 years now ...very very happy with it
constructed many versions many pcb and my favourite remains a version equiped with 2SA1302 - 2SC3281 + BD139 BD140 biassed at 120ma ( from various i tested this what i liked most ) 2x10.000 mfd per board at 40+40 volt dual mono
now days i constructed one version with fairchild 1943-5200 biased at 50ma but seemed to me that the sound was much more worst I wonder if it was the bias or the semis
After that one idea crossed my mind to make one that is biassed at 200 or 250 ma presuming that has big enough heatsinks and may be bit more capacitance but for sure rails something like 35+35 volts
question:
1 do you think that this will worth to try ???
2 then again my Vbe in attached to one of the drivers in all my previous versions to you think for a 200 -250ma bias should be attached to the heatsink ???
thanks sakis
constructed many versions many pcb and my favourite remains a version equiped with 2SA1302 - 2SC3281 + BD139 BD140 biassed at 120ma ( from various i tested this what i liked most ) 2x10.000 mfd per board at 40+40 volt dual mono
now days i constructed one version with fairchild 1943-5200 biased at 50ma but seemed to me that the sound was much more worst I wonder if it was the bias or the semis
After that one idea crossed my mind to make one that is biassed at 200 or 250 ma presuming that has big enough heatsinks and may be bit more capacitance but for sure rails something like 35+35 volts
question:
1 do you think that this will worth to try ???
2 then again my Vbe in attached to one of the drivers in all my previous versions to you think for a 200 -250ma bias should be attached to the heatsink ???
thanks sakis
I have experimented a bit with Vbe multipliers and bias.
I found I got best results with the bias transistor glued to an output transistor.
If the bias transistor is put on the heatsink you will get thermal lag and if the output transistor gets hot very qucik the reaction time will be too slow.
On the other hand if I run my amp with minimum bias I can get away with the bias transistor on the pcb. We are probably talking 10mA bias here just sufficient to lose cross over distortion.
I found I got best results with the bias transistor glued to an output transistor.
If the bias transistor is put on the heatsink you will get thermal lag and if the output transistor gets hot very qucik the reaction time will be too slow.
On the other hand if I run my amp with minimum bias I can get away with the bias transistor on the pcb. We are probably talking 10mA bias here just sufficient to lose cross over distortion.
The P3 amps are CFP. You don't want the bias transistor mounted to the heatsink or an output device like with EF output. It should be mounted in contact with a driver transistor on the board.
For high bias I'd probably drop the rails to +/-30Vdc. Or just build the class A version.
For high bias I'd probably drop the rails to +/-30Vdc. Or just build the class A version.
Last edited:
on the other hand i may as well construct something that is designed to be class A from starts ...
never done that before ....
never done that before ....
Last edited:
- Status
- Not open for further replies.