do it, doit, do it....
Marc
+/-50Vdc rail????
I have allready 2 matches mjl3281/1302 pair in stock and 500va rcore 2x36V in case of...
Last edited:
Hi sbrads
Allow me first ...
Exactly, 200 mA is an absolute max forward current for 1N4148 diode. Incautious or unaware DIY-er should use serial resistor and trimmer as safer solution. Even better to put CCS-s which will rise AC performance of the input pair to much higher level.
Hi sbrads
Allow me first ...
Exactly, 200 mA is an absolute max forward current for 1N4148 diode. Incautious or unaware DIY-er should use serial resistor and trimmer as safer solution. Even better to put CCS-s which will rise AC performance of the input pair to much higher level.
Meaning much higher gain from FE to throw away?
Member
Joined 2009
Paid Member
Who wants to minimize even order distortion
Exactly - the power resistor should be higher in value than what you would calculate for the current you want to achieve, then use a pot that is even higher value so that it draws less current than the power resistor and shunts the power resistor to increase the overall current - adjustably.
If you want to idiot proof it (just in case you're building it for somebody else
) instead of a putting a pot in parallel with the power resistor, put an extra resistor in series with the pot, then put the Pot + series resistor combination in parallel with the power resistor. This extra resistor protects you in the case that the pot is turned down to zero I could easily do TSSA BIGBT HP since it is much simpler than SSA and potentially even better performer.
yeah ! (I just like the elegance of 'simplest')
hehe almost
Problem is to find and get some good p-jfet, any suggestions
don't do it, CCS's are evil, just say 'no'.... just kidding, although I've never used one (CCS or JFET for that matter)
Last edited:
Hi Bigun
Over the weekend I measured few of them, not for the first time, I did it before also and was never realy satisfied. BJT as cascode in CCS had ten times higher Z100 kHz, noise and much more nonlinear impedance curve over audio band. But the one I posted with jfet cascode impressed me deeply. For rock solid DC current there's precision zener, for perfect AC performance there's jfet.
Jfet itself to form simple CCS is way too much temperature dependent (DC stability is bad) but here in cascode position it feels just perfect, made for this.
Attachments
Hi guys
TSSA all BJT version needs to be thermally compensated, for that reason 1N4148 were replaced by BD139/140 mounted on the heatsink together with outputs, also input BJT pair flats are glued together. In that way amp should be thermally stable even at high output bias current.
TSSA all BJT version needs to be thermally compensated, for that reason 1N4148 were replaced by BD139/140 mounted on the heatsink together with outputs, also input BJT pair flats are glued together. In that way amp should be thermally stable even at high output bias current.
Attachments
Or even less parts solution with thermal track output transistors, where tracking diodes have double function: bias input pair and compensates thermal drift.
LC is it worth to work on a pcb for this, i think i have all parts in stock
Marc