I think the reverse bias transistor reference in mentioned in Analog Devices app note on Alexander Current mode audio p[ower amplifier design
I know that Jung uses it in "Don't Shunt the Shunt Regulator". But it's outdated and uses the ancient 2N3638. Does it make sense that lower noise transistors will make lower noise references when reverse biased like this?
I think I remember there being another app note or something by Jung where he does this again with a more modern transistor.
- keantoken
I think I remember there being another app note or something by Jung where he does this again with a more modern transistor.
- keantoken
I spent a day experimenting my Marantz SACD player with the K-Multiplier.
I have to say that I like the resistor (47R) much better than the jFET (2SK170BL with Idss=11.2mA) at R1 position. I have no technical explanation. To me the resistor version sounds correct and more musical, while with jFET it sounds a bit harsh. I have 47R for R1 now. I have not compared different values of resistance, but only compared the resistance to jFET. The result was so obvious that repeatitions were not needed.
Bad jFET?
I have to say that I like the resistor (47R) much better than the jFET (2SK170BL with Idss=11.2mA) at R1 position. I have no technical explanation. To me the resistor version sounds correct and more musical, while with jFET it sounds a bit harsh. I have 47R for R1 now. I have not compared different values of resistance, but only compared the resistance to jFET. The result was so obvious that repeatitions were not needed.
Bad jFET?
Last edited:
Keantoken:
in your post at 146 http://www.diyaudio.com/forums/powe...tokens-cfp-cap-multiplier-15.html#post2376286
will you spec out the two capcitors in the FET version of the shunt. Also are the resistors all quarter watt?
Thanks
in your post at 146 http://www.diyaudio.com/forums/powe...tokens-cfp-cap-multiplier-15.html#post2376286
will you spec out the two capcitors in the FET version of the shunt. Also are the resistors all quarter watt?
Thanks
1/4W resistors will work for all resistors.
The capacitors I used for the 1000uF cap are Panasonic FC (25V), but I don't know how much it matters. The output decoupling can be film, I tested with polypropylene snubbers and 6.8uF polycarbonates here.
- keantoken
The capacitors I used for the 1000uF cap are Panasonic FC (25V), but I don't know how much it matters. The output decoupling can be film, I tested with polypropylene snubbers and 6.8uF polycarbonates here.
- keantoken
Last edited:
I think Salas would be the one to ask, but in Zout terms, both will work as long as you make the current through it 15-17mA.
- keantoken
- keantoken
That J110 shows low on resistance which is good. Is it's 10mA Idss spec enough for biasing the slave transistor correctly in your circuit?
To get the low output impedance specs I gave, you need 15-17mA through the driver transistor. Output impedance is directly related to the quiescent of the driver (bigger is better). Try not to go above 20mA, because this is the onset of beta droop for the BC550/560. So say the outputs have an Hfe of 100 and your load draws max 300mA. Well give some headroom by biasing the driver 3mA lower, at 17mA.
Zout is impacted dramatically by the driver bias current. IE, for 1mA it's 800mR and for 10mA it's 80mR, or you get the idea.
- keantoken
Zout is impacted dramatically by the driver bias current. IE, for 1mA it's 800mR and for 10mA it's 80mR, or you get the idea.
- keantoken
So we look for something available in the house of mouse with circa 15mA Idss.
P.S. The driver transistor is called also master in a Darlington, yes?
P.S. The driver transistor is called also master in a Darlington, yes?
What's all this about the house of the mouse? Everyone's talking about it now... Party? Where's my invitation?
Pretty much.
- keantoken
Pretty much.
- keantoken
Slang for Mouser among DIYers. So I just found J109 is available at Mickey's. It has 40mA IDSS @ 15V VDS. He can get that and degenerate with a source to gate resistor. Taking current from the gate node. How much VDS will it see in your CCT?
Abort!🙂 Revert to a resistor. J111 is -3 to -10V Vgs(off). Bill in Australia could put a K170BL 11mA (just about for adequate pinch off) but still did not like it if you remember. That ties with your explanation that just 10-11mA to 17mA recommended impacts the Zo of the complementary Darlington negatively, since each mA in the driver counts in lowering Zo enough.