But not simultaneously. 😊
That would be 80W for SOT223 !!!!
Maybe cascode, like the LU1014 triode cell ?
But then maybe just use LU1014 ?
Patrick
@6L6 suggested doubling up the transistors and cutting the power per transistor in half. Each transistor moves further away from the SOA cliff-of-death.
I think it's a pretty good idea! Wayne's source degeneration resistors ballast the MOSFETs; they soak up any deltaV which may be present due to (Q1a, Q1b) not matching perfectly. Same for (Q2a, Q2b). Bravo @6L6 !
Parts count increases by four.
Q1 with source resistor R3 ----> becomes ----> (Q1a with source resistor R3a) parallel (Q1b with source resistor R3b)
Q2 with source resistor R5 ----> becomes ----> (Q2a with source resistor R5a) parallel (Q2b with source resistor R5b)
And naturally R3a = R3b = (2 * R3) . Also R5a = R5b = (2 * R5).
I think it's a pretty good idea! Wayne's source degeneration resistors ballast the MOSFETs; they soak up any deltaV which may be present due to (Q1a, Q1b) not matching perfectly. Same for (Q2a, Q2b). Bravo @6L6 !
Parts count increases by four.
well, deducing source resistor value for 90mA/2 is sorta trivial
for someone having BSP129 at hands
for someone having BSP129 at hands
speaking of nature of the beast(s)........ what I got to as practical recipe, using 2* triple of 2SK2145BL for same sex buffer, I found that white follower is worse (in sound and in measurements) than plain non-modulated buffer***
White follower arrangement is better "only" in condition of really heavy load..... for which I even didn't think of using puny JFets buffer
just a reminder that different parts are different with reason, or from reason
***which can be sorta good, lower portion (CCS) can be made with whatever best for occasion, no need to use same 3 JFet cell as above
White follower arrangement is better "only" in condition of really heavy load..... for which I even didn't think of using puny JFets buffer
just a reminder that different parts are different with reason, or from reason
***which can be sorta good, lower portion (CCS) can be made with whatever best for occasion, no need to use same 3 JFet cell as above
I know, but everything about AC loop. No?
Without servo, cap doesn’t matter for DC either, of course. But FB might change depending on non-servo DC bias arrangement.
Without servo, cap doesn’t matter for DC either, of course. But FB might change depending on non-servo DC bias arrangement.
Last edited:
my set of criteria was different; with said JFets in said circuit, performance of (optimized) White follower is too much load dependent for my liking
same phenomena why some ppl are liking tube Mu stages, some don't ..... you can optimize them as part of (bigger) circuit but if used as stand-alone, another story
same phenomena why some ppl are liking tube Mu stages, some don't ..... you can optimize them as part of (bigger) circuit but if used as stand-alone, another story
clarify .....
as it is, same servo is going to be fully functional, with or without that cap; buffer being White follower or plain follower
(Maybe too obvious to everyone else) that it’s easy to add switch to select between white or plain.
Maybe servo part of my comment is irrelevant as well as confusing.
Maybe servo part of my comment is irrelevant as well as confusing.
