Lazy Singing Bush

[Zen Mod]

same , 4R load graphs

 

[Zen Mod]

then old trusty (even if somewhat on the verge, 50W of heat) IRFP9140 in place of IXYS puck

 

[Zen Mod]

same, 8R load graphs

 

[Zen Mod]

same, 4R load graphs

 

[Zen Mod]

......... How about a standalone frontend to drive it?

pretty low on my list

major obstacle is - deciding how much parts selection can be exclusive?

if Lundahl Iron is allowed, it can be done in short time

if strict Paesano parts choice is allowed, I'm still waiting for crystallization particle

 

[Ben Mah]

I really like my Sony VFET 2SK79 preamp for driving my followers. It has a reasonable amount of second harmonic distortion, similar to the Luminaria, but can output enough voltage for 50W at 8R. For less sugar, the BA-3 preamp also outputs enough voltage and then there is also ra7's latest Schade Common Gate preamp.

 

[Zen Mod]

yup

!

 

[codyt]

pretty low on my list

major obstacle is - deciding how much parts selection can be exclusive?

if Lundahl Iron is allowed, it can be done in short time

if strict Paesano parts choice is allowed, I'm still waiting for crystallization particle

Lundahl is allowed, you may proceed…

 

[Zen Mod]

Now, as I promised - play with 2SK2087C
......
Iq decreased to 2A; did that prior to reading Ben's post, so for my tests I did stay at 1:1 ratio of resistor groups , decreasing them to 0R33 (3*1R in parallel); easy to recalculate to 3:2 ratio

.....

Following Ben Mah's findings, about 3:2 Mu resistors ratio, and counting that same applies to 2SK2087C, be it combined with puck or IRFP, that would be 3*1R2/3W MOX at Mu MOS Source side, and 3*0R82/3W MOX at SIT side

so, 400mR + 273mR

resulting in sum of 673mR, pretty close to desired sum of 666mR

clearer:
R5,7,9 being 1R2/3W MOX
R4,6,8 being 0R82/3W MOX

 

[rickmcinnis]

Is this an evolution of the SINGING BUSH or another flavor?

I see the input buffer is four instead of six FETs and the LM317 is no longer used along with things moving about.

ZM, do you consider this an improvement of SINGING BUSH or just a simplification?

Is it a change worth making?

 

[codyt]

There's no voltage gain, so its a different beast in that aspect

 

[rickmcinnis]

Aha!

Cody, what do you see is the advantage in needing another stage in the chain?

Is this for people who already have powerful line stages?

 

[codyt]

The Singing Bush is a Common source amplifier, giving both voltage and current gain, hence only a buffer up front to help drive the gate of the SIT (largish input capacitance). The Lazy Singing bush is a common drain (follower) amplifier, providing only current gain. The advantage is less distortion, at the cost of no voltage gain. So, you can have your choice of how you want to provide that voltage gain – you could wire up a line transformer/autoformer (like M2, SissySIT, etc), or you could just use a linestage with the necessary cojones. BA3 is always suggested as a perfect example of this. Ben also shared his 2SK79 linestage earlier, which can do the job.

I love my Singing Bush, but am also very excited to try the same parts in a common drain amp (like Lazy SB). The SIT follower I have experience with is the recent Sony lottery amp, and it had a nice sense of clarity/lucidity that I don't necessarily get with my other amps. That said, adjusting the mu resistors in the original Singing Bush can give me something closer to that sound.

 

[andynor]

The Singing Bush is a Common source amplifier, giving both voltage and current gain, hence only a buffer up front to help drive the gate of the SIT (largish input capacitance). The Lazy Singing bush is a common drain (follower) amplifier, providing only current gain. The advantage is less distortion, at the cost of no voltage gain. So, you can have your choice of how you want to provide that voltage gain – you could wire up a line transformer/autoformer (like M2, SissySIT, etc), or you could just use a linestage with the necessary cojones. BA3 is always suggested as a perfect example of this. Ben also shared his 2SK79 linestage earlier, which can do the job.

I love my Singing Bush, but am also very excited to try the same parts in a common drain amp (like Lazy SB). The SIT follower I have experience with is the recent Sony lottery amp, and it had a nice sense of clarity/lucidity that I don't necessarily get with my other amps. That said, adjusting the mu resistors in the original Singing Bush can give me something closer to that sound.

But the question is: is there still room on Codys list to accomodate all of Mightys brain children?

And: wazzup with the birth rate? Schlivo induced or plain mighty+cofee magic?

 

[codyt]

I know! The pace is staggering...

 

[codyt]

Oh, I should also point out that I don't make additions to The ZM List™ until there's an official ZM post, with explanations, etc.

"Link porn or didn't happen" I always say

Now, back to Lazy SB....

 

[rickmcinnis]

Thanks, Cody, for that clear explanation.

I think I will remain with my current level of complication before adding even more!

 

[BRN]

I’m excited about this amp. I have a pair of THF-51S, Randy’s PS boards, a BA-3 pre, and need to finish ra7’s Schade Common Gate preamp.

 

[ra7]

ZM, should there be a resistor from drain to ground for a Schade MOS? It might be there because your measurements for Schaded MOS are different than pure MOS. It is also there on the Warbler schematic.

I’m trying to learn the biasing scheme for the 2sk2087c from your many designs!!! Trying to use an LND150 CCS to develop a stable reference voltage. It needs a gate to ground “grid-leak” type resistor but then it is rock steady.

 

[Zen Mod]

as I said many times (more clever people have better understanding and certainly explanation for that than Mighty Moi) - GND is state of mind

here, GND is "nothing more and else" than positive rail for SIT

we chose (me, looking at mentioned guys) GND as upper rail simply counting on practical fact that GND (whichever way we got to it) is in most cases lesser polluted rail of them all

so, ignore GND when thinking of Schade arrangement; there is resistor from D to G, there is series resistor to G - all you need



now, "biasing"

you need to differentiate several purposes of it*** :

biasing with purpose of setting Iq

biasing with purpose of setting voltage level

biasing in purpose of setting both of these


in this case , entire shebang is having Iq set in Mu Follower itself, ( optothingie)

with biasing input JFet buffer gates, in voltage realm, we are automatically doing same thing for SIT gate - thus setting its Source voltage, which we wanna keep at approx. 1/2 of sum rail value

here, that's 60Vdc-ish, so goal is to get SIT Source somewhere at -30Vdc level

easy peasy



***that's why I'm often insisting that "biasing" term must be either clarified or clearly established in context; if nothing else - just to avoid Mighty Moi being confused ( easy)

example - if you say - "this part is biased to Iq of 3A2", everything is clear ........... but it's even clearer if you say "this part is biased to Iq of 3A2 with Iq setting of Mu Follower, while output node is biased to half rail value with setting of P1", everything ref. to pic above

🙂