Was your determination of NFB level more for stability or a targeted THD number?
Also, have you published measurements of the P FE by itself? I bet that could be a killer stand-alone pre for those with CD/Buffer amps.
Tl:dr will it drive an F4??
Also, have you published measurements of the P FE by itself? I bet that could be a killer stand-alone pre for those with CD/Buffer amps.
Tl:dr will it drive an F4??
there is a Beauty Factor Resistor; I believe I did explain its role somewhere in Babelfish XA252 thread
here set for maxumum OLG, thus NFB being maximised too
didn't exactly shoot for anything specific regarding OLG/NFB, more gone with the flow, varying few things till I was satisfied with outcome and possible variations of mentioned resistor
of course, dominant 2nd always being condition sine qua non
didn't filed measurements of FE itself, was lazy , implementing it right away in amp environment
there is possibility to make full preamp package, once when I'm done with implementation of Babelfish XA252 US, where full Ultrasymmetry FE is called for
that being one of next fun stints, I hope..... being already postponed for some time
I started, then it became clear that resulting OLG ( full blown balanced) needs taming and few tweaks
as always, either All The Glory, or Major Disgrace
damn faceooklikingandpopularity functions ........ 
what's wrong with simple old chatting form?
asking questions and getting replies
mocking included, of course .... I can't do without
what's wrong with simple old chatting form?
asking questions and getting replies
mocking included, of course .... I can't do without
Dear Mighty ZM,
Do you foresee any problem (aside from perhaps some level shifter and BFR adjustments) with using a Semisouth R085 depletion part in place of the Tokin?
Thanks,
Dennis
Do you foresee any problem (aside from perhaps some level shifter and BFR adjustments) with using a Semisouth R085 depletion part in place of the Tokin?
Thanks,
Dennis
no problems at all
BFR doesn't need to be touched for that change, sole reason to that is possible accommodation for user
even value of level shifting resistor isn't that much an issue, as it is with passive setting of DC Offset ( as it is in M2 derivatives), where we like to see similar values of resistors from rails to gates, to achieve best DC Offset behavior overall
here FE is acting as servo, so easy peasy
now, thinking of levels ....... and various parts not exactly being in Ugs range of regular IRFP mosfets; keeping Level shifting Source follower in place, we can think of any depletion mode part as final in positive rail, and IRF510 as little one
LuDEF P , logical example ( most likely next one in entire P installment)
SJDP Semisouth family
them we can think of enhanced mode , with Ugs lower than common (IRFP) 4V - so Semisouth SJEP family; possible use of either Toshiba TO220 mosfet ( not easy and expensive) or DN2540 ..... wanting to have resulting Ugs of Level Shifetr mosfet at similar absolute value as lower output mosfet ..... well, some symmetry is still wanted there
I don't have any SJDP , so can't try it and do all mumbojumbo piccies ...... but I'm more than sure it'll work flawlessly
and it'll even have that funny current distribution -SIT SJDP fighting with mosfet, while speaker getting their current difference, not sum
BFR doesn't need to be touched for that change, sole reason to that is possible accommodation for user
even value of level shifting resistor isn't that much an issue, as it is with passive setting of DC Offset ( as it is in M2 derivatives), where we like to see similar values of resistors from rails to gates, to achieve best DC Offset behavior overall
here FE is acting as servo, so easy peasy
now, thinking of levels ....... and various parts not exactly being in Ugs range of regular IRFP mosfets; keeping Level shifting Source follower in place, we can think of any depletion mode part as final in positive rail, and IRF510 as little one
LuDEF P , logical example ( most likely next one in entire P installment)
SJDP Semisouth family
them we can think of enhanced mode , with Ugs lower than common (IRFP) 4V - so Semisouth SJEP family; possible use of either Toshiba TO220 mosfet ( not easy and expensive) or DN2540 ..... wanting to have resulting Ugs of Level Shifetr mosfet at similar absolute value as lower output mosfet ..... well, some symmetry is still wanted there
I don't have any SJDP , so can't try it and do all mumbojumbo piccies ...... but I'm more than sure it'll work flawlessly
and it'll even have that funny current distribution -
Last edited:
It could be fun to do a modular approach (not as though we have enough🙂 - stand-alone FE card and swappable OS cards.
😀 I know how much you like legos…
😀 I know how much you like legos…
@codyt
well, you're not first to think of that, when we are speaking about my concoctions
I mean - besides me, few other Greedy Boyz did observation that , with care and methodical work, one can make all of my M2 derivatives with just one (properly and adequately made) pcb
but, however idea is tempting, after some daydreaming time invested, I realized it's exact recipe for disaster ...... no, make it - DISASTER!
just imagine scenario of regular Greedy Boy being utterly confused with number of jumpers, parts not needing population for actual iteration of amp etc. ....... and each one pulling me to help with dilemma ( in best case) or help to put gray smoke back (much realistic scenario)
having dedicated "One for All" FE Daughter pcb and various iterations of OS Motherboard pcbs is just milder version of DISASTER scenario, keeping on mind that - if I'm going to try to attend my duty of helping, I need to have at least illusion of some context and continuity and, somehow, that impression/illusion is firmer when I'm thinking about separate construction, not just another combo of piled up pcbs
applicable pretty much in same amount to Plethora of Pinjatas family as it was for M2 derivatives family
having that all analyzed, there is still one more relaxing thought for me (pretty much exact words of someone smarter than Mighty ZM) - mine is to make them, not to choose between them ......
anyway - Greedy Boyz likes to build things; many of them ( I clearly remember my naive and innocent time of it, doing some other important life things while audio electronics being "just a hobby") are having just enough energy to play with it few hours a week here and there, wanting and finding relaxation and joy .... while they're simply not having neither time (right now) nor energy to invest in (too much) thinking and learning
example - actual Aleph J Store pcbs - certainly and without doubt - made with best intentions, are resulting in just too much grief and sweat in too many cases, simply because majority of builders are confused with too many choices left on pcbs
go figure; straightforward is sweet, straightforward is easy ..... especially when activity in case is there to keep bleak side of life furthest possible away
so, that's main reason why ACA is so good; it is not as good as people are happy with it - it is so good because people are not tired while building it and they got in the end soooo much more than they expected
(fact that it's singing so much better than 65% of expensive Drek sold around, that's separate part of the ACA Story)
well, you're not first to think of that, when we are speaking about my concoctions
I mean - besides me, few other Greedy Boyz did observation that , with care and methodical work, one can make all of my M2 derivatives with just one (properly and adequately made) pcb
but, however idea is tempting, after some daydreaming time invested, I realized it's exact recipe for disaster ...... no, make it - DISASTER!
just imagine scenario of regular Greedy Boy being utterly confused with number of jumpers, parts not needing population for actual iteration of amp etc. ....... and each one pulling me to help with dilemma ( in best case) or help to put gray smoke back (much realistic scenario)
having dedicated "One for All" FE Daughter pcb and various iterations of OS Motherboard pcbs is just milder version of DISASTER scenario, keeping on mind that - if I'm going to try to attend my duty of helping, I need to have at least illusion of some context and continuity and, somehow, that impression/illusion is firmer when I'm thinking about separate construction, not just another combo of piled up pcbs
applicable pretty much in same amount to Plethora of Pinjatas family as it was for M2 derivatives family
having that all analyzed, there is still one more relaxing thought for me (pretty much exact words of someone smarter than Mighty ZM) - mine is to make them, not to choose between them ......
anyway - Greedy Boyz likes to build things; many of them ( I clearly remember my naive and innocent time of it, doing some other important life things while audio electronics being "just a hobby") are having just enough energy to play with it few hours a week here and there, wanting and finding relaxation and joy .... while they're simply not having neither time (right now) nor energy to invest in (too much) thinking and learning
example - actual Aleph J Store pcbs - certainly and without doubt - made with best intentions, are resulting in just too much grief and sweat in too many cases, simply because majority of builders are confused with too many choices left on pcbs
go figure; straightforward is sweet, straightforward is easy ..... especially when activity in case is there to keep bleak side of life furthest possible away
so, that's main reason why ACA is so good; it is not as good as people are happy with it - it is so good because people are not tired while building it and they got in the end soooo much more than they expected
(fact that it's singing so much better than 65% of expensive Drek sold around, that's separate part of the ACA Story)
in fact , semi-true
I just found measurement files of PoP FE, for several values of BFR , all the way up to 16Vrms of output (most likely wasn't in the mood to shoot for more)
so, I did filed, just didn't published them
anyway - numbers are ridiculous (low THD)
though, I did learn one thing about (this) FE itself - when decreasing OLG , BFR going up in value, thingie is becoming 3rd dominant on higher output levels
which is logical - symmetry being increased with more local degeneration in LTP .......LTP halves more counteracting each other
Dennis - you won't look back
ZM will La Zucca have AVC option? I do love the Italian name 😊
nope
La Zucca imagined with resistor (relays) binary matrix attenuator
2^6, 64 steps of ~1db
Iron Pumpkin is one and only; if someone needs both exquisite attenuator and different gain stage than it's already given with autoformer, then just use Iron Pumpkin as exquisite attenuator and whatever outbox gain stage is adequate
there is no easy way to physically combine Iron Turtle, necessary buffers , regs for them ( all that being Iron Pumpkin already) and additional gain stage ...... all of that in one box
You sold me a long time ago. 🙂
The issue is always the time and resources to make one worthy of the design.