Bimo, are you proposing to steal it from Roender? 😛
Sorry, just joking 😀
Current coupling between the first differential stage and the second one - they actually share the same dynamic load - is cool and very fast
Sorry, just joking 😀
Current coupling between the first differential stage and the second one - they actually share the same dynamic load - is cool and very fast
Bimo, are you proposing to steal it from Roender? 😛
Sorry, just joking 😀
Current coupling between the first differential stage and the second one - they actually share the same dynamic load - is cool and very fast
I want to steal it 😀
I still learn how this thing work. May be I want to simplified this.
New for 2015 ....
And so does this one (below 1) 😀 .280V/us - easy !
Absolutely ideal. Another Russian member applied the current feedback like this -
and I "stole" it .. 😎
Got me 10ppm on my older NAD design.
(below 2) is the slew WITHOUT tweaking.
(below 3) is the PSRR without the cap multipliers.
(below 4) is the THD into a 4R load.
(below 5) is the Bode with the 40mhz "peak" (baxandall)...
I know this , as with a hawksford VAS , it disappears.
The Kypton C has this too , but I'm listening now (it's in negative gain territory) 🙂
Infidel taught me some things , on this one Q1-4 fully cancel thermally .
All is ideal with this 🙂🙂.
It's going to sprint (layout) now !
Kypton C is the "close cousin" , I already
aced the NAD , this is just the refined best of the two.
Truly stunning simulations (rock solid stability) .
Edit - (below 6) ... the best part , rounded "Valve like" overload behavior -
"all hail the super-pair" !!! 😀
OS
And so does this one (below 1) 😀 .280V/us - easy !
Absolutely ideal. Another Russian member applied the current feedback like this -
and I "stole" it .. 😎
Got me 10ppm on my older NAD design.
(below 2) is the slew WITHOUT tweaking.
(below 3) is the PSRR without the cap multipliers.
(below 4) is the THD into a 4R load.
(below 5) is the Bode with the 40mhz "peak" (baxandall)...
I know this , as with a hawksford VAS , it disappears.
The Kypton C has this too , but I'm listening now (it's in negative gain territory) 🙂
Infidel taught me some things , on this one Q1-4 fully cancel thermally .
All is ideal with this 🙂🙂.
It's going to sprint (layout) now !
Kypton C is the "close cousin" , I already
aced the NAD , this is just the refined best of the two.
Truly stunning simulations (rock solid stability) .
Edit - (below 6) ... the best part , rounded "Valve like" overload behavior -
"all hail the super-pair" !!! 😀
OS
Attachments
Last edited:
Bimo, are you proposing to steal it from Roender? 😛
Sorry, just joking 😀
Current coupling between the first differential stage and the second one - they actually share the same dynamic load - is cool and very fast
That's a folded cascode "symasym"(symasui) - I built it. I was a member when Roender designed it.
Member MJL21193 "stole" it for his "abomination" amp ,too !!
OS
Because, if you don't, you do not measure the slew rate capability of the amps, but just the bandwidth limitation of the input filters witch is useless.
So you'll have poor numbers, not reflecting the real possibility of the electronic, witch have an impact on TIM.
OS,
So is this an improved version of the Krypton C or a very distinctly new input section you are just showing in your plots?
So is this an improved version of the Krypton C or a very distinctly new input section you are just showing in your plots?
It's actually old AND new.
First CFA I "tinkered with" (the NAD). And the Kypton C , which I'm listening to.
Apply the old to the new ... get a better one.
I changed it a little , actually simpler (below).
Dropped the second stage cascode - now I have about .15ma VASout thermal
from 10-50C. I know this is close , as my "C" does <1/4ma in real life.
Esperado - this one will max out at about >300V/us right at overshoot. And
yes , I did remove the input filter. With the 150pf input filter , this design will
still do >250V/us.
All this is also very similar to a VSSA type design. What is superior here ,
is the PSSR , THD (more CLG). The servo return is not "working" hard
and does not interfere with any audio path.
Just all around "well behaved" (for a simulation).
PS - cool , I'm down to 12 semi's !
OS
First CFA I "tinkered with" (the NAD). And the Kypton C , which I'm listening to.
Apply the old to the new ... get a better one.
I changed it a little , actually simpler (below).
Dropped the second stage cascode - now I have about .15ma VASout thermal
from 10-50C. I know this is close , as my "C" does <1/4ma in real life.
Esperado - this one will max out at about >300V/us right at overshoot. And
yes , I did remove the input filter. With the 150pf input filter , this design will
still do >250V/us.
All this is also very similar to a VSSA type design. What is superior here ,
is the PSSR , THD (more CLG). The servo return is not "working" hard
and does not interfere with any audio path.
Just all around "well behaved" (for a simulation).
PS - cool , I'm down to 12 semi's !
OS
Attachments
Thanks OS,
Simpler and better is always nice. So you have basically updated the Krypton C with some minor changes.
Now that you have been listening to the Krypton-C for awhile what is your thinking on the sound? I know you are the kind of guy who will for curiosity sake keep trying new things and looking for improvements but what say you from the point you are at now, are you satisfied with the design sound wise?
Simpler and better is always nice. So you have basically updated the Krypton C with some minor changes.
Now that you have been listening to the Krypton-C for awhile what is your thinking on the sound? I know you are the kind of guy who will for curiosity sake keep trying new things and looking for improvements but what say you from the point you are at now, are you satisfied with the design sound wise?
Hi OS. Well designed (enough output attenuation) a servo never interfere with audio path in the audio bandwidth.
On my point of view, there is no difference if you set the return to a css or to the feedback return point. they both will interfere the same way to the signal.
(CSS *are* in the signal path, don't they ?)
The C is so close to perfection , I'll put them in my mini OPS's .... I won't shelve them.
This one's "sound" might be so close you would not be able to tell .
Electrically , this one is more "precision". It performs better thermally, servo
controls the offset "easier".
TOTALLY DC coupled , with no big 470uf NP electro's ... is the attraction.
A true 50 year IPS - tough !
Hey , Kindhornman - What's up with that "blowtorch crowd" over there ...
They keep recycling the same old audiophile "lore" and never build or design
anything (new) ??
OS
It seems like we have a new baby🙄
It's time for dirty job
If i correct remember we have 3 NAD?
Last edited:
Yes , you are technically correct. If you inject an error signal into the servo ,
you will see it reduced by >20db by the CCS "buffer".
But it still acts as this "buffer" , which is far better than a direct return.
PS - the way this is ... the CCS diodes are 95% of the reference ,of which
5% is "shifted" by the servo (which has low pssr itself).
Look below - the CCS actually drops LF psrr by >50db to the input pair..
OS
Attachments
OS,
To me at least half of the Blowtorch thread is pure bs. Frank is the biggest bs'er most of the time and says the same cr*p repeatedly. I do learn things from a few on that thread and when I want a real answer I send a PM and get some real detailed responses. There are some real smart guys on there.
John seems to like to stir the pot once in awhile with Bybee bs and that can take a week to die down once he does that. He also acts like he doesn't understand some things about digital audio but I and some others don't believe that is anything but a call to stay with his analog only designs, of course digital in his talking points is inferior to his analog designs, he just won't concede that things have move forward in the past 40 years.
Richard Marsh is another one to stir the pot but he also knows better about what he does, I think he likes to get others to answer his questions for his own edification but he will almost never answer a question directly, he just answers a question with a question most of the time. This is not without reason, he is fishing and staying tight lipped at the same time. he does have some ownership in his old company in Asia and I think this is where his gathered information is going.
You just have to know the personalities and who will give you a straight and accurate answer. I have a couple of these people on my list in LinkedIn and private E-mail, those are the people who use real science and knowledge in what they do.
I have finally gotten to the point of getting tooling quotes for my speakers to produce the molds for the enclosures. I still plan on using your designs here in my commercial speakers. I am now leaning on the Krypton-C or your new version improvement here and a few other things that I need for a self powered speaker. The xo is still up in the air, whether to use DSP or just a simple electronic Saleen-Key type cross-over. The DSP may be the way to go as I also have been considering a USB input and a couple other things so I can hook these speakers up to any audio source. Stay tuned on that development. You may just see a Kixkstarter or Indi-go-go campaign. Hell I'd even paint then yellow if someone wanted them to match their Pono device! You will not be forgotten in all this and I will do what I originally said I will do, a royalty for every unit sold to those who have helped me along the way on this long journey.
PS, Christophe / Esperando is one of those who's opinion I hold highly, he has the experience on both sides of the puzzle, as both a recording engineer and an electronics R&D researcher.
To me at least half of the Blowtorch thread is pure bs. Frank is the biggest bs'er most of the time and says the same cr*p repeatedly. I do learn things from a few on that thread and when I want a real answer I send a PM and get some real detailed responses. There are some real smart guys on there.
John seems to like to stir the pot once in awhile with Bybee bs and that can take a week to die down once he does that. He also acts like he doesn't understand some things about digital audio but I and some others don't believe that is anything but a call to stay with his analog only designs, of course digital in his talking points is inferior to his analog designs, he just won't concede that things have move forward in the past 40 years.
Richard Marsh is another one to stir the pot but he also knows better about what he does, I think he likes to get others to answer his questions for his own edification but he will almost never answer a question directly, he just answers a question with a question most of the time. This is not without reason, he is fishing and staying tight lipped at the same time. he does have some ownership in his old company in Asia and I think this is where his gathered information is going.
You just have to know the personalities and who will give you a straight and accurate answer. I have a couple of these people on my list in LinkedIn and private E-mail, those are the people who use real science and knowledge in what they do.
I have finally gotten to the point of getting tooling quotes for my speakers to produce the molds for the enclosures. I still plan on using your designs here in my commercial speakers. I am now leaning on the Krypton-C or your new version improvement here and a few other things that I need for a self powered speaker. The xo is still up in the air, whether to use DSP or just a simple electronic Saleen-Key type cross-over. The DSP may be the way to go as I also have been considering a USB input and a couple other things so I can hook these speakers up to any audio source. Stay tuned on that development. You may just see a Kixkstarter or Indi-go-go campaign. Hell I'd even paint then yellow if someone wanted them to match their Pono device! You will not be forgotten in all this and I will do what I originally said I will do, a royalty for every unit sold to those who have helped me along the way on this long journey.
PS, Christophe / Esperando is one of those who's opinion I hold highly, he has the experience on both sides of the puzzle, as both a recording engineer and an electronics R&D researcher.
Last edited:
It seems like we have a new baby🙄
It's time for dirty job
If i correct remember we have 3 NAD?
Yes we tried no servo , both non-inverting and inverting direct servo.
here we have nearly a "non-servo" servo 😀 , I just did my sims with
NO servo - NO change (except offset).
But , we have it hooked to the super VAS (I'm still listening 🙂).
Might be a very good combo !
There really are no other CFA topologies out there beside paralleled
NAD's (accuphase amps).
OS
Thank you , Kind !
I actually glean snippets from the blowtorch "pub" .. 😀
I don't comment much , as they know I am a "pitchfork villager". I'm sure my
relative ignorance offends them.
I want an optical DAC now.
I use my PC the most for listening , but an old CD
player has a bit more clean-ness as a source. I know it's the PC's dirty SMPS.
If I run a real long line cable and plug into another house circuit , better sound
but a very faint ground loop is crated by the 100' run of earth wire.
Nothing plugged in - absolute silence out of the amp.
All the IPS's , so far -
- Kypton C , real clear perfect imaging .... as my kids say "ghost sounds".
Things you could not hear on either OEM amp I (had).
- wolverine , just about as good as the C (but not quite).
- spooky , a VFA - highs are like the wolverine .... but the BASS !!
for some reason , the leach topology pushes them woofers harder.
Its tests the same as the others , but you can hear the difference 😕 .
I've listened for days to each IPS, loud/quiet - 5 different genre's of music -
FLAC/CD/ even radio-movies.
Topologies do sound different
.Listening is better than debating the technical side.🙂
OS
Zetex video transistor type ZTX 558 pnp E line npn also type to92 ,400v 0,5 W only,cob 10 pf.
fast but no high power version, and also type smd FMMT 491 ? any know about these transistors.
What about mosfet drive bjt , to discrete IGBT? Accuphase use that's with Toshiba fet.
fast but no high power version, and also type smd FMMT 491 ? any know about these transistors.
What about mosfet drive bjt , to discrete IGBT? Accuphase use that's with Toshiba fet.