So by this are you saying you think the C is lacking in bass? I didn't notice that at all. But to qualify, I'm not a fan of sub woofers except for movie soundtracks.
Not lacking , but I did put it on the sub OPS to replace the wolverine.
Just a little less "grunt" than a VFA. but , my sub has a 5KG magnet and 60mm
VC .... not the same load as my full range setup. 😱
OS
Back to work .... Kypton- ND
Just finished my 4th new design ---> Kypton -ND.
"Kypton -C" needed a "brother" 😀 ....
The old NAD ... with -
- CCS/servo (not in audio path).
- Super hawksford/baxandall VAS.
- Cascode on 2'nd stage.
- Q1/5 and Q2/6 fully negate thermally (just leaving the negative LED CCS's).
-also .... "clips like a tube" !!
(schema below 1)
-Upcoming ... the bootstrap amp , simetri , this Kypton , and I have permission
to do ONaudio's " SYMEF.
SYMEF = a blameless/LIN with a mirrored cascoded symmetrical drive VAS. The
level shifter approach is different. (symef concept- below 2)
Edit-( below 3 )... Kypton-ND does the "overload thing" better than "C".
plot - super pair base (top) , clamped super-pair emitters (middle) , and tube-like
rounded overload (bottom). 15ppm and nV offset , as well.
OS
Just finished my 4th new design ---> Kypton -ND.
"Kypton -C" needed a "brother" 😀 ....
The old NAD ... with -
- CCS/servo (not in audio path).
- Super hawksford/baxandall VAS.
- Cascode on 2'nd stage.
- Q1/5 and Q2/6 fully negate thermally (just leaving the negative LED CCS's).
-also .... "clips like a tube" !!
(schema below 1)
-Upcoming ... the bootstrap amp , simetri , this Kypton , and I have permission
to do ONaudio's " SYMEF.
SYMEF = a blameless/LIN with a mirrored cascoded symmetrical drive VAS. The
level shifter approach is different. (symef concept- below 2)
Edit-( below 3 )... Kypton-ND does the "overload thing" better than "C".
plot - super pair base (top) , clamped super-pair emitters (middle) , and tube-like
rounded overload (bottom). 15ppm and nV offset , as well.
OS
Attachments
Last edited:
So you've finally found out that the subtraction on a single device is the way to go. What you have here is VSSA principle with enhanced VAS, but still with wrong feedback R-C configuration, the reason for weak bass. Just to inform you VSSA can even do better, ie. her bigger brother is the answer.
In my controlled situation , I'll compare the "ND".
As far as the "wrong R/C" ... do you mean more current and a bigger DC cap?
Simulation does show this gives lower LF THD , but would this be audible ?
I'm about to make the "transistion" Kypton-ND ,will be mixed SMD/through-hole.
OS
Hardcore CFA tests - kypton-ND
(below 1) - the PSRR is acceptable for a CFA.
Improvements in the power supply (dual mono or variants) will
make for good SQ.
(Below 2) better gain margin and higher UGLF than most.
(below 3) absolutely perfect rail to rail squarewaves with 270V/us slew.
As I said , thermally ... less than 5% VAS Ic variation (0-50C)
The infamous "Infidel" was the source of this perfection.
The layout and performance should actually be as good (or better in some aspects) ,
than the kypton C. NO DC caps , single current FB path , single simple compensation path.
Oh , better 1st/2nd stage BCxxx devices .... (high 500Hfe)
Thimios built 2 versions of the old "NAD" - I know it performs.
This performs X2 !! 🙂
OS
(below 1) - the PSRR is acceptable for a CFA.
Improvements in the power supply (dual mono or variants) will
make for good SQ.
(Below 2) better gain margin and higher UGLF than most.
(below 3) absolutely perfect rail to rail squarewaves with 270V/us slew.
As I said , thermally ... less than 5% VAS Ic variation (0-50C)
The infamous "Infidel" was the source of this perfection.
The layout and performance should actually be as good (or better in some aspects) ,
than the kypton C. NO DC caps , single current FB path , single simple compensation path.
Oh , better 1st/2nd stage BCxxx devices .... (high 500Hfe)
Thimios built 2 versions of the old "NAD" - I know it performs.
This performs X2 !! 🙂
OS
Attachments
That's for me to know and you to find out with a toner transfer. 😛
It's the other method for a flawless CFA IPS.
PS - It's not the "infidel" ( and thimios built it twice - before you came).
OS
I can't do the toner transfer without the file. I was here when thimios built the NAD. However you gave the impression that the cfa-xh was the newer, better design so I built that instead.
Wasn't that the version that was going to be the gonad that got a different but more pc name to it?
Os why are you talking about DC blocking when you're DC coupled..??
Also after building quite a few CFAs I have found that the diamond input types sounds the best, I thnk it's because of the singularity in the feedbac-path and the fact that you don't have/need capacitor blocking.
Also after building quite a few CFAs I have found that the diamond input types sounds the best, I thnk it's because of the singularity in the feedbac-path and the fact that you don't have/need capacitor blocking.
The current is not the probllem, it is the impedance of the feedback path VS the parasitic capacitance of the emiter of the input transistor. You need to minimize phase shift at very HF.
Same thing if you use a DC blocking capacitor, with phase turn at LF.
I use one of these per channel for +- 73 vdc rails. Give or take depending on the line voltage.
Evan
Evan
Wow, that is some crazy wattage. You must be driving some kind of massive load to need that kind power. 😱
The bigger transformers don't cost much more and since I build mono blocks space isn't an issue. Two channels of 5 output pair slew monsters are running at about 83 volt rails driving a pair of 15" open baffle woofers each. Lots of eq to get the lower bass so they seem to like lots of power.
Last edited:
Nope , Brother Valery got me just 1 - 1KVA 50-0-50 triad (20A).
Triad Magnetics Product Specs
Split the 16ga secondaries into 2-18ga ... then to 2-35A bridges. This connects
to 4 X 22Kuf capacitor banks @ 73V rails.
Might be enough for china breaking 😀 .
EvanC's "real man's" trafo is really a 400VA , and it gives 65V rails for the sub.
Just a humble 2.1 home theater/music system.
PS - I had to re-machine the main amps to put the outputs in the
middle - my only major screw-up. (below)
Off - center more than 25% horizontal makes the end output pair much higher
bias. I'm under 10% now ... 🙂
OS
Attachments
