Slewmaster - CFA vs. VFA "Rumble"

[ostripper]

Is this a comparison of CFA vs VFA or is it more about the benefits of the Cascode? - I'm looking for some opinion on the CFA vs VFA thingy ?

At the end of the day, do you feel you'll be building CFAs more often than VFA's or is there really no reason to pick one over the other because in the end it's the detailed building blocks that make the difference. Or are they just all the same when you design them this darn good !?

This was my "rumble" goal - to design 1 real good CFA and one REAL good
VFA.

NAF's NAD w/servo is 1mV offset with 6-8ppm thd20 ... 90db PSRR.

This one ,the wolverine (current work) ... is 2-3ppm without clamping ,5-6ppm with protection).
- - 120db PSRR !!
The IPS's are so dang cheap to build (10-15$ apiece ) , and the OPS
is modular and standardized.
Plug either one of these IPS's in for a day. unplug it , listen to the other
for a few days.. the builder can decide. (which sounds better ??)

This also eliminates waste. The EF3 is 21'st century HK680/990 tech.
Any amp on the forum will work with this EF3 and be super stable - I have about 20 IPS's to try !!
I also have a simpler CFA (VSSA variant) and the symasym VFA up next
they are 20-40ppm "contenders" for those who like it simple.


Below - are the 5 IPS's

1- NADservo (CFA) - all done, NAF loves it ...

2- Wolverine (VFA) - better blameless - done real soon (tomorrow).

3- CFA-X , real simple intro to the "CFA world" (next layout)

4- Symasui (VFA) - like the symasym , but with sansui z3900 comp./ltp.

5 - a 21'st century leach amp (spookyamp -symmetrical VFA)

PS - 3 and 4 are almost done , as well.

OS

 

[bimo]

Simple VFA with symmetrical VAS.
Slew rate is 140V/uS. Distortion is not low but may be "good enough".
OS, can you tweak it for better THD?

 

[MiiB]

The leach frontend has huge potential for a superb VFA for even better performance a current mirror in the two branches will be good.

 

[bimo]

The leach frontend has huge potential for a superb VFA for even better performance a current mirror in the two branches will be good.

Yes, I know. I was sim Bonsai's e-Amp and get better THD dan slew rate. This is about being simple as I can but still can have high symmetrical slew rate.

 

[Bigun]

It clearly notice even when I use laptop/notebook internal soundcard and mp3.
Next time, I will try amp with simple as possible, try to get highest slew rate and then trade the slew rate a bit for lower THD.
I want to know why amp can sound more "airy" than other amp.
Slew rate, pattern of harmonic distortion? Unfortunately, I do not have instruments to confirm it. If I can borrow RNMarsh's instruments.....

I believe you are right, that lowest THD is not the only option for good sound. I do like a clean sound and find that my best amps currently are those with relatively high feedback. However, I suspect that 'air' is something that is diminished in such designs if special attention is not given to it. Hugh Dean (AKSA) knows how to design 'air' into an amplifier and he has history with tube amps. I believe the answer is complex since it likely depends on IM distortion more than the harmonic profile itself but he has said that the 'Hiraga' profile of monotonically decreasing harmonic distortion is the best. I notice that Hugh mentions the use of nested feedback in his amplifiers - perhaps this is to help get this profile. I do the same in my TGM8 amplifier (CFA with nested feedback - but relatively high open loop gain - sounds really good).

 

[buzzforb]

OS,
I must admit, I am interested in the "simple" CFA, although the spooky looks entcing as well. While I am sure that the others offer incredible performance, I hestitate to build them based on my limited experience.

 

[Kindhornman]

Buzzforb,
There does not seem to always be an improvement in the actual sound quality to the high parts count implementations. From my reading of this thread and others it seems that many times the simpler circuit is the preferred for sound even if the more elaborate circuits may have lower ppm distortion or some other measurable improvement in electrical signal. Why is that I couldn't answer, but perhaps like anything else the less you stir the pot the better?

 

[buzzforb]

I tend to agree. Accuracy does not always lead to enjoyment. I somethimes think of each transistor used as creating a copy, and the more copies you use, the less the copy looks like the original. This is soley opinion and i am overwhlemed with what OS has done, but it would be interesting to compare a simple version vs a complex version.

 

[PETR NICONOV]

http://www.diyaudio.com/forums/soli...imate-fidelity-amplifier-305.html#post3816330 To [URL="http://www.diyaudio.com/forums/members/bimo.html"]Bimo
Simple VFA with symmetrical VAS You can see if review the report link
http://www.diyaudio.com/forums/soli...imate-fidelity-amplifier-305.html#post3816330
[/URL]

 

[ostripper]

Simple VFA with symmetrical VAS.
Slew rate is 140V/uS. Distortion is not low but may be "good enough".
OS, can you tweak it for better THD?

Yes, sir bimo ....

Ask and you will receive.
Leach/hawkford dropped thd from 45 to 8ppm.

OS

 

[ostripper]

New models/ low voltage semi's

Mouser has 123000+ of these in stock , both P/N y grade.
PERFECT low voltage IP stage semi's.
Low vsat ,4pF cob, 80mhz Ft ....
Current production models !!

http://www.fairchildsemi.com/ds/KS/KSA1015.pdf
http://www.fairchildsemi.com/ds/KS/KSC1815.pdf

Models are below ...

OS

 

[ostripper]

The leach frontend has huge potential for a superb VFA for even better performance a current mirror in the two branches will be good.

Leach topology does not "get along" with CM's too well. Many have
tried/discussed this drawback.

To get more loop gain , beta enhance the VAS.

OS

 

[mcd99]

What are your feelings about the c1845/a992 transistors? They have lower Cob. Aside from the current capability they look pretty good.

Really enjoying this thread by the way.

 

[mlloyd1]

OS:

regarding symmetrical front-ends, have you ever explored for differences between the leach and the psuedo-folded cascode (ala borbely/curl) ips?

mlloyd1

btw, did you every try Dip-trace? you get schematic capture and the error checking that goes along with it. also, not forced to use autoroute for layout. i have & use Sprint also but need schematic capture to catch my errors .

Leach topology does not "get along" with CM's too well. Many have
tried/discussed this drawback.

To get more loop gain , beta enhance the VAS.

OS

 

[ostripper]

What are your feelings about the c1845/a992 transistors? They have lower Cob. Aside from the current capability they look pretty good.

Really enjoying this thread by the way.

Thanks .... I enjoy creating it.
The high Vceo device saturation characteristics are worse in low I/V applications. The 992/1845 pair is wonderful for high Vce applications ,
such as cascodes /CCS's ... the beta enhancement device.

At 2-3pF , they are in the same range as the SS9014/15 , and just
a little better than the 1015/1815 pair I just posted (4-7pF).

All my designs are 1-2ma at IP , 4-8 ma at VAS , low Ic devices are
perfect. I run em' cool ....

OS

 

[ostripper]

OS:

regarding symmetrical front-ends, have you ever explored for differences between the leach and the psuedo-folded cascode (ala borbely/curl) ips?

mlloyd1

btw, did you every try Dip-trace? you get schematic capture and the error checking that goes along with it. also, not forced to use autoroute for layout. i have & use Sprint also but need schematic capture to catch my errors .

I never played around with JC's designs . I'd be 'a sliding in the snakeoil. I
like japanese designs !!

Never tried dip trace (I'll google it ), just LT/sprint ...

OS

 

[mcd99]

Thanks .... I enjoy creating it.
The high Vceo device saturation characteristics are worse in low I/V applications. The 992/1845 pair is wonderful for high Vce applications ,
such as cascodes /CCS's ... the beta enhancement device.


At 2-3pF , they are in the same range as the SS9014/15 , and just
a little better than the 1015/1815 pair I just posted (4-7pF).

I see. Maybe good alternatives to BC550/560.
The SS9014/15 look good but where to get the SS9015?

All my designs are 1-2ma at IP , 4-8 ma at VAS , low Ic devices are
perfect. I run em' cool ....

OS

We see eye to eye here. I like the low current VAS too.

Keep up the good work

Paul

 

[Carl_Huff]

I really like LT/ Sprint. My only grumble is that they don't have a way of creating or importing schematics as a check to be sure that I have the parts and their pins connected properly. I get really stupid much too easily!

 

[ostripper]

Drippin' with parts ... Wolverine V1.2

Here it is.

Self would like it ...

2-4ppm THD 20 , transistor or diode clamp , buffered cascode ...
All the embellishments ...

ksa1015/sc1815-ss9014 for IP low voltage devices ,
ksa992/sc1845 for cascode/ccs/vas beta , and

mje340/350 or any other high Vceo device for main VAS-CCS.

PS -(below) the ss9014/15 models , mouser has 10K 9014(npn) ... fairchild stopped
making the 9015(pnp) ...
OS

 

[ostripper]

I really like LT/ Sprint. My only grumble is that they don't have a way of creating or importing schematics as a check to be sure that I have the parts and their pins connected properly. I get really stupid much too easily!

I had to adapt .... I'll have a glass of wine with both schema/layout in
front of me - the errors "pop out" and bite me on the nose ...

OS