CFB-SKA -- Hybrid of VSSA and SKA Amps Concept

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
fab

fab

Member
Joined 2004
fabaudio.online
Paid Member
Hi

This new thread is to explore the concept of an hybrid amp comprised of VSSA input/feedback stage and SKA GB150 bootstrap input stage and inverted buffered output stage. The main difference with the SKA GB150 is the removal of the diff input stage and associated CCS. The result is an amp of "CFB" type with less component thus preliminary called "Very Simple Killer Amp"....;)

See attached schematics. I have not put the components value b/c this is a concept at this stage (but I simmed one version fine so far).
The temperature compensation is intended to be done using VBE multiplier (mounted on output mosfets) in the positive and negative sides in the input stage.
The bias current and DC offset are managed with 2 potentiometers in series with the VBE multipliers.

The GBW is very wide.

Related threads:
http://www.diyaudio.com/forums/solid-state/238593-ska-gb150d-now-public-domain.html#post3546728
http://www.diyaudio.com/forums/vendors-bazaar/225747-vssa-lateral-mosfet-amplifier.html#post3285859

Please let me know if it is worth persuing.

Thanks

Fab
 

Attachments

  • VSKA-fab v 2B0.png
    VSKA-fab v 2B0.png
    20.1 KB · Views: 1,132
Last edited:
fab

fab

Member
Joined 2004
fabaudio.online
Paid Member
Lazy Cat said:
Hi Fab

Of course it is, the more CFB amps on the market the better. ;)

Briefly checked the schematic and it seems OK, will comment after you'll insert the values. :up:
Click to expand...

Thanks Lazy cat. Somehow I knew it would have some interest for you:p
I like the version of your VSSA amp with MOSFET for VAS..;)

JOHN BALI said:
:yes::nod:
I'm waiting too
Click to expand...

See attached. Waiting for comments.:)

BobEllis said:
Cool. Reminds me of a BJT drive version of the Pass F5T V2. Might make an interesting experiment to compare them.
Click to expand...

In fact my first idea was to use a symetrical jfet stage (like F5) in lieu of the SKA diif input before I decided to persue first with bjts symetrical input stage to keep the idea of the highest open loop gain...;) However, I will eventually try the jfet stage too...

Note that the VBE multiplier resistors divider will need to be adjusted to match more closely the VGS drift in temperature of the MOSFET....

All comments are welcome.:cool:

Thanks

Fab
 

Attachments

  • VSKA-fab v 2D0-values.png
    VSKA-fab v 2D0-values.png
    30.1 KB · Views: 910
Last edited:
fab

fab

Member
Joined 2004
fabaudio.online
Paid Member
davidsrsb said:
Q5 and Q6 are Vbe multipliers, not current sources, so supply rail variation goes straight through to modulate the quiescent current
Click to expand...

That is a very good comment :up: thanks!
I will have to revisit that part... :scratch2:
In fact even the vbe multiplier would need to probably be of higher ratio than in my posted circuit...:headbash:
I could still try to adapt and use the original temperature compensated current source of the SKA but I wanted to reduce the parts count ....:rolleyes:
Maybe I can replace R13 and R14 with Jfet as current source as in the VSSA....

Fab
 
Last edited:
fab

fab

Member
Joined 2004
fabaudio.online
Paid Member
CFB-SKA concept now...

Now with the SKA original Current sources...

It is no more a "Very Simple Killer Amp":rolleyes: but a conversion from Voltage feedback to current feedback amp...

The values are preliminary but any comments are welcome!

Fab
 

Attachments

  • CFB-SKA Fab v 1A0-values.png
    CFB-SKA Fab v 1A0-values.png
    31.8 KB · Views: 786
J

J. Wright

Member
Joined 2004
If you use a cascoded jfet front end like that of the old Luxman L-480 you can essentially tie the jfet sources together for self biasing and use that as your feedback point as well. You'll have to juggle values a little because the jfets will run a bit more current than your BJT front end.
 
fab

fab

Member
Joined 2004
fabaudio.online
Paid Member
CFB-SKA

J. Wright said:
If you use a cascoded jfet front end like that of the old Luxman L-480 you can essentially tie the jfet sources together for self biasing and use that as your feedback point as well. You'll have to juggle values a little because the jfets will run a bit more current than your BJT front end.
Click to expand...

Hi

Thanks for the suggestion.
This would be like an F5 Turbo input stage. This could be a next project but for now I want to stick with the bjt input to get the maximum open loop gain...
I have already an F5 amp and I am looking for a different sound with a lot higher feedback...:)

Fab
 
fab

fab

Member
Joined 2004
fabaudio.online
Paid Member
davidsrsb said:
The lack of decent p channel jfets these days makes this symmetrical circuit hard to convert
Click to expand...

2SJ74 are obsolete since a long time and I have only a few left... Also, it is preferable to be matched for IDss with the n channel in the symetrical self bias scheme...

Do you have any recommendation on the last schematics posted?
Should I start prototyping?

Thanks

Fab
 
Last edited:
Bigun

Bigun

Member
Joined 2009
diyaudio.com/community/threads/tgm8-an-amplifier-based-on-rod-elliot-p3a.245619/
Paid Member
I did have a look at some kind of CFB version of the GB150 in some simulations. I didn't really get something working well and didn't keep the file. I may have to revisit if you make some good progress here! - I can see the attraction of marrying the current hot-topic (CFB) with the pedigree of the SKA. I guess it's important to remember that the current sources in the SKA are key to the topology for achieving high PSRR - which is the driving principle behind it's design. I don't think the CFB version will have such a high PSRR since the LTP has a high CMRR by itself ???

Of course, to know if the CFB sounds better than the original SKA you'd have to build both.... :D


p.s. I believe the SKA GB300 uses cascode in the front.
 
Last edited:
fab

fab

Member
Joined 2004
fabaudio.online
Paid Member
CFB or Voltage feedback...

Bigun said:
I did have a look at some kind of CFB version of the GB150 in some simulations. I didn't really get something working well and didn't keep the file. I may have to revisit if you make some good progress here! - I can see the attraction of marrying the current hot-topic (CFB) with the pedigree of the SKA. I guess it's important to remember that the current sources in the SKA are key to the topology for achieving high PSRR - which is the driving principle behind it's design.

Of course, to know if the CFB sounds better than the original SKA you'd have to build both.... :D

p.s. I believe the SKA GB300 uses cascode in the front.
Click to expand...

Hi Bigun

Thanks for your interest. I will surely look closely at your thread since I do intend to compare both original and CFB SKA amps ...eventually:violin:

My simulation shows it works well in CFB (see my latest schematics). However, I may have to experiment with this bjt current source to get the proper compensation for the mosfet VGS drift... I think Greg has worked a lot to achieve the required goal.

There are several brilliant ideas in the GB SKA design and one of which is the absence of the VAS where the gain loss is compensated by the bootstrapped input stage...

I will be looking forward to receiving your comments:cheers:

Fab
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom