Leach amp - "spooky" is done.
WHAT an amp ! Never tried my hand at this topology , I like it.
Massive slew , got it down to single digit PPM without beta enhancement.
Goes real well with the Hawksford/servo combo.
Stole the HK-990 servo (inverting servo) , amp will run without an
input cap (DC). Slew is awesome with the DC input (200+).
(below) is schema + .asc.
Slewmaster.txt is now (Keen + ksa1015/ksc1815).
(Below 2) is static offset ... WOW !
PS- I like laying out the ones that are " a' drippin' with parts.
Sorry...Mlloyd1 , the hafler folded cascode is not even close (.05% vs. 5ppm)
OS
This is the classic fully balanced symmetrical VFA. They are NICE!
Good job OS - looks like with your std OPS and a selection of IPS+VAS everyone has a plug 'n play choice
If you want that inverting amp's feedback to be independent of the source output impedance the you need an input buffer, I used it in mine gain clone amp, JLH buffer is good for that.
[looking for 200W-250W/8]
Thx-RNMarsh
Richard,
There is 200 W amp in my thread with lateral MOSFET OPS and Chrisfr made nice PCB layout. Maybe if you ask him to send to you the gerber files then you can order the PCB production. Are you wait for assembled PCB to drop in to your lap?
By the way, you probably did not start to assemble GainWire.
BR Damir
Boards might be available soon ..... professional.
These days , even the best amps are willing to go back to the "classic"
(below) . I heard this one (repair) /// it rocks !! 21'st century Leach ...
They put the whole IPS in SMD. (small)
OS
Attachments
CFA - X is done.
Had to check the other VSSA variants out there ...
Peeceebee had the decoupling of both the IP pair and
the current feedback on the same ground traces ... but it
still works !
Jkeutmann's through hole variant is better , the original lazy cat
thread shows just about every mod that can be done .
Here... (CFA-X) , we might drive 400W EF3's and use 85V rails.
More considerations ....
-Separate zener diode regulator and current feedback ground returns
which are fully symmetrical and terminate at "star-OV ".
Current feedback as designed is 20mA at 70V rail clip. SOME
tweakers might want to lower CF resistance to hit 300V/us+
....50ma could be passed to the ground centerpoint ... balanced
(same length) traces will cancel this out here. You don't
want the regulator decoupling to "see" any of this garbage .
-The input pair ... At high rails and heavy current feedback , all
you have between you and "magic smoke" is Q1 and Q2 (below 1).
THEY BETTER BE GOOD !
KSA916/KSC2316 are 30p Cob 120 Vceo heavy duty vertical output devices ...
http://www.mouser.com/ds/2/149/KSA916-66917.pdf
http://www.mouser.com/ds/2/149/KSC2316-59562.pdf
MOUSER has 6K each of the "Y" grade. These are better for
a big output voltage swing monster amp running at 70V+ rails.
CCS's just supply 1.6-1.8mA (ksa992/sc1845) is alright here.
Look it over ... I'll have my glass of Argentine Vino while I stare at it
(below 2) for awhile ... I think it's 100% ( and the best of the Variants) -
for "ungodly" power levels (HD VSSA).
OS
Had to check the other VSSA variants out there ...
Peeceebee had the decoupling of both the IP pair and
the current feedback on the same ground traces
... but itstill works !
Jkeutmann's through hole variant is better , the original lazy cat
thread shows just about every mod that can be done
.Here... (CFA-X) , we might drive 400W EF3's and use 85V rails.
More considerations ....
-Separate zener diode regulator and current feedback ground returns
which are fully symmetrical and terminate at "star-OV ".
Current feedback as designed is 20mA at 70V rail clip. SOME
tweakers
might want to lower CF resistance to hit 300V/us+....50ma could be passed to the ground centerpoint
... balanced(same length) traces will cancel this out here. You don't
want the regulator decoupling to "see" any of this garbage
.-The input pair ... At high rails and heavy current feedback , all
you have between you and "magic smoke" is Q1 and Q2 (below 1).
THEY BETTER BE GOOD !
KSA916/KSC2316 are 30p Cob 120 Vceo heavy duty vertical output devices ...
http://www.mouser.com/ds/2/149/KSA916-66917.pdf
http://www.mouser.com/ds/2/149/KSC2316-59562.pdf
MOUSER has 6K each of the "Y" grade. These are better for
a big output voltage swing monster amp running at 70V+ rails.
CCS's just supply 1.6-1.8mA (ksa992/sc1845) is alright here.
Look it over ... I'll have my glass of Argentine Vino while I stare at it
(below 2) for awhile ... I think it's 100% ( and the best of the Variants) -
for "ungodly" power levels (HD VSSA).
OS
Attachments
Last edited:
I you look at the circuit , it is different than the others. The current feedback
nodes are current sourced from a +/- 12v supply (the zeners).
Each input device only sees only the rail + 12V. This automatically doubles (-12v)
the voltage this IPS can handle.
Globally , this IPS could run from 35V -85V (my design goal).
With lower Voltages , R13/16 might need to be reduced to keep the
Zeners at 12V. Input devices could be "cheeper" lower Ic (ksa1015/sc1815)
, as well.
That's all ... good for all the OPS's.
PS - you would want to do the reverse ... the OPS can run "boosted rails".
(IPS at a regulated 70V - OPS at 60V , for example)
OS
Last edited:
Thanks for the explanation OS. I have a couple of Lazy Cats VSSA boards, how different is your CFA-X design than that? I look forward to a complete tested amp from input to output and the opinions on the sound.
ps. I guess I should have said the input section as your output section is very different I believe than what LC did in that regard.
ps. I guess I should have said the input section as your output section is very different I believe than what LC did in that regard.
Last edited:
There is not much to the VSSA , it's just a simple I-V converter where
there are just 2 devices doing most of the amplification (I).
CFA-X just level shifts the NFB node to the 12v supplies to double
the voltage capability. This amp , lazy cat's ,VSSA, peeceebee are all the
same input section.
This variant is just customized for the slewmaster OPS.
OPS is already tested , this IPS is so dang simple/stable .. it can't fail.
OS
OS,
Since this a very fast moving thread it is getting difficult to keep track of what is the current version of things are. Can I make a suggestion? Perhaps there might be some benefit to using your ability to edit post one by linking to the salient posts containing the final version of the various OPS and IPS as they are completed. That will make it much easier to find the relevant schematics, layouts, Sprint files etc. Just a thought.
Since this a very fast moving thread it is getting difficult to keep track of what is the current version of things are. Can I make a suggestion? Perhaps there might be some benefit to using your ability to edit post one by linking to the salient posts containing the final version of the various OPS and IPS as they are completed. That will make it much easier to find the relevant schematics, layouts, Sprint files etc. Just a thought.
Balanced or bridged ? A bridged is fed from 2 op-amp's (or directly),
fed by a balanced input....
Two slewmasters could be run this way with lower rails.
OS
Holy poop !! They did fix that issue ... I had no idea and was going to
suggest this to the "powers that be".
All (v1.2) work will be on post #1 .... everyone !!
OS
Holy poop !! They did fix that issue ... I had no idea and was going to
suggest this to the "powers that be".
All (v1.2) work will be on post #1 .... everyone !!
OS
Excellent! The extra effort I'm sure will be appreciated by many. The 'issue' has been fixed for some time now. Thanks for all your hard work OS.