F6 Amplifier

Zen Mod · 2012-09-13 12:24 pm

F4 is

BA2 PP OS (also useful for BA3)

buzzforb · 2012-09-13 12:28 pm

Why dual TL over single(like f4/ba2 output). Better control. Perhaps stupid question. Why would TL outperform ccs loaded LED?

Antoinel · 2012-09-13 12:35 pm

lhquam said:
Since my amplifier power needs are happily satisfied with less than 25W, I have no real interest in spending (wasting) time of Class-AB circuits.

IMHO, the challenge will be to operate Conceptual F6 and its variants in Class A/B or even Class B modes and still make great sounding amps. I'll do it when the diy PCB by Mr. Pass becomes available for sale.

buzzforb · 2012-09-13 12:46 pm

Antoinel, just about every major amp manufacturer i know of tries to maintain class A as much as heat allows. There is really no doubt it improves performance. A/b is perfectly viable above low power levels, but you can keep it for anything else, IMO.

Antoinel · 2012-09-13 1:19 pm

buzzforb said:
Antoinel, just about every major amp manufacturer i know of tries to maintain class A as much as heat allows. There is really no doubt it improves performance. A/b is perfectly viable above low power levels, but you can keep it for anything else, IMO.

Please count me in for PCB, parts and semis to assemble Conceptual F6 as you make them available..

Zen Mod · 2012-09-13 1:26 pm

buzzforb said:
Why dual TL over single(like f4/ba2 output). Better control. Perhaps stupid question. Why would TL outperform ccs loaded LED?

dual vs. single because you need two bias nets :clown:

TL will not outperform LED,diode or zenner based bias net

in fact , appropriate tempco will win stability wise but soundwise there will not be difference

sonnya · 2012-09-13 1:58 pm

Nelson Pass said:
ZM has been nagging me for information on the F6. I have a very nice
sounding prototype, but the design and the article that accompanies it
have some details remaining.

However, I can afford to put out this teaser:

Buzzforb Mentioned something..

Sorry for taken the first post and quote on, but.....

looks like it can use a pulse transformer buti would suspect a ratio 1:10:10 -> 1:100:100???

Am i right???

Zen Mod · 2012-09-13 2:48 pm

pulse xformer is certainly not adequate

those are for switching duties

whatever - define - are you asking about 1:10 , or about 1:100 xformers ?

sonnya · 2012-09-13 2:56 pm

Zen Mod said:
pulse xformer is certainly not adequate

those are for switching duties

whatever - define - are you asking about 1:10 , or about 1:100 xformers ?

Thats the ration of the transformer. This would give you system gain.

But you are correct. And the transformer will never be saturated....

Zen Mod · 2012-09-13 3:06 pm

too much gain , so too much feedback then

sonnya · 2012-09-13 3:10 pm

Why is it to much gain???

Zen Mod · 2012-09-13 3:12 pm

go back and read what lhquam wrote about OLG amount

whatever Rs value you choose ( so amount of OLG ) , with 1:10 xformer that goes by factor 10

to keep amp in normal gain range (say 1V5 for full out ) , amount of feedback must rise , for my liking too much

lhquam · 2012-09-13 3:25 pm

I was looking for a low part count approach. Perhaps batteries are the solution for that.

buzzforb said:
Why dual TL over single(like f4/ba2 output). Better control. Perhaps stupid question. Why would TL outperform ccs loaded LED?

Zen Mod · 2012-09-13 3:31 pm

what's wrong with plenty of parts ........ just look at my Lightning version - 18 LEDs per channel

:rofl:

buzzforb · 2012-09-13 3:45 pm

sonnya said:
Why is it to much gain???

Perhaps different design goals, but Nelson seems to try to get away with as little gain and feedback as possible. I believe using the semi south in one amp, he stated that with the added linearity achieved from the part, he reduced gain and feedback giving back some of the performance gains as a result. He truly believes in as simple and minimalistic as possible, but no less.

Antoinel · 2012-09-13 4:02 pm

Zen Mod said:
what's wrong with plenty of parts ........ just look at my Lightning version - 18 LEDs per channel

please direct me to the Lightning version. Thank you.

Zen Mod · 2012-09-13 4:04 pm

schematic posted by me - hand drawn - few pages back

tiefbassuebertr · 2012-09-13 4:54 pm

Zen Mod said:
pulse xformer is certainly not adequate
those are for switching duties

Why are SMPS transformer ferrite cores like that one about
ferrite core transformer - Detailed info for ferrite core transformer,ferrite core transformer,ferrite core transformer,EE/EI on Alibaba.com
are not suited for linear audio driver transformer?

Which special core requirements are necessary for best results in an application like the F6?

Concerning output transformers in tube amps there are several design rules like follow:
http://www.geofex.com/article_folders/xformer_des/xformer.htm

But the idle current values so as peak current values in transformers for fit power output stages are much more lower.

Zen Mod · 2012-09-13 4:57 pm

BudP is a guy to tell that

lhquam · 2012-09-13 5:27 pm

TL will not outperform LED,diode or zenner based bias net

You appear to be correct about that. Here is a Spice simulation of the two. The plot shows the ripple rejection of the two approaches. As you can see, the LED+CCS has a much lower corner frequency than the TL and is about 16dB better at frequencies about 100Hz.

Zen Mod said:
dual vs. single because you need two bias nets

TL will not outperform LED,diode or zenner based bias net

in fact , appropriate tempco will win stability wise but soundwise there will not be difference

F6 Amplifier

Official Court Jester

Member

Member

Member

Member

Official Court Jester

Member

Official Court Jester

Member

Official Court Jester

Member

Official Court Jester

Member

Official Court Jester

Member

Member

Official Court Jester

Member

Official Court Jester

Member

Attachments