F5 power amplifier

[juma]

Hi estman,
your idea seems atractive (adding source follower to drive the output MOSFETs) but it implies higher value drain resistors for input JFETs thus giving the circuit much higher open loop gain (about 70 dB) and consequently larger NFB which will alltogether change the amp's character - it won't be F5 any more, it won't sound that nice any more.

What we might need (in order to drive two or even three pairs of MOSFETs) is lower impedance of the previous stage. That can be achieved (in some extent) by using V grade JFETs and cascoding can take care of JFETs higher dissipation if higher PSU voltage is chosen.

[jackinnj]

Originally posted by juma
What we might need (in order to drive two or even three pairs of MOSFETs) is lower impedance of the previous stage. That can be achieved (in some extent) by using V grade JFETs and cascoding can take care of JFETs higher dissipation if higher PSU voltage is chosen.

Paralleling BL k170's and j74's the JFETs and cascoding?

[juma]

Originally posted by jackinnj

Paralleling BL k170's and j74's the JFETs and cascoding?

It's easy for you - you got a store full of them

[twitchie]

Finally got a chance to fire up my F5 boards (one of them so far)

Thanks to Papa for sharing this with us. It's going inside a Sumo Polaris chassis for now, at least until I can decide how much I like it ... cases and heatsinks don't grow on trees around here


My current Kitchen Table

[estman]

Originally posted by juma
Hi estman,
your idea seems atractive (adding source follower to drive the output MOSFETs) but it implies higher value drain resistors for input JFETs thus giving the circuit much higher open loop gain (about 70 dB) and consequently larger NFB which will alltogether change the amp's character - it won't be F5 any more, it won't sound that nice any more.

Hi juma,
you make a good point about higher OLG. It must be avoided. I plan to use Toshiba mosfets as they have ~2x lower Vgs compared to IR & Fc parts.
Also, if you have paralleled output pairs with lower current in each mosfet, you have lower drop across R11,R12. Increasing the value of R11,R12
(to have same voltage drop as orig.) lowers the gain of output stage.

[juma]

Hi estman,
it's surely worth experimenting - especially if you have lateral MOSFETs at hand. They are not that hard to find, it's just that they are quite a bit more expensive - on the other hand they are known for being able to sound good even with lower bias - not so deep in class A. Certainly an amusing experiment Good luck and don't forget to report the results.

[jackinnj]

Originally posted by twitchie
Finally got a chance to fire up my F5 boards (one of them so far)

Thanks to Papa for sharing this with us. It's going inside a Sumo Polaris chassis for now, at least until I can decide how much I like it ... cases and heatsinks don't grow on trees around here

I have a Sumo Polaris sitting to my immediate right -- here in the shop I use 2 amps -- a Hafler 200 and the Sumo Polaris. Interesting coincidence.

[enzedone]

Is the track layout correct for P1 & P2?

Sorry for pleb question, but just want to make sure.

[estman]

I'm astonished - like there isn't original F5 layout at all.
From picture I can see that input circuitry is kept away from high current pow supply-mosfets-output to loudspeaker loop.
Isn't it important ?

[zinsula]

Originally posted by juma
.............it's surely worth experimenting - especially if you have lateral MOSFETs at hand. ...........

But Toshiba MOSFET's are not laterals...
They are very complementary verticals, with low Vgs threshold.

Tino