"Circumcising" the F3

[Fuling]

Hi

Let´s say I´m building an F3 amp but I don´t like the Aleph CCS.
Removing R18, R19 and C9 (refering to the ZV9 article) should take care of that, but wouldn´t it also be a good idea to decrease the rail voltage and increase the quiescent current a bit?

F3 operates at 50V (minus some losses in the cap multiplier and cascode circuits) and 2A, perhaps something like 40V and 2,4A would be better in this case?

[traw]

Brings up something i was curious about. Seems F3 actually at 1.5amp or so, the 0.6 volt across 1-1-2 parallel resistors versus the 1-1-1 in zen 9. So, maybe bringing amperage up to 2amp for first observation worthwhile.

[GRollins]

I'm not sure I follow your reasoning on this. It sounds as though you're saying that the Aleph current source is better at higher voltages/lower currents and a normal current source is better at lower voltages/higher currents.
To some extent you can make the argument that higher bias current is always good, but you can also make a case for higher voltage...if for no other reason than the fact that a MOSFET's capacitance drops with increasing voltage.

Grey

[flg]

Originally posted by Fuling
...but I don´t like the Aleph CCS... but wouldn´t it also be a good idea to decrease the rail voltage and increase the quiescent current a bit?

Well, I like it, but I have this thing about trying to find something else as good or better On the rail Voltage, it effects your linearity somewhat and maximum output level.. I found the opposite was benificial, see below. Maybe drive an F4 with it or something. On the Iq, same deal, but in this case, the configuration was developed to use the load line cancellation technique with that size Rsource and that Iq. But, read on...

Originally posted by traw
Brings up something i was curious about. Seems F3 actually at 1.5amp or so, the 0.6 volt across 1-1-2 parallel resistors versus the 1-1-1 in zen 9. So, maybe bringing amperage up to 2amp for first observation worthwhile.

Was that elluding to something I thought odd in the ZV9 schematic and voltage and current refrences. Something like a 2A
current source but the JFET is set up for 1.5A? Am I confused here?

Originally posted by GRollins
I'm not sure I follow your reasoning on this. It sounds as though you're saying that the Aleph current source is better at higher voltages/lower currents and a normal current source is better at lower voltages/higher currents...
Grey

Is it possible that you want the ability to clip the output into 8 ohms without any wasted Iq or voltage rail? What, no headroom

[Fuling]

What I meant was that if I turn the Aleph CCS into a static CCS the maximum output current (for a specific idle current) would be reduced, right?
With less output current swing available there would be no need for such a high(ish) rail voltage (not taking mosfet capacitance and stuff into consideration).

I think that even if I use a static CCS and increase the current to, say, 2,4A it would still be enough voltage headroom if i cut down the voltage a bit (to save the planet and my mosfets...)

[Fuling]

flg, you posted while I was writing.

The whole point in reducing the rail voltage would of course be to decrease power dissipation, not the headroom.
I recall Nelson wrote something about "prefering current over voltage" but that was in different context.
The Jfet loadline stuff might complicate matters a bit if I choose to go astray from the proven path...

Another interesting option would be to insert switches or relays to switch between "active" and "passive" CCS.

[GRollins]

One of us is confused--quite possibly me; I may not be understanding the question.
The Aleph current source doesn't require a lot of extra headroom. You'll lose the Vgs of the MOSFET, but then you'll lose the Vgs of the MOSFET in a "normal" current source, too. If you're really concerned about reducing the overhead, you'll need to go to bipolars. That will take you from roughly 3-4V for a MOSFET's Vgs, to about .65V for the bipolar.
You can adjust the rail voltages and bias current by ten or twenty percent before you need to look at recalculating some of the parts values. But as I've said before, the rail voltage determines the output wattage and the bias determines how low an impedance the amp will drive. It's up to you to decide what your targets are. Heat dissipation falls directly out of the rail and bias numbers.

Grey

[Fuling]

From the ZV9 article:

"Referring to fig. 8, the current source has the addition of C9, R18 and R19 that set the current source AC value to track an arbitrary fraction of the output. Usually we set this value around 50%, and we can use it to "ghost" the complex load impedanceat a negative multiple of the load value. With the values chosen in figure 8, an 8 ohm load would look like 16 ohms to the Jfet gain device"

So without the Aleph CCS an 8 ohm load would look like 8 ohms to the Jfet, right?
An 8 ohm load requires 1,414 times more current and 0,707 times less voltage to develop the same power as an 16 ohm load, unless I´m completely lost.

With the stock 50V 1,5-2A the amp would run out of current before it runs out of voltage IF the Aleph CCS was disabled (and the load was 8 ohms, no "ghosts"involved).
Compare to the old ZV2 whích operates at 34V (IIRC) and 3A.

[Babowana]

Originally posted by Fuling


F3 operates at 50V (minus some losses in the cap multiplier and cascode circuits) and 2A, perhaps something like 40V and 2,4A would be better in this case?

It might be.

But, the alteration of Aleph CCS to the plain CCS means that we are ready to accept lower efficiency of the amp as well as lower damping factor.

Among others, the most important point is to be how the sound would change, IMHO.

Good luck!

[Fuling]

the most important point is to be how the sound would change

In the end, that is the only thing that matters.

So far this is only a theoretical discussion, I have a pair of F3 boards and most of the other (expensive) stuff already and I will report as soon as anything happens.