Is this a choke loaded Xsoz?

[Fuling]

Since I don´t have any idea of what I have created (mentally), I must ask for advice here. Would this qualify as a choke loaded XSOZ with CCS "tail"?? Would it even work??

BTW: Sorry for the lousy quality

 

[Fuling]

I hate to sound desperate, but c´mon, I know there is a lot of wise men here who fiddles with amplifier circuits all day, someone must have some comment on this?

 

[Henrik]

Hi Fuling

I had this in my "drawer".
It is not a finetuned circiut, but it is close to your scematic.

You need some resistors over the cockes to stabilize them - Nelsons idea.

And the inductors shoud be one inductor with a centertab. This way they are locked to each other through the magnetic field, so when one is going down the other is going up - Circlotrons idea.

Regards Henrik

 

[Fuling]

Circlotrons idea

No offense to neither you or Circlotron, but I believe that is what´s called Push Pull in the tube world. Still a good idea though, especially since I happen to have a pair of homewound chokes that should fit.
I also have a bunch of bigger chokes but I´ll save them for my balanced SE follower (schematic is somewhere else on this forum).
At least ´til this design has been tested properly.

Some resistance across the chokes might be a good idea, even though I haven´t noticed any need for it in my earlier experiments.
If resistance is not enough I guess RC networks would be, a quite common sight around the OPT´s in tube amps.

BTW, is there any reason for the negative rail voltage to be as high as the positive? As far as I can see all the voltage swing takes place in the positive rail (so to speak), so maybe the negative voltage could be reduced drastically?

 

[Circlotron]

You're both right, really.

If I was making a voltage amplifier + phase splitter then I would definitely use a single centre-tapped winding because the outputs are referenced to earth so they *must* be balanced.

For an output stage I would use separate DC biased chokes as per Fuling's cct because the floating load speaker gets driven from the *difference* of the two fet drains, i.e. it is not ground referenced. Hopefully, this will lead to near cancellation of even order harmonic distortion. Besides, this will mean that each fet gets negative feedback from itself only, not from itself in parallel with the other one. Seems tidier to me.

I like that cct Fuling, with the ccs in the tail. I wonder if the tube guys have tried it? I think -5v should be enough for the ccs.

 

[Fuling]

So, CT chokes for gain stages and separate airgapped chokes for output followers?

Of course it would be easier to design a symmetric PSU, but if we can cut down the idle dissipation with 60W by going from -15 to
-5V I´d definitly do it!

 

[gnomus]

I understand the the choke symbols on the outputs. But, what is the little box at the other end?

 

[eLarson]

The constant current source? Or the resistors (euro-style symbol)?

Erik

 

[Fuling]

It´s supposed to be a constant current source

 

[gnomus]

Thanks. Why not a choke threre too? One problem with SOZ is the large heat sinking. The chokes could reduce size and the work of drilling and tapping bias resistors holes.

How much will the chokes improve efficiency?

I have read posts that bias resistors sounded better than an active current source like mosfets in SOZ. I wonder how the chokes will influence the sound.

 

[Fuling]

Why not a choke threre too

I have thought about it, and it may work fine. It won´t regulate the bias in the same way though.

How much will the chokes improve efficiency?

Allmost all of the power that is dissipated in the resistors can be spared, so the difference should be quite big.