Why no CCS in PP output stage?

[leadbelly]

Why is CCS so rarely used in PP output stages?

Planet10 gave me this reference:

EL84 at AA

I can't find much else that uses it. Is it a question of specific topology required to implement it that I can't discern, or builder opinion/preference?

 

[SY]

Using a CCS output stage as its own phase splitter is fine as long as you don't mind the cost, inefficiency, and distortion. The RDH discusses this topology and pretty much dismisses it. And possibly for good reason.

 

[EC8010]

Also, a CCS in the output stage means that the stage folds catastrophically when asked to leave Class A.

 

[leadbelly]

Using a CCS output stage as its own phase splitter is fine as long as you don't mind the cost, inefficiency, and distortion. The RDH discusses this topology and pretty much dismisses it. And possibly for good reason.

If you mean the section in Chapter 13.3.viii, I don't think that applies to the case where the input section is the phase splitter. Is there a more detailed reference somewhere else?

Oh, and I did find another example:

http://home.pacifier.com/~gpimm/1624.gif

 

[planet10]

Allen Wright claims to have been the 1st (he knows of) to use a CCS on the outputs -- he says so cautiously given how often something new is actually done. He does cite its use in a German atomic reactor.

It was at the last VSAC he started talking about it and showed his amp -- this amp set a few people on their ears. Since then the amp has won product of the year, 2 years in a row at enjoy the music.

It requires the amp run in class A (ie probably 4 watts max for your EL84s) and clipping is abrupt -- more like a SS amp.

The AA stuff is all spawned from that VSAC lecture, and has had active input from Allen.

I'm looking forward to his pending power Amplifier cookbook where he will discuss it in detail.

dave

 

[leadbelly]

Thanks Dave!

BTW, my latest idea on how to build my power stage was to have a triode/UL switch, and in the UL position have partial feedback with a 100k resistor (don't want to use an OPT connection since the input stage is in a different box). I haven't yet figured out if the 70 mA/tube CCS setting in the AA amp jives with that. In Jones' book he glosses over how he biases the EL84, and that's what I need spelled out in crayon!

 

[SY]

If you mean the section in Chapter 13.3.viii, I don't think that applies to the case where the input section is the phase splitter. Is there a more detailed reference somewhere else?

In the linked John Swenson schematic, the input stage is not a phase splitter, it's a plain vanilla single-ended voltage amp.

 

[audiopro]

Allen Wright claims to have been the 1st (he knows of) to use a CCS on the outputs

Actually, it was Bill Beard who was first ( in his P30 power amp) to use a CCS in the output stage. I believe he used LM317's or 7812's (I'm not sure which) on the cathodes of each EL84.
By the way that was one fine sounding amp. The only bug was that if a tube went bad you had to replace the regulator as well almost every time.

Dan
P.S. I'm not trying to knock Allen

 

[planet10]

A lot of Allen's stuff use a cascode differential stage in them (which he pinched from HP test gear). His comment when i sent him a map from an early 50s mag that had a 7F7 cascode differential stage driving PP 807s (ie very similar to his PP1c) was that he well knew that very little we do with tubes is actually new.... hence his cautious statement :^)

Allen is at least responsible for the recent interest in this.

dave

 

[planet10]

PS: any references to the Bill Beard amp?

dave

 

[reinhard]

Bill Beard is using one current source for each tube, bridged by a big capacitor, that is a completely other thing, more a bias regulation, nothing differential.
Reinhard

 

[planet10]

In the linked John Swenson schematic, the input stage is not a phase splitter, it's a plain vanilla single-ended voltage amp.

The V1 version of swenson's amp used a trafo phase splitter. Tony's "preamp" is already a differential stage with balanced output -- the idea was to use the output of that to directly drive the grids of the EL84s.

dave