advances in tube amp design?

[Steve Eddy]

Originally posted by SY
I had no idea that JC had gone the woo-woo pricing route. Sorry.

I guess you're just not privy to the Divine Secrets of the Woo-Woo Priesthood.

se

[PRR]

> winding physics is now way better.....but the iron hasn't changed........perhaps more exotic mixes.

Winding design was WELL understood in the 1930s and 1940s. Winding computations are summarized in RDH 3rd and derived in great detail in RDM 4th.

The actual winding is always limited by the wires and insulators you can get and what your winder wants to do. Poly varnish can be somewhat thinner and a LOT stronger than the old enamel. Fishpaper still has uses in hi-Z windings, but Kapton and Teflon tapes give better insulation in less space. Winding precision.... people still wind by hand, scramble or "neat", but even the neat wasn't too precise until the recent Swiss precision winders. Jensen used a very laborious quadfilar process until they got the Swiss machine, now they can do better by just laying the wires RIGHT where they need to be.

Iron has improved markedly in saturation, in both low-level and high-power flavors. I'm old, and when I saw the Gauss levels used in today's power transformers I was shocked. Permalloy of today will carry 6dB more than the old days.

Of course this is another case where better numbers do not assure better sound. Audio transformers never come close to saturation. Good audio iron almost has to be different from power iron: they work on opposite sides of the knee.

And ferrite used to be unthinkable, but now there are "ferrites" that will carry audio bass. (They have new fancy names... to me, if it ain't iron or air and is a powder-stuff then I call it Ferrite.)

Iron lamination insulation is another detail. We used to dip stampings in varnish. Now it is common practice to "rust" the iron. Laminations do not need 1,000 GigOhm insulation, eddy current impedance is OTOH an Ohm. Rust is thinner, more iron in the stack, but also rust has magnetic and electric properites which start to show far above the bass limit (i.e. in the critical midrange). Rupert Neve was a connoisseur of lam-rust.

> I don't understand why the ceased highly popular 6550A hasn't been re-introduced

Is the new "Tungsol"-brand 6550 anywhere near as good as the generic 6550s I bought in the 1970s, or just another 6L6 in new bottle?

But in general: making good sturdy tubes is hard work. Special metals and processing. When military and railroad contracts buy many thousand tubes a year, and the buyers track failure rates, the tubes have to be good and you can afford the work needed. When all your sales are pairs/quads to guitarists, who may not know a bad tube if it bites them, can't get a proper warranty, and certainly can't keep failure statistics, then you just build tubes "good enough".

[CharleyW]

That $2850 is the price of the MQ-100 - not the ST-70 rebuild.

Joe hasn't gone woo woo.

I would suspect between Joe's boards and Uncle Ned's transformers, you could do a primo primo ST-70 for no more than 750-800 even without an old Dyna as a donor.

I did it for around $450 with a donor chassis and transformers, and am very happy with the results.

[smoking-amp]

"I recall (a little dimly now) that the Audio Aero had an unusual topology, that they claimed someone had tried to do in the 1950s, and failed. Maybe wrong, but I think it was this amp that operates SET up to a certain watts, then switches automatically into pentode mode. " rick57


The "Capitloe / T.R.A.C." design appears to be a P-P triode class A / P-P pentode class AB implimentation of the Yundt composite amplifier. Published in the 80s.

for an explanation of composite amplifier operation see:
http://www.diyaudio.com/forums/showt...505#post521505
The important point is that A2 (in diagram) is a current output amplifier, while A1 is a conventional voltage output amplifer. Hence, you will notice A2 implimented as pentode or MosFet (drain outputs) for high output impedance. A2 seeks to minimize current draw from A1 by the load.

They also have a P-P triode class A / MosFet class AB design, "Capitole TransTrac". This has the Mosfet amp (A2 in my diagram) connected to the secondary side of the transformer instead since it can handle the output current directly.

Marketing hype describes these designs as a "little local feedback" when in fact this topology is the premier case of elegant NFB. Nothing wrong with that however. This topology is the ultimate approach to low distortion impedance conversion, allowing a small class A amplifier to perform exquisitely into a light load, while producing high power output. Even the commercial solid state amplifiers haven't caught on to Yundt's topology yet.

Don

[rick57]

SY:
Were you (“6DJ8 cascode diff amp with CCS and active regulation of the rails”) referring to the MQ-100 or the ST-70 rebuild?

There are many many (tube amp) fish in the sea ~ I just need to work out which one (in the confusing array) to fry.

At say 15 w, just for mids & bass, I’m leaning to SE, perhaps www.audiodesignguide.com/my/572se.html.
Need to be careful with the volts, but what do people think of the design overall.
(Not that it matters but I’ve wondered for years if Andrea is an Italian male or female name. He/ she certainly has been productive).

Could someone explain what exactly is meant by a “fully transformer coupled system”. Coupled from x to y, to achieve z?

smoking-amp

Do you know if anyone get the Yundt composite amp working at the time? It seems a very good idea, for more work and parts.

[SY]

The Glass Audio rebuild of the ST-70.

[audiousername]

Fixing the link to the Andrea Ciuffoli SV572-3 SE amp:
http://www.audiodesignguide.com/my/572se.html

Originally posted by rick57
At say 15 w, just for mids & bass, I’m leaning to SE, perhaps www.audiodesignguide.com/my/572se.html.
Need to be careful with the volts, but what do people think of the design overall.

I think that the SV572-3 might be difficult to get considering Svetlana USA went belly-up and I don't think the the new owners still distribute them.

If you're thinking of an SE amp with a transmitter triode, something along the lines of a SE GM70 or SE triode-strapped 813 might be better. The valves likely will be cheaper, too.

Originally posted by rick57
Could someone explain what exactly is meant by a “fully transformer coupled system”. Coupled from x to y, to achieve z?

The Ciufolli design is not fully transformer coupled:

RC coupling exists between the 6SN7 (input valve) and 6BX7 (driver valve)

Most all valve amps have a transformer coupled output stage because valves are intrinsically high impedance devices (high V, low I), and the transformer allows them to couple to speakers (which are low impedance). Some OTL (output transformer less) valve amps exist, but they are wacky in the extreme...

There are four main benefits of interstage transformer coupling:

• The inductive load allows the driver valve to swing above the voltage rail, thus increasing the amount of voltage it can swing;
• It allows step-up/step-down ratios;
• A high impedance load is presented to the driver valve, thus increasing gain and reducing distortion (in triodes, anyway)
• The secondary of the transformer presents a low impedance to DC, allowing the output valve to stray into Class *2 operation on overload without causing 'blocking' distortion.

Well, that's how I understand it, anyway.

The disadvantages include cost (a very major disadvantage), and the phase shifts that occur across interstage transformers which are more of a problem when designing a global feedback amplifier.

[fdegrove]

Hi,

Some OTL (output transformer less) valve amps exist, but they are wacky in the extreme...

Wacky?

Nah, they're basically a wish to do away with all the conflicting demands of a good OPT and introducing their own set of problems by the same token.
For ordinary, medium to lowish efficiency speaker they're usually fine and among the best sounding amps of the PP variety when properly executed and interfaced.

Looking back and given that OPTs and chokes have greatly improved over the past twenty years or so, I could probably be quite happy without my OTL amps provided I'd change the speakers for more efficient, probably less linear units....

But OTLs whacky?
No, they can be absolutely fabulous despite their shortcomings....
Moreover, they're one kind of amp that's hardly been developed at all...Still mucho potential untapped there.

Cheers,

[audiousername]

Sorry, Frank. I should have said "Unusual"

[Brett]

Originally posted by audiousername
[*]It allows step-up/step-down ratios;[*]A high impedance load is presented to the driver valve, thus increasing gain and reducing distortion (in triodes, anyway)

The load on the driver goes down if there is a step-up.