My best sound amplifier of these last 20 years

[audiodesign]

see on:

http://www.diyaudio.com/forums/tube...-amplifier-only-3-components-signal-path.html

 

[WuYit]

Hi,
congratulations, you have good taste in sound.

Everything is in the signal path, even the air.

 

[djk]

http://www.lundahl.se/pdfs/datash/1671.pdf

Is the connection Alt.V ?

 

[audiodesign]

Please check the transformer connections in the image at the end of project webpage

 

[audiodesign]

The web page of this project has been update with more info.

Hybrid Amplifier by Andrea Ciuffoli

After 3 different listen sections on the prototype with different friends I can confirm that this amplifier is superior to my 813/GM70 SE, Hybrid 2010, PP2010 and New Class D.

The propagation speed of sound is terrible and attracts and involves everyone who hears.

The level of details is extreme and never is present a listening fatigue.

It is very difficult return to ear any other amplifier after these listen tests.

I suggest it to any DIY in any version, with 5842 or 6C45, with CLR power supply or simple and with any type of mosfet suggested.

You can build it also with surplus components and the unique "big" cost will be the interstage transformers Lundahl LL1671.

I will continue to develop this project in this year and I am selling all the other amplifiers at low cost.

 

[Lee Knatta]

Why are there two resistors connected "antiparallel"???
Do you believe that resistors have a polarity???
Or is it just for increased wattage and the polarity mark not to taken serious?

 

[Lazy Cat]

Inductance cancellation for sure.

 

[djk]

Sometimes non-inductive resistors may be hard to find, or expensive.

 

[Bonsai]

Nice to see your efforts up on the web Andrea. While not everyone would do things the way you do, that really is not an issue. As long as it sounds good to you, and you get pleasure out of building it and writing about it, that's all that matters.

Good luck!

 

[audiodesign]

Yes, the unique neutral resistors are Caddock MK132 and some similar Vishay.

If you use normal 1% metal resistors in the parallel connection and with inverted phase these combination give a result near to the Caddock MK132.

http://www.audiodesignguide.com/Hybrid2011/Conn1a.gif

 

[Lee Knatta]

Sounds interesting. Can you explain how exactly the cancellation effect works?

 

[audiodesign]

Simple two choke in parallel give 1/4 of their inductance value so if the metal resistor have some parasite inductance this will be reduced.

In theory the opposite phase can cancel some electromagnetic field.

Many years ago the proprietary of a famous company told me this idea and he have compare it to Caddock.

 

[AKSA]

Andrea,

Clever design, hats off.......

Only if you try new and strange ideas will amps sound different, but it takes a special kind of engineer to do this.

Clever biasing system, very close to fixed bias on tubes!

Hugh

 

[atx]

Simple two choke in parallel give 1/4 of their inductance value

??
Should not it be 1/2 of the value? i.e Xt= X1*X2/(x1+X2) ??

 

[Telstar]

Can you do a square wave test? I smell some ringing.

 

[audiodesign]

Here the first info about the final project

Amplifier End by Andrea Ciuffoli

 

[willysan]

I'm not so sure about the "antiparallel" resistors. The magnetic coupling between resistors is very small because there is no magnetic core. Truly, the equivalent inductance would reduce as L/N (L is the inductance of a single resistor and N the number of resistor), but soldering them along opposite directions isn't important. I also doubt whether parasitic inductance could be important in resistor of some hundreds ohms

 

[audiodesign]

Willysan,
I am agree with you about the very low magnetic coupling between resistors and the low parasitic inductance in low value resistors but in my experience the sound performances with this method are the same of my reference: Caddock MK132.

 

[audiodesign]

av-trouvaille,
You can use more mosfet in paralel to increase the output current so the power power but a single device can sound better.
Use one separated gate resistor for each mosfet.

 

[digits]

Something odd I saw about the PSU, you simed putting a small C as first then L then larger C... Go look what happens when you swap the capacitor banks to have large C value first, also look up some actual ESL values to plug in, from capacitor datasheets.