Hybrid Class-D

[johnrtd]

I've seen the thread with the hybrid (tube + class-D) experiences. We've been busy last year with the same kind of configuration, but we use the old Jean Hiraga ideas for a phase shifter. The result being
http://www.audio.nl/downloads/MS-TUBE-aanvulling.doc

It's in Dutch but the figures are clear.

[revintage]

Hey Johnrtd,

Where´s the link ?

[johnrtd]

The link is there now. Note that the differential inputs for the UcD400 are NOT equal.

The presented circuit has a noise and distortion figure of around 0,005%!

John

[revintage]

Hi John,

You are partly wrong about the input. When driven differentially the input is fully symmetrical with a Zin of 3,6k. But driven separately/unbalanced - is 1,8k and -is 10k.

Also there is something wrong with the linked file, have tried to download and open it in two different computers with no luck!

[johnrtd]

Dear Lars

We discussed that differential input with the designer of the UcD's, Bruno Potzeys. And we also meaured those impedances by ourselves.

Those impedances are NOT a problem when driven from a relatively low source impedance of, say <10 Ohm (such as op amps).

With the SRPP the output impedance is around 1,5 kOhm but, as found by Jacco Dekkers, the distortion from the SRPP has a relationship with the load impedance giving (in our configuration!) the best result with a load of around 3 kOhm. So we combined the input impedance of the UcD with resistors to get the best results (lowest distortion). And I must say it works fine.

Quite a number of these kits have been built by amateurs (and technicians) and all stated that it performs with the "character" of a very good tube amp combined with a very low output impedance.

John

[revintage]

John,

I also discussed this with Bruno as we are thinking of producing a hybrid but with a completely different(and truly differential ) topology than this.

He confirmed 3,6k Zin when driven diffferential.

When you measure the impedances separately they are different (1,8k/10k) but when you measure from+ to - it will still be 3,6k.

In the drive you use they will of course be different.

I also believe in that the concept with tube driver will be nicesounding.

[Koifarm]

John, i followed your and Jacco's project. But i think Lars is right about the 2x 1k8 impedance ( 3K6 ). You know i do not like measurements but i use the balanced input now for 2 years with 2 SRPP identical tubestages optimized for 1k8 impedance driving unbuffered Hypex UCD180st. And i can not hear distortion at high levels, so it must better then 0.5 %. My latest project with a unbalanced tube driven UCD400oem works also great.

Saturday my hybrid designs will be object for testing and listening at the "Roehren und Hoeren" meeting in Germany. I will see what the germans think of the design.