DIY Hybrid tube/ss amplifier help

[famousken]

Hello everyone, I'm in the design process of building a 100W/ch audio amplifier. I would like it to sound as good as possible (willing to sacrifice power output for sound quality) and am looking for some design pointers. The output stage and driver stage will be solid state using 2sd845/2sb755 complimentary pair for the output transistors (120v,12A,Hfe=15) and c2235/a965 for the drivers (complimentary) (120V,800mA,Hfe=180). From that point I would like to use tubes as the required drive current is within realistic ranges (Aprox. 60mA) The tubes I was considering using are 5963 dual triodes simply because I have a bunch of them and they seem to have the right specs http://www.tubebooks.org/tubedata/HB-3/Receiving-Type_Industrial_Tubes/5963.PDF.

Where I am having trouble is in two areas, the first being the biasing of the driver transistors for class AB operation. Should I bias at the driver transistors, after the tube amplifier stage? I am worried about stability problems at power on/off as the tubes need to warm up. Also, any suggestions as to coupling the tubes to the driver transistors? I'm thinking that I will need to use pure class A, cathode follower, biased to 0V relative to the power supply rails (+65V-0- -65V) and this would mandate capacitor coupling between stages and biasing of the final stage at the driver transistors (I think) however I would like to use direct coupling in all stages if possible, so if there is a way around this please let me know!

Here are the datasheets for the driver and output transistors:
http://www.datasheetcatalog.com/datasheets_pdf/2/S/B/7/2SB755.shtml
http://www.datasheetcatalog.com/datasheets_pdf/2/S/C/2/2SC2235.shtml

 

[sophiecentaur]

What's so attractive about thermionic valves? Is there ever any good reason for choosing them, these days - apart from a 'retro' thing?

 

[sophiecentaur]

Oh yes. They are glass and shiny and they glow inside. But they are not particularly good as amplifiers - not as linear amps, anyway.
Tube amps have a 'particular sound' which must mean they contaminate the original - or you wouldn't know they'd been used. It's a bit like Vinyl records - people reckon they like the clicks and hiss and they even like the distortion due to the original analogue magnetic recording.
It strikes me a bit like choosing to climb Mount Everest wearing steel shod boots and a tweed jacket.

But, as an ex broadcast Engineer, I still take my hat off to the the hundreds of kW that a good valve transmitter can produce.
Would it be a guitar amp that you're making?

 

[Bob S]

Also, any suggestions as to coupling the tubes to the driver transistors? I'm thinking that I will need to use pure class A, cathode follower, biased to 0V relative to the power supply rails (+65V-0- -65V) and this would mandate capacitor coupling between stages and biasing of the final stage at the driver transistors (I think) however I would like to use direct coupling in all stages if possible, so if there is a way around this please let me know!

When I worked at HP in the mid 1950's, everything was vacuum tube. For dc coupling in amplifiers, we would use a string of NE-2 neon bulbs (with a small dc current through them to keep them lit) to couple a plate circuit from one tube to the grid of the next. Put a capacitor in parallel with the NE-2s to minimize the white noise from the NE-2s. If ± 65 volts is not enough room for a NE-2, then use a zener. See photo of NE-2 in

http://en.wikipedia.org/wiki/Neon_lamp

Bob S

 

[sophiecentaur]

I guess you could go the whole hog and use transformer coupling to the output stage.

 

[famousken]

Why am I using tubes? Because they sound good, I know their noise floor is higher than solid state components, I know they have more distortion than ss, but I like the way they sound. There actually is a bit of truth to tubes sounding better though because when they distort, they do so softly and the distortion produced is mainly 2nd harmonics, which are present in music anyways; when a transistor distorts, it does so abruptly and the distortion is mostly higher-order harmonics, which sound abrasive and not related to the music at all.

Bob, thank you, I never would have thought of using neons for biasing, although I have seen leds used for ss gear. I am still a little foggy about where to apply the biasing as this is actually new territory for me (most of the stuff I do is microcontrollers and smps stuff) and the output transistors are quite expensive so I don't want to blow them up.

Sophiecentaur, I have thought of that actually, but am looking for alternatives so as not to add the frequency peaks and rolloffs at very high and very low frequencies, so I'm looking for a safe (for my components) way to couple the two stages without a transformer or capacitor.