4+4-set-stereo-2a3
    
    
    
    4 WATT + 4 WATT SINGLE ENDED 
           TRIODE STEREO AMP.
            This stereo amp was made to power
      horn speakers. I was able to get 4W from each SET channel. 
      Picture 1. 
      
      The large OPT behind the 2A3 are Hammond 1627SEA OPT for 2.5k :
      4,8,16 ohms. 
      I found they gave exemplary technical performance and the amps
      gave the best 4WI have ever heard 
      from a small tube even with my own low sensitivity speakers of
      about 88dB/W/M. 
      With horn loaded speakers with sensitivity at least 100dB/W/M,
      there is no need for more than 4W.
      The two "transformers" nearest the output terminals on left side
      of picture are a Hammond choke and 
      Hammond power transformer for HT for B+ and 6.3V for input triode
      heaters. There is another transformer 
      under the chassis for the 2.5Vac heaters for the 2A3s. 
      
      Picture 2.
      
      
      I had a schematic which I had drawn up neatly somewhere but can't
      find it now. Design your own! Nobody 
      would want to build my idea anyway, everyone wants their way, not
      mine, even though mine is very simple.
      Each channel has 1/2 a 12AU7 input triode feeding a paralleled
      6SN7 driver for the 2A3 with approx 10dB 
      global NFB from output to 12AU7 cathode to reduce Rout from 2r5
      with zero global NFB to 1r0 when using 
      the 8r0 output tap connection. Nothing is technically
      extraordinary, but there is beautiful sound with horns or 
      sensitive speakers such as Tannoy dual concentrics.
      
      Someone asked me "Don't you reckon those OPTs are a bit too big?"
      "Nah, not for the big sound from the tubes...." I replied.
      
      When in doubt, ALWAYS buy bigger iron than you need; size matters.
      
      The chassis is made as follows :- A length of 50mm x 25mm x 3mm
      wall thickness anodized aluminium 
      channel is mitre cut at the corners with hack saw, then filed
      carefully to make mitre joins nearly invisible. 
      I forget overall plan size, but about 500mm long x 250mm wide.
      Accuracy of the cut and joined corners 
      is important, and the 4 sides must be tried together while clamped
      down onto a template and holding blocks 
      on a bench so that when all 4 sides are together, the two diagonal
      measurements are each equal, within 
      1mm. The chassis top and bottom plates are 1.6mm thick natural
      aluminium and made to sit over top and 
      bottoms of channels but kept back 3mm from channel edges. All
      sharp exposed plate edges were filed 
      neatly and sanded with 400 grit paper. Top plate was kept free of
      scratches while drilling holes for sockets, 
      ventilation holes around power tubes, bias pots, bolts etc. The chassis parts were all well held
      together 
      with much masking tape while drilling machine screw "pilot" holes
      through top and bottom plates with 
      3mm drill for 4mm machine screws, and at not more than about 120mm
      apart. Always use too many screws,
      not too few! 
      The tape was removed from plates only, but left around outside of
      channels so the relative positions of 
      screw holes are not disturbed. 
      
      Pilot holes for screws were tapped for machine screws, plates were
      "linished" with 600 wet and dry sand 
      paper using turps as lubricant, and keeping paper washed clean,
      while sanding in ONE direction only. 
      After all holes were drilled and de-burred and or linished and
      looking good, screws were all tried to check 
      positions. Tape around the rectangle of channel was removed, and
      when I was happy there were no 
      horrid gaps and all looked well, the bottom plate only was
      removed, and 4 lengths of 25mm x 25mm 
      Alum angle 40mm long were glued into each internal channel corner
      using Selleys 401 industrial strength 
      silicone. Then the bottom plate was screwed back on using 4
      screws, and chassis left to cure a day 
      before the next lot of work. The angles in internal corners will
      hold the channels together if both top and 
      bottom plates are ever removed, but while the top and bottom
      plates remain fixed, the chassis is strong.
      
      You NEED the channel wall thickness to be 3mm to prevent screws
      stripping threads and to give the 
      chassis enough strength to resist bending if dropped with heavy
      transformers bolted to it. 
      
      This amount of metal work is going to challenge those who have
      never done anything much with any 
      tools, and have no tools or workshop, and or any patience. The
      above is how I did the chassis in the 
      pictures. It looked very neat, and just like the work of the
      dedicated expert. 
      
      An easier way to make a chassis :-
      Fig 1.
      
      My customer who ordered me to make the amp placed the Hammond
      trannies and 2A3s on my bench 
      and bribed me suitably to use them. This was the rare occasion
      where I have not wound my own 
      OPTs and PT and choke, but it all worked out fine.
      
      To Power amps directory.
        
        To Index Page.