From: info@turneraudio.com.au   
      
   >>   
   > I recently rebuilt a couple of Quad II's from what was pretty bad shape   
   > with parts missing etc. Everything now works and sounds great and even   
   > looks mostly original... except for the highish B+ Im getting. This is the   
   > new configuration:   
   >   
   > Rectifier tube (5AR4) - 33uF/450v - the original choke - 47uF/450v   
      
   I am not sure you know what you are doing.   
      
   ** The OP knows - his description is of the entire DC supply circuitry.   
      
       The changes are the rectifier and filter cap values.   
      
   Well, from what he was saying,it looked like he didn't know all that could be   
   known, and for which we must allow for.   
      
   If nothing else is changed in an original Quad-II amp, the The use of GZ32   
   or GZ34/5AR4 makes only a slight rise of B+ which is entirely OK and nothing   
   to worry about.   
      
   ** The voltage rises by about 30-40V  so is significant.   
      
   Its about 10% B+ rise, and quite OK in itself. However, with a rise in B+ the   
   Iadc rises a bit so the idle Pda rises, which is not so good, because tubes   
   run hotter than they should. So 220r common cathode Rk would be OK.   
      
   You don't need to do every single thing I usually do to old Quads to get   
   better music from them. The use of a single 180r common cathode Rk is   
   terrible.   
      
   ** Hardly a fair comment seeing as it works so well with the specified   
   output valves ( and most 6L6GCs ) long as they are still in good condition.   
      
   Yes but anyone can build a better amp than the Quad-II, its not a world   
   beater. While KT66 or 6LCGC are new and in good condition and matched, the Idc   
   in each tube will remain balanced but after some time they age differently and   
   become unmatched and    
   there is nothing to prevent the Idc imbalance. Its a major flaw in Quad-II   
   design. If one 6L6GC begins to get a tiny bit gassy, the 680k grid biasing Rg   
   begins to draw current and the grid goes slightly positive, and tube conducts   
   more Iadc, which heats    
   up the tube which causes more grid current and Ia, and more heating. The   
   higher Ia of this tube raises Ek, which tends to turn off the other tube. So   
   the Idc can become unbalanced sooner in Quad-II than in any other better   
   designed amp which has    
   individual cathode biasing networks.   
   The 680k bias R is way too high and should be no more than 120k, but then the   
   piss weak EF86 have a load that's too low to get the wanted voltage gain.   
   Just why Walker used EF86 is unknown when better ideas with say a pair of   
   12AU7 would have been so much better. Money is the root of most amplifier   
   crapitological practices used by so many minimalist amplifier makers of 1950s.    
      
   In Quad-II I always use a 470r plus 470uF from each 6L6 cathode to end of CFB   
   winding which has its CT taken to 0V. This allows slightly worn old tubes to   
   last a lot longer without one tube beginning to glow red while other is too   
   cold, and sound has    
   gone to mud.    
      
   The distortion measurements one is supposed to get with Quad-II are often a   
   complete fiction and once output tubes become unbalanced the mild amount of   
   GNFB fails to reduce it much. Quad-II benefit from being completely gutted,   
   and starting all over    
   again with a better schematic that ensures output tube Ia balance, has lower   
   value grid biasing R, has a lower distortion input driver amp with at least 2   
   x 12AU7, or perhaps 12AU7 input plus 6CG7 LTP driver   
   with CCS in cathode circuit. There must be individual RC cathode biasing and   
   PSU needs total revision. Then the amp works fairly well in class AB with   
   recent made speakers of say 6r0 which have low sensitivity.    
   Quad-II have RLa-a when 16r load is used with OPT strapped for 16r. But 4k0   
   does not give pure class A. For that you must use 32r, so RLa-a becomes 8k0,   
   and each 6L6GC sees a class A load of about 4k0.    
   Many ppl strap Quad-II for 9r, then connect 4r0 speakers, and I have done it   
   myself when testing amps I've modded. They cope because of my mods as long as   
   high levels are not used. having 4r0 with OPT set for 9r, means the RLa-a =   
   1k8, and then the    
   amount of class A is tiny before class AB threshold, so at high levels the   
   Quad-II THD-IMD-noise is bloody atrocious.   
      
   I've never accepted that Quad or any other old fuddy-duddy brand was fabulous   
   or even adequate when I could see so many shortcomings. I refuse to bow down   
   to the past.   
      
   Patrick Turner     
      
      
   ....  Phil    
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)