From: info@turneraudio.com.au   
      
   Snip.....   
      
   I said...   
   > Amp has wear and tear and looks 20 years old, but it sure ain't from 1965.   
      
   You said...**  Those 3 PCBs are a dead give away.   
      
   Well yes, but it is know as a Fender "Deluxe Reverb" reissue.   
   The full name on schema is "Deluxe Reverb Amplifier". No need to nit pick Phil.   
      
   > This tiny amount of NFB does hardly anything to flatten the   
   > signal at the speaker. The signal response at has a 6dB peak   
   > at 100Hz and +6dB at 2kHz, relative to 1kHz level. So the   
   > Vo merely outlines the speaker Z which is to be expected..   
      
   **  Err   -  no   
    Regular guitar ( and most other) speakers don't double impedance with a one   
   octave change in F above 1kHz.   
      
    It takes 2 octaves, at least.   
      
   OK,at bass F the speaker has a peak in response at around 100Hz and caused by   
   its Fs, due to RESONANT behavior and speaker becomes equivalent to an R damped   
   parallel LC network.    
   The rise in Z above the minimum Z at about 300Hz is due to increasing XL in   
   series with RL of speaker, and the eventual increases in ZL IS 6dB per octave.   
   The amp with such low NFB and with 6V6 in tetrode is a virtual current source   
   because amp Rout is    
   far more ohms than the speaker minimum Z. So hence the current applied to   
   speaker remains fairly constant and because Z changes, so does the speaker   
   voltage, and the response outlines the ZLs approximately.   
   Tetrodes like 6V6 need a lot more GNFB if a flat response was to be expected.   
   But in a Guitar amp flatness = Dull & Boring sound, artistic crap quality,   
   like a view on the sound in black and white and no color.   
      
   So, the Fender amp with its speaker acts like a tone control with useful   
   boost at around 100Hz, and above 1kHz. Nobody complains about it, they like it.   
   The NFB pot I added can increase GNFB to 8dB max which DOES make a change to   
   sound of a tone from a sig gene, and does slightly flatten the response   
   because of reduced Rout. With ZERO GNFB, the sound is kinda sloppy, could be   
   irritating to many ppl, but    
   some may like that. With max FB the sound becomes sort of firm, and maybe   
   liked/disliked, and when driven deep into overload, there's not huge sound   
   change because the signal is a mass of clipped square waves, and twidding the   
   NFB control does nothing    
   much when much overdriven.   
      
   But below overdrive, there is considerable sound change.    
      
   Using more FB increases the clipping threshold of OP stage so that instead of   
   needing 2.5Vrms input to power amp you need about 8Vrms which means slightly   
   more THD is produced by stages upstream of the power amp.     
      
      
   >   
   > Phil implies moving grid stoppers 1k5 at PCB to tube sockets might do   
   > something, because Fender did it in the past, but such a move would do   
   > SFA.   
   > Maybe **increasing** grid stoppers to say 4k7, 6k8, 10k0 *might* have an   
   > effect.   
      
   ** What ever it takes....   
      
   > But what I done loads the 6V6 with 9.4k a-a above about 30kHz, thereby   
   > stopping the oscillations.   
      
   ** A simple zobel at the speaker output would do as well.   
      
   I tried that, various R and C values, but OPT leakage inductance is so high   
   that OP tubes are not loaded enough at high F to stop oscillations.    
   It is much better to have critical damping Zobel networks across each 1/2   
   primary in ALL tube amps where the OPT was designed for LOW bandwidth needed   
   by musicians with electric guitars.    
      
   I've not measured the LL, but many years of fixing crappy amps tells me the   
   Fender OPT must have little interleaving, and it probably has just two primary   
   windings with a central single secondary winding, P-S-P interleaving pattern.   
   Interleaving is generally only used where it needs to be - in good quality   
   hi-fi amps which should have 65kHz BW with a designed resistance load, and be   
   stable unconditionally with no load at all, or with any speaker ever made or   
   with any combination of    
   L, C and R that anyone can devise. To attain this standard for a 22W amp the   
   OPT should weigh twice what is in the Fender and have at least a P-S-P-S-P   
   pattern, with S-P-S-P-S-P-S being best, but mass producing such OPTs makes   
   them cost $10 more, and    
   this would ruin many hi-fi amp makers and they get a marketing BS artist to   
   lie instead about the amp.   
      
   > Possibly this amp could be converted to use EL34,   
      
   ** Sacrilege  !!!   
      
   Why? many ppl like EL34. Musicians are usually Heathens, and follow no God.   
   They spend as many years as their body will take at getting gigs, getting   
   laid, getting pissed, avoiding divorce costs, and enjoying the sound, until   
   their ears begin to play up real bad. So they don't much give a stuff what   
   tubes are used, as long as    
   the amp with its speakers makes the magic in whatever circumstances and venue   
   they find.    
      
    Conversion to use Sovtek 5881s is OK, though.   
      
   Yeah, probably quite OK, biasing is the same, and tubes with the Ea at only   
   400V approx should last very well. Even NOS 6L6 might survive with 300Vdc   
   screen rating. But 6L6GC, 5881, yeah, sure.    
      
    Otherwise, using JJ brand 6V6s is practically mandatory.   
      
    SFA difference between  the two tubes - really.   
      
   Well, not really. 5881 are virtually same as 6L6GC, but have twice the Pda   
   rating than old style 6V6, which I don't recommend in this Fender. But the JJ   
   6V6 are sort of a poor man's 6L6, and have higher Pda rating and from what I   
   have seen they are good    
   versions of the 6V6.   
   The Ra, gm of JJ6V6 and Russian 5881/6L7GC are indeed similar and not many ppl   
   would notice the difference in sound. The 5881 would be able to swing anode   
   voltage slightly lower in Fender and thus produce slightly more Po. But if the   
   speaker was changed    
   to lower Z the 5881 will swing the Ea much lower than 6V6 and you get probably   
   40Watts at clipping, class AB1. But it seems ppl have this devotion to small   
   output tubes because of the sound before overdrive, and because the sound with   
   overdrive is good,    
   but not too loud for some venues.    
      
   But I dare to generalize. I once knew a young bloke in a Dark Metal band. He   
   wanted an extra 12AX7 gain stage in a Marshall, and he wanted the pair of 6550   
   converted from tetrode to triode operation. OK, too easy. When he brought his   
   guitar to try it, I    
   put ear plugs in and ear muffs on, and sound was like a continuous number of   
   747s crashing on my shed. Ah, Such Music! Such tone! such magic!, and I just   
   agreed and smiled when I got paid, which was a wonder.   
   Patrick Turner.   
      
      
   ...   Phil   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)