Fender Style Reverb Grid Leak Bypass Cap

[stratomaster]

At some point in the Silverface era Fender started bypassing the 220k grid leak resistor on the reverb recovery triode with a 0.0022uF cap.  The usual justification is that this helped with noise from generally poor lead dress.  Looking at the corner frequency of just that bypass network (ignoring tank output impedance and grid impedance and g/k and g/a capacitance) this is a severe high cutting of desired musical content, something like 330Hz if my math is right.  This mirrors the input low cut 500pf/1M network. 

It seems this is responsible for the reduced overall gain and "sparkle" of the Silverface reverb tone. 

Quite separately, I've played around with the low and high cuts on spring reverb modellers like studio effect plug ins and Line 6 effects units.  I found that replicating the Fender low cut of around 330Hz and dialing the high cut to 10X at 3.3kHz gives a nice musical reverb that doesn't swamp the dry signal and allows the mix to be dialed up a good deal. 

I'm therefore proposing that the less tonally obtrusive noise fix (along with neatening the reverb circuit lead dress) is a 220pF grid leak bypass cap or a 470pF plate bypass cap (or both for a steeper roll off).  This will leave the meat of the reverb intact but roll off the hash and "noise".  It also has the side effect of accommodating the higher effect levels that this circuit produces without having to resort to dwell pots or tweaks to the fixed voltage divider network at the circuit output. 

I know a passive tone control will accomplish something similar, but I'm thinking just along the lines of adjusting the recommendation of 0.0022uF across the 220k grid leak for those looking to tame the reverb hash. 

Bottom line: If you're already adding a cap try a 220pF instead of the usual 0.0022uF.

[stratomaster]

On the other hand, if we ignore the grid leak as sufficiently large, then the low pass filter calculation is dominated by the tank output impedance and the cap to ground.  This would put the corner frequency around 32kHz and the addition of the cap would be basically inaudible.

So which simplification is more correct?  Looking at the grid leak and bypass cap alone, or ignoring the effect of the grid leak and looking mainly at tank impedance and the cap to ground?

[PRR]

> ignoring tank output impedance

Tank inductance dominates this node. The cap sets a resonance in upper midrange.

Don't like it? Change the cap, it's cheap.

The "12,000 ohm" coil is like 1.7 Henry.
1.7H with 0.002uFd  = 2700Hz
1.7H with 220pFd  = 8200Hz
At the higher 8kHz tuning the reactance is OTOO 14K so the 220K resistor has very little effect.

Should a reverb "sparkle"? Very much a matter of musical taste and application. In real rooms as much as amp tanks.

[tdvt]

I just happened to be rebuilding the board of a '79 Pro Reverb yesterday, that originally had the later circuit changes you mention.


There are any number of small changes Fender made in that area, beyond the ones you mentioned, but I am not clear as to your end-goal. A restoration to that specific era/spec or addressing a particular noise issue?


Most would (me included) rewire things to the blackface circuit, removing the "distortion" circuit in the process & generally tidying things up there.

[stratomaster]

The goal is partly to understand better what is happening, why the cap was added, what it does and how it interacts with the rest of the circuit.  From there make informed choices about how to better accomplish the goals of the cap, propose alternatives, and think about other potential programs that can be simultaneously addressed. 

Put something out there and see how others with deeper understanding and experience poke holes in it.

With the info posted by PRR above, it seems like keeping the 0.0022uF cap and adding a 470pF across the plate resistor will give an upper mid lift and treble roll off starting around 3.3kHz to help the reverb from totally swamping the dry.  I would have never looked at the filter as an LC otherwise.