B+ R/C Filters in Gretch Preamps; Split load?

[W5FH]

Hello-
     I have a question regarding the B+ R/C filters as used in some of the Gretch amplifiers. I have attached the model 6162 schematic as a reference. In most amps the B+ string as it progresses through stages passes through several dropping (smoothing) resistors, each with an associated electrolytic filter capacitor. A typical dropping resistor may be 3.3K to 10K, with a 33 uFD to 47 uFD electrolytic filter cap. I have read Blencowe's treatise on these RC filters and how to calculate the -3 dB cut-off frequency (goal is 0 Hz DC; FC= 1/2piR). In the Gretch amp it appears they flip these component values. They use a much larger smoothing resistor (100K) and a much smaller filter capacitor (0.1 ufd).
    What messes with my understanding here is that the Gretch arrangement appears as a split load plate resistor arrangement with a 0.1 uFD cap dumping the signal between the resistors (220K and 100K plate load) to ground. Can anyone comment on the concept here that Gretch is using as compared to the conventional manner of B+ distribution?

[HotBluePlates]

... model 6162 schematic ... the Gretch arrangement appears as a split load plate resistor arrangement with a 0.1 uFD cap dumping the signal between the resistors (220K and 100K plate load) to ground. Can anyone comment on the concept here that Gretch is using as compared to the conventional manner of B+ distribution?

Do you have a 6162 amp in-hand to verify parts values & voltages?  If "No" then don't put 100% faith in all voltage shown on the schematic, beyond being "in the ballpark."  V2 has a cathode follower after the tremolo Intensity control, whose cathode is directly connected to one half of V4.  They have 2 different voltages shown on the schematic, but must be a single value due to the direct-connection.

Such issues out of the way, what the Gretsch demonstrates is a difference between "DC Load Line" and "AC Load Line" for a special case we don't encounter very often today.

... In most amps the B+ string as it progresses through stages passes through several dropping (smoothing) resistors, each with an associated electrolytic filter capacitor. ... Can anyone comment on the concept here that Gretch is using as compared to the conventional manner of B+ distribution?

Gretsch didn't do it "Fender-style."

It's kinda "Vox-style" in that a small-ish B+ voltage appears to be delivered by the power transformer, probably very near the 350v indicated at the 6973 plates.  There is a 1kΩ resistor to the 2nd filter cap that appears to have 340v present, and feeds the 6973 screens.  From this point, raw B+ voltage is sent multiple places like spokes on a wheel.

   -  Through 27kΩ to a filter cap feeding the paraphase inverter.
   -  Through a 1kΩ to a filter cap feeding the reverb driver transformer.
   -  Through a 4.7kΩ and filter cap to another 4.7kΩ and filter cap to feed all the other preamp stages.

The Gretsch designer apparently needed to avoid a bunch of voltage-dropping in the power supply, so used the "spokes-on-a-wheel" arrangement.  But now they had 1 filter cap to feed 6 different gain stages.  What's worse, in the effects channel the 1st (half of V2), 2nd (half of V4) and 3rd (other half of V4) gain stages all have the same filter cap.

   -  The last of these gain stages ("Channel Mix stage" of V4 after the Tone controls) handles the largest signal level, and so needs the highest plate voltage.  So its plate load resistor connects directly to the filter cap.

   -  The other gain stages don't need as-high a plate voltage, and they need their power supply voltages strongly-isolated from V4's Channel Mix stage.


The Gretsch designer chose to solve this problem by adding the 100kΩ resistor between the shared filter cap and the "rest of the gain stage."

   -  The DC conditions are established based on the voltage at the shared 20µF filter cap, with a "DC Load Line" of 270kΩ + 100kΩ = 370kΩ.

   -  The AC conditions are established by the 270kΩ plate load in parallel with whatever load follows after the gain stage's coupling cap.

   -  Gretsch saves the cost of 1 more filter cap (though to be fair they already have 7 filter caps, probably in 2 expensive multi-section cans).


York's Amplifiers instructs on drawing a loadline where the B+ decoding resistor is assumed to be a significant % of the total resistance back to a source voltage, and is therefore a part of the "DC conditions" (page 40).  We normally think of Fender-style power supplies where the B+ decoupling resistor is a small % of total resistance and so we tend to ignore including it (Merlin seems to ignore the decoupling resistor for loadlines here; I'm not sure about his other publications).

   -  I'd normally argue we don't need to include the B+ decoupling resistor in a loadline, because we have a "large filter cap" on the other end of the plate load resistor.

   -  The Gretsch 6162 is the special case that refutes my reasoning above, because that 0.1µF is only a very-solid bypass for guitar frequencies.  It becomes no-bypass at the bottom of the audio range.  And the 100kΩ B+ decoupling resistor is >25% of the total resistance back to the next "solid bypass" so it has significant influence on moving the AC loadline to a different place on the tube's curves.