... Interesting there's only one grid leak resistor and the Presence network I've not seen before?
This circuit is really a detour from the intent of your thread (and uses a 12AT7/ECC81 rather than a 12AU7/ECC82). But...
The single 1MΩ "grid leak" is incorporated the way it is because the "fixed biased" by virtue of the d.c. coupling from the prior stage cathode follower. There is no cathode resistor for biasing, whose "ground end" (top of the tail resistor) is where you'd normally return your grid leak.
- At the end of the day, that grid leak resistor is a voltage reference for the grid.
- In a self-bias long-tail, returning the grid leak to the bottom of the self-bias resistor ensures the correct grid-to-cathode voltage to bias the tube.
- Since there is no self-bias resistor in this circuit (d.c. coupled), the grid resistor is
not needed for the
input grid (bias is set when tube current drops a voltage across the tail, placing the cathode at the correct design bias compared to the fixed grid voltage from the cathode follower output).
- But you have no voltage reference for the non-driven grid; the 1MΩ from grid to grid
establishes the same voltage for the non-driven grid, as there is no current & no voltage drop across that resistor.
The 1MΩ looks to the circuit the same as if this was a self-bias long-tail with 500kΩ grid leaks from each grid in the typical arrangement. IOW, there is a virtual ground at the midpoint of the grid-to-grid resistor.
This grid-to-grid form of a grid-leak is really only relevant for the direct-coupled input to a differential or long-tail type circuit. Which explains why you likely haven't seen it much.
On to the feedback circuit... This is simply a different approach to tonally tweaking the feedback of an amp.
- The 10kΩ from the speaker to the 2.2kΩ shunt feedback resistor is your basic feedback loop, setting overall gain reduction.
- In parallel with the 2.2kΩ is a 10nF cap in series with a 270Ω resistor; the loop appears to be 10kΩ:2.2kΩ for low frequencies (lot of feedback), but 10kΩ:240Ω (little feedback) for very-high highs. The network appears to start reducing feedback above ~750Hz, and probably gradually rises to counteract speaker/OT treble roll-off.
- Possibly because the previous 2 items were the first to be designed and in order to avoid upsetting their effect, the Presence circuit is implemented in an unusual manner. Moving the wiper to the right side of the Presence pot connects the 47nF essentially to ground across the other network above, greatly reducing feedback to highs & upper mids, while moving the wiper to the left side isolates that 47nF cap and connects a 1nF cap from a plate output to ground (darkening sound).
So while unusual, it does all the things you'd ask of a Presence control, plus with more apparent power given the ability to roll-off highs before they even go into the cathode follower (which has its own bright network at its input).
Topic: 12AU7 LTPI help (Read 13791 times)