10/26/2026
On the workbench today: Yamaha T100 Tube amp head
I have a Yamaha T100 on the bench—the 100W Mike Soldano-designed head from the early ’90s. While sorting out other issues, I noticed the circuit around V4 and the reverb recovery didn't match the factory schematic. After flipping the board, I found cut traces, jumper wires, and an intentional re-architecture of V4 and the reverb routing.
Here is a breakdown of what the factory circuit originally did, what the modifier changed, and the "why" behind it.
In the factory T50/T100 layout:
* V4 is wired in parallel (both plates tied, both grids tied), serving strictly as an active FX Loop Return recovery stage.
* Its output (Node D) is sent straight into V5 pin 2—the Clean channel’s cathode follower. From there, it passes through the Clean tone stack and Clean Master volume.
* The reverb return from V6 runs passively through two level trimmers and optocouplers (Vactrols 53 and 54) to toggle between channels, relying on passive mixing right into the masters.
This factory design has two well-known quirks among T100 owners: the FX loop return only feeds into the Clean channel's front end, and the reverb circuit suffers from crosstalk, bleed, and weak wet recovery.
What Was Done: Splitting V4
A previous tech completely decoupled V4's two halves to repurpose them for two distinct jobs. Two factory PCB traces were cut, and point-to-point jumpers were run under the board between socket legs:
1. V4A (Pins 1, 2, 3): FX Return Re-routed to the Overdrive Channel
* The Mod: Pin 2 remains the input from the FX Return jack. Trace cuts isolated plate pin 1, and an under-board jumper was run from V4 pin 1 directly to V5 pin 7 (the grid of the Overdrive cathode follower).
* The "Why": In the stock circuit, V5 pin 7 is driven solely by the Overdrive preamp output (V3B plate). By connecting V4A's plate to V5 pin 7, the modifier turns the Overdrive cathode follower into an active summing node. Instead of being trapped on the Clean side, wet rack effects or external preamps can now feed straight into the Overdrive channel's tone stack and master volume.
2. V4B (Pins 6, 7, 8): Dedicated Active Reverb Recovery Buffer
* The Mod: The second triode half was decoupled to run independently:
* Cathode (Pin 8): Fitted with an unbypassed 6.8k 1/2W resistor to ground. With this cold bias, pin 8 sits around 3V to 4V DC, while pin 6 (plate) idles high on the B+ rail because of the reduced current draw.
* Grid (Pin 7): Fed directly by the photocells of the channel-switching Vactrols (53 and 54).
* Plate (Pin 6): Runs out through an added 4.7 nF capacitor that lands directly into the 22 nF input cap of V7 pin 2 (the Phase Inverter grid).
* V6 Feedback: A 100k resistor was added between the grid and plate of the V6 reverb tube.
* The "Why":
* High Headroom / Clean Recovery: The unbypassed 6.8k 1/2W cathode resistor biases this triode cold. This significantly knocks down the stage's gain while widening the input headroom, ensuring the returning wet reverb signal will not clip or add unwanted grit.
* Independent Injection: By routing the buffered reverb signal through the 4.7 nF cap directly into the Phase Inverter (V7 pin 2), the wet reverb bypasses the stock channel-switching Vactrols (35 and 36) entirely.
* Voicing: The 4.7 nF series capacitor acts as a tailored high-pass filter ahead of the 22 nF PI coupling cap. It deliberately strips muddy sub-bass rumble out of the spring tank decay before it can flub out the 6L6 power section.
* Stability: The 100k plate-to-grid local feedback loop on V6 tames the drive stage's excess gain, killing off the capacitive buzz and high-frequency ringing that often plagues these spring drive circuits.
* Replacing the 4.7 nF Cap:
The capacitor used here is a RIFA PME 271 series metalized paper safety/suppression capacitor (specifically marked PME 271 Y 447)—widely known across electronics and amp repair benches as a "RIFA smoke bomb." These clear epoxy-encapsulated paper caps develop micro-fractures over decades, absorb atmospheric moisture, and fail catastrophically in a cloud of acrid smoke.
Worse yet, in this circuit it is being used as a high-voltage audio coupling cap, sitting directly between the 300V+ plate of V4B and the sensitive grid of the phase inverter at V7 pin 2. Any DC leakage across that deteriorating paper dielectric will put positive DC directly onto the PI grid, destroying the power section's bias and potentially taking out tubes or transformers. It is an immediate discard-and-replace item. It will be swapped out for a fresh, high-reliability 630V Vishay MKT1813 film cap before this chassis ever leaves the bench.
Summary of the Modification
Rather than living with an awkward loop and a finicky reverb circuit, whoever worked on this amp separated the parallel triode to solve both issues in one move:
1. FX Return: Injects straight into the Overdrive tone stack/master via V5 pin 7.
2. Reverb: Gets its own buffered, voiced recovery stage via V4B that dumps directly into the power section at V7.
It is a clever, well-thought-out piece of custom bench engineering—even if it took some head-scratching over the schematic, cut traces, and an infamous paper cap to piece the whole story together.
It's waiting on a fresh set of filter caps and diodes, and then I can find out how it sounds.
Stay tuned.
