FAQ
TL;DR: Startup damage and hum come from a surge through the coupling cap; the thread reports “one of two things take place” and 2 outcomes total. Add input clamps and slow the B+ edge. [Elektroda, Charles Johnson, post #21675853]
Why it matters: This FAQ shows how to stop LM4562 inputs from seeing destructive startup transients in cap‑coupled tube/solid‑state hybrids.
Quick Facts
- Symptom: audible hum at power‑on implies C2 surge into the op‑amp input. [Elektroda, Charles Johnson, post #21675853]
- Root cause: fast B+ edge couples through C2; large R6 cannot damp the pulse. [Elektroda, Dean Franks, post #21675857]
- Protection: use a pair of ESD diodes from the cap node to the rails, plus an input series resistor. [Elektroda, Dean Franks, post #21675857]
- Alternative clamp: anti‑parallel diodes to ground at the + input were also suggested. [Elektroda, Charles Johnson, post #21675859]
- Topology note: R7 “serves no useful purpose” in the shown configuration. [Elektroda, richard gabric, post #21675854]
What causes the LM4562 to blow on power‑up here?
A fast B+ turn‑on edge couples through C2 into the non‑inverting input. With R6 large, the pulse is not damped, so the input sees an overvoltage and the op‑amp can fail. Adding a clamp and limiting the pulse current fixes this. [Elektroda, Dean Franks, post #21675857]
Why do I hear an audible hum from C2 at startup?
The hum flags a surge current through the coupling capacitor during B+ application. That transient current injects a pulse into the op‑amp input until the node discharges through R6. Once it settles, the hum disappears. [Elektroda, Charles Johnson, post #21675853]
How do I protect the op‑amp input from the surge?
Clamp the cap’s output node with a pair of ESD diodes to the V+/V− rails and add a small series resistor before the op‑amp input. This limits current and keeps the input within safe limits during the pulse. [Elektroda, Dean Franks, post #21675857]
Where exactly should I place the ESD or clamp diodes?
Connect the ESD diodes from the node after C2 to V+ and to V−. Keep leads short. This shunts the pulse energy away from the input. A quoted tip: “A pair of ESD diodes… would be a good addition.” [Elektroda, Dean Franks, post #21675857]
Do I need a series resistor on the op‑amp input, and will it change gain?
Yes, add a series resistor between C2 and the op‑amp input to limit surge current. It does not change the closed‑loop gain when placed before the non‑inverting input. It just protects during transients. [Elektroda, Dean Franks, post #21675857]
Should I add resistance or inductance in the B+ line to slow the edge?
Yes. Put a resistor or small inductor in series with B+ so C1 forms an RC/LC low‑pass. You can also increase C1. Slowing the edge reduces how much of it C2 couples forward. [Elektroda, Dean Franks, post #21675857]
Is R7 actually doing anything useful in this schematic?
No. As drawn, R7 does not contribute to a valid op‑amp configuration and should be removed or repurposed. Recheck the gain network and biasing. [Elektroda, richard gabric, post #21675854]
What does “B+” mean in this context?
B+ is the high‑voltage supply feeding the tube plate. Switching B+ on creates a fast edge that can capacitively couple into the op‑amp front end through C2. [Elektroda, Dean Franks, post #21675857]
What is a coupling capacitor, and why can it surge?
A coupling capacitor (C2) passes AC while blocking DC between stages. When B+ steps quickly, the capacitor momentarily passes a large transient until the downstream node discharges via its resistive path. [Elektroda, Charles Johnson, post #21675853]
Quick 3‑step fix to stop startup destruction?
- Add a pair of ESD diodes from the C2 output node to V+ and V−.
- Insert a small series resistor before the op‑amp input.
- Add series R/L in B+ and consider increasing C1 to slow the edge. [Elektroda, Dean Franks, post #21675857]
Can I use back‑to‑back zeners instead of ESD diodes?
Yes. Use a pair of fast zeners head‑to‑head from the C2 output node to ground, plus the series input resistor. This clamps the transient similarly to rail‑clamps. [Elektroda, Dean Franks, post #21675857]
Why did the circuit sometimes work and sometimes kill the LM4562?
The startup pulse amplitude varies with conditions, so outcomes alternate. The thread reported two outcomes: normal operation or failure after the hum. Clamping removes that variability. [Elektroda, Charles Johnson, post #21675853]
Do I need to change C1 or C2 values to help?
Increasing C1 helps form a stronger low‑pass with added series R/L, reducing the edge. Keep C2 as designed for bandwidth, but protect its output with clamps and a series resistor. [Elektroda, Dean Franks, post #21675857]
Is there a simple ground‑referenced clamp option?
Yes. Anti‑parallel small‑signal diodes from the op‑amp + input to ground were suggested. This provides bidirectional clamping at that node. [Elektroda, Charles Johnson, post #21675859]
What edge case should I watch for with slow‑ramp supplies?
If B+ ramps slowly, the transient shrinks, and failures may appear to vanish. However, a later fast restart can still exceed limits. Keep the input clamps. [Elektroda, Dean Franks, post #21675857]
Was forum etiquette addressed in this thread?
Yes. Participants asked to keep discussions constructive, then the original poster apologized and proceeded with technical fixes. Stay focused and share results. [Elektroda, Kevin Angelo Ma, post #21675856]
