You power on a receiver you haven’t touched in five years. The tuner comes alive, you dial in a station, and then you hear it: a high-pitched whine underneath the music, or worse, a low-frequency rumble that doesn’t match anything on air. You turn up the volume to check if it’s the input, and the noise gets worse—way worse. Then you smell something warm from the chassis. Your first thought is probably “the amplifier is dying,” but what you’re actually hearing is oscillation—and it’s almost always caused by something far more fixable than a failed output stage.
Oscillation in vintage stereo receivers is one of the most misdiagnosed problems in the hobby. It looks like an amp failure, sounds alarming, and creates enough heat that people immediately assume they need a recap or new output transistors. In reality, oscillation usually comes from a broken feedback network—a circuit feature that’s supposed to stabilize the amplifier but has degraded just enough to turn it into an uncontrolled oscillator instead. The physics is elegant. The troubleshooting is methodical. And almost always, the fix is both cheaper and faster than you’d expect.
This guide walks you through what’s actually happening inside your receiver when it oscillates, how to diagnose where the problem lives, and most importantly, how to fix it without guesswork or expensive parts replacement.
What you’ll learn and why it matters
Oscillation happens when an amplifier’s output signal feeds back into its input in a way that reinforces rather than stabilizes. In vintage receivers, this usually means a resistor has drifted out of value, a capacitor has dried out, a connection has corroded, or the feedback network topology has been disrupted. The symptoms range from audible high-frequency whistling to subsonic rumble, sometimes with visible distortion on the speaker output.
Understanding the feedback network itself—what it does, how it degrades, and why specific components fail—lets you diagnose the problem in minutes rather than days. You’ll learn to recognize oscillation signatures, test feedback components under real operating conditions, and identify which parts are actually responsible. Most importantly, you’ll know when you’re dealing with something you can safely troubleshoot yourself versus a situation that needs bench work.
How negative feedback works in a vintage amplifier
A basic amplifier without feedback is simple: a weak input signal goes to a gain stage, which amplifies it. But there’s a problem. The gain is incredibly unstable. Temperature changes, component drift, and even the load impedance on the speaker output all change how much the amp amplifies. An amp designed to have 100x gain might actually have 90x or 120x gain depending on the moment. The output signal shape also distorts—harmonics appear that weren’t in the input, the frequency response wobbles, and the amp becomes unpredictable.
Negative feedback fixes this. A small portion of the output signal is sent back to the input, but inverted (180 degrees out of phase with the original signal). This reversed signal subtracts from the original input before amplification happens. If the output is too loud, more inverted signal gets subtracted, reducing the next gain cycle. If the output is too quiet, less inverted signal is subtracted, letting the next cycle amplify more. The system self-corrects continuously.
In a typical vintage stereo receiver, negative feedback is applied across the entire power amplifier stage—from the output transformer secondary (or speaker terminals, depending on the topology) back to an earlier point in the signal chain, usually the input of the voltage amplifier or sometimes even the preamp output. A feedback network is built from resistors and capacitors strategically placed to set the amount of feedback and its frequency response characteristics.
The beauty of this system is that it dramatically reduces distortion, widens the bandwidth, and stabilizes the gain. A naked amplifier might have 2-3% THD (total harmonic distortion) and drift like crazy; with feedback, you get under 0.5% THD and rock-solid gain. The problem is that feedback networks can degrade in very specific ways, and when they do, the amplifier stops being stable and starts being an oscillator.
Why feedback networks oscillate
An oscillator is just an amplifier with positive feedback—feedback that reinforces the signal instead of opposing it. In a feedback network, oscillation typically starts when one of three things happens:
The phase shift becomes too aggressive. Feedback only works if the inverted signal actually arrives at the input in the right phase relationship. Inside the amplifier, there are multiple stages (input buffer, voltage amp, driver, output stage), and each stage introduces a small phase shift, especially at frequency extremes. Capacitors in the circuit also shift phase—a capacitor in the feedback path, in the power supply, or elsewhere can phase-shift the signal so much that it’s no longer properly inverted at the input. When the phase shift exceeds 180 degrees, the “inverted” feedback signal is actually reinforcing the input, and the amp starts to oscillate.
The feedback network’s resistors or capacitors drift out of value. The resistors in the feedback network set how much signal is fed back—usually 5-20% of the output. If a resistor’s value drifts high, less feedback is applied, reducing the damping effect. If it drifts low, the feedback becomes excessive, which can also cause instability. Capacitors in the feedback path drift too, and dried-out electrolytic capacitors are especially bad: they lose capacitance and gain impedance with frequency, distorting the phase relationship at high frequencies.
Corrosion or poor connections in the feedback path break continuity. If you’ve ever looked at the board inside an old receiver, you’ve probably seen oxidized solder joints and discolored resistor leads. A corroded connection in the feedback network path—even one with a tiny bit of resistance—can introduce an unexpected impedance into the signal path, phase-shifting it and disrupting the feedback balance. This is especially common at component legs, PCB traces, and connector pins.
The result is that the amplifier no longer damps oscillations—it amplifies them. A tiny noise, a stray RF signal, even thermal noise in a resistor becomes an unstoppable signal that grows louder and louder until it hits the limits of the power supply or the output transistors clip.
The oscillation signatures you’ll hear and measure
High-frequency oscillation (10 kHz–several MHz) sounds like a piercing whistle or hiss underneath the program material. It’s often described as “squealing” or a “computer-like” noise. This is very common in vintage receivers and is usually caused by phase shift at high frequencies—typically a dried-out capacitor in the feedback network or the power supply filter. The frequency usually correlates with the RC time constant of the problematic capacitor. Higher oscillation frequencies suggest the problem is in a high-frequency coupling path (a smaller capacitor); lower ones suggest a larger capacitor in a low-frequency path.
Midrange oscillation (1–10 kHz) sounds like a lower-pitched whistle or buzz and often seems to modulate with the program material. It’s sometimes mistaken for a mechanical hum from the amp chassis or transformer. This often indicates a problem with the main feedback resistor or a capacitor in the voltage amplifier stage.
Low-frequency oscillation (subsonic, 20 Hz–500 Hz) appears as a rumble or boom that seems to come and go. You might not hear it clearly, but you’ll feel it in the speakers as they move back and forth. This is usually caused by a dried-out capacitor in the power supply (filter or bypass cap) or problems in the feedback network at low frequencies. It’s particularly dangerous because the low frequency can push the output transistors and output transformer hard without clipping, potentially causing thermal runaway.
When you measure oscillation on a scope, it’s unmistakable: the signal has become a regular waveform (usually triangular or sawtooth) instead of a clean amplified version of the input. The frequency is steady and doesn’t change much with input frequency, which is how you know it’s oscillation and not distortion.
Which components actually degrade in feedback networks
Resistors in the feedback path. Vintage film resistors rarely drift catastrophically, but carbon composition resistors absolutely do. Carbon comp resistors are hygroscopic—they absorb moisture from the air, which increases their resistance, sometimes dramatically. A 10k resistor might creep to 12k or 15k over decades. In a feedback network where the resistor values are critical, this drift changes the feedback amount. A 20% change in feedback resistor value can be enough to trigger oscillation, especially combined with capacitor degradation.
Electrolytic capacitors in the feedback path and power supply. This is the big one. Dried-out electrolytic capacitors degrade in impedance as they age, losing their ability to pass high-frequency signals. A capacitor that was 100 µF at 1 kHz might be 10 µF at 10 kHz after 30 years. This impedance increase changes the feedback network’s frequency response, cutting off feedback at high frequencies. Without feedback, those high frequencies oscillate. Even worse, the ESR (equivalent series resistance) of old capacitors rises dramatically, introducing resistive loss that phase-shifts the signal.
Power supply filter capacitors. A sagging power supply—one with weak filter capacitors—introduces ripple onto the rails. This ripple can couple into the signal path through the amplifier’s power supply bypass network, causing phase shift and instability. If a 100 µF filter cap has dried out to 50 µF or less, the ripple frequency (usually 100–120 Hz for a full-wave rectifier) can couple into the amplifier, causing low-frequency oscillation or modulation of the entire output.
Coupling capacitors between stages. These set the high-pass cutoff frequency between amplifier stages. If a coupling capacitor has drifted or developed high ESR, it can restrict the bandwidth of a stage, causing phase shift. In a multi-stage amplifier, this cumulative phase shift can exceed 180 degrees and trigger oscillation, especially at frequency extremes.
Corroded connections and solder joints. Oxidized RCA jacks, corroded resistor legs, and cold solder joints introduce unexpected impedance into the feedback path. Even a tiny bit of resistance (a few ohms) can be enough to destabilize a feedback network that’s already marginal. This is especially true in the feedback connection itself—any corrosion here directly affects the signal being fed back.
How to diagnose oscillation systematically
Before you tear into the amp, confirm that you’re actually looking at oscillation and not something else (hum from a failing transformer, RF interference pickup, a failing speaker, or a bad input connection).
Step 1: Confirm oscillation with a scope and signal source
Connect a known-good signal generator (or even your phone playing a clean sine wave from a music app) directly to the amplifier’s input, bypassing the tuner and any preamp switches. Use shielded cables to avoid RF pickup. If the oscillation persists and the signal generator output is clean, the oscillation is coming from inside the amp itself.
If the oscillation disappears when you switch inputs or use the signal generator, the problem is in the preamp, selector circuit, or input stage—not the power amp feedback network itself. That’s still important information and narrows your search significantly.
With a scope on the output, look at the waveform. Pure oscillation will show a regular waveform (sine, triangle, or sawtooth) that’s independent of the input signal. If the waveform changes with the input frequency, you’re looking at distortion or a different problem.
Step 2: Identify the oscillation frequency
Use a spectrum analyzer (a smartphone app like Spectroid works, or a real analyzer if you have access) to measure the oscillation frequency. This tells you which part of the feedback network to suspect.
Above 50 kHz: Almost always a high-frequency coupling path issue, usually a small capacitor coupling the feedback signal or in the power supply. The output stage might be oscillating at its own natural frequency.
10–50 kHz: Likely a capacitor in the main feedback network or a compromised connection causing phase shift at high frequencies.
1–10 kHz: Usually the main feedback resistor or a midrange coupling capacitor.
Below 1 kHz: Power supply filter capacitor degradation or low-frequency feedback path problems.
Step 3: Measure DC voltages on the amplifier rails
A sagging power supply is often an overlooked culprit. Measure the positive and negative rails under load (with the amp oscillating and at high volume). A healthy amp should show rails within ±1-2V of their rated value at idle and under load. If the positive rail sags to 30V when it should be 35V, or if it wobbles, that’s a sign of weak filter capacitors or a failing power supply.
Check especially at the output stage. Some vintage amps have separate power supplies for the preamp and output stage. If the output stage’s rail is sagging, oscillation becomes much more likely because the output transistors have less headroom to fight against positive feedback.
Step 4: Trace the feedback network visually and measure component values
Pull the schematic and locate the feedback network. In most stereo receivers, it’s a cluster of resistors and capacitors connected from the output back to the input of the voltage amplifier or even the preamp input. The values are usually printed on the schematic.
Measure the resistance of each feedback resistor with a multimeter. Compare it to the schematic. A 1% variance is fine; 5% or more suggests drift, and anything over 10% is a red flag in a feedback network where precision matters. Remember: you’re measuring in-circuit, which means the multimeter is measuring the resistor with other components in parallel, so the reading might be lower than the actual resistor value. For a true reading, unsolder one leg.
Measure capacitor values if your meter has a capacitance function. Old electrolytics often read 20-30% lower than their rated value. If the schematic calls for a 10 µF cap and you’re reading 6-7 µF, that’s probably your oscillation culprit, especially if it’s in the feedback path.
Step 5: Check for corrosion and poor connections
Look at the feedback network physically. Are there any discolored resistor leads? Oxidation on component legs? Solder joints that look dull or grainy instead of shiny? Take photos if you see corrosion—you’ll want to address it. Even light corrosion can introduce enough resistance to destabilize a tight feedback network.
The feedback connection itself is critical. If there’s a wire running from the output transformer secondary or the speaker terminals back to the input, inspect it for corrosion, bent pins, or loose connections. A feedback wire that’s partially disconnected will cause serious oscillation.
Step 6: Measure AC impedance in the feedback path under operating conditions
This is where you need a scope or function generator and a load resistor. The idea is to inject a small AC signal at various frequencies into the feedback network and measure its impedance response. If the impedance jumps unpredictably at certain frequencies, you’ve found a capacitor that’s lost its ability to pass that frequency.
For a practical approach: inject a 1 kHz sine wave at 100 mV amplitude directly into the feedback network (at the point where feedback is summed with the input signal). Measure the voltage before and after the feedback network. Repeat at 5 kHz, 10 kHz, 50 kHz, and 100 Hz. If the impedance is high and frequency-dependent, a capacitor is degraded. If it’s consistently high across all frequencies, a resistor has drifted high or a connection is corroded.
When to replace components versus troubleshoot further
At this point, you probably know which component or connection is causing the problem. Now comes the decision: do you replace it, or is something deeper wrong?
If the culprit is a resistor that’s drifted high: Replace it with a modern metal film resistor of the same value. A 10% tolerance metal film resistor is stable and will likely solve the problem completely. Cost: <$0.50.
If the culprit is a capacitor: This is more nuanced. Some people advocate for replacing every electrolytic capacitor in a vintage amp as routine maintenance. Others replace only the failed ones. The honest answer is that capacitor failure is probabilistic—old capacitors fail, but not all at once. If you’ve identified a specific capacitor causing oscillation, replace it. You can replace it with a modern audio-grade electrolytic (Nichicon, Elna, or Vishay are good brands) of the same capacitance and voltage rating. For feedback network capacitors, try to match the ESR (equivalent series resistance) of the original if possible—modern low-ESR capacitors have different frequency response characteristics than vintage ones, which can subtly change the feedback behavior.
If the culprit is corrosion or a cold solder joint: Clean and resolder the connection. Use 60/40 tin/lead solder if possible for vintage work (it flows better on old boards), or modern lead-free if that’s what you have. Don’t use super-flux core solder; use rosin core. Clean the joint afterward with isopropyl alcohol and a small brush to remove flux residue.
If the power supply voltages are sagging: Before you replace filter capacitors, measure the transformer secondary voltage with no load and with the amp at full volume. If the secondary voltage drops more than 5%, the transformer itself might be failing (or the power supply loading is too high because of a short in the output stage). If the secondary is stable but the rails sag, the filter capacitors are almost certainly the problem. Replace the filter bank capacitors with modern equivalents of the same or higher capacitance. A 10,000 µF filter cap can usually be replaced with 15,000 µF or even 22,000 µF; the extra capacitance will only improve ripple rejection.
Edge cases and complications
Oscillation that appears only at certain frequencies: If the amp oscillates when you tune to a strong FM station but not a weak one, or when you play certain recordings, you’re probably looking at a partial feedback failure. The amp can handle weak signals, but strong signals push it into a region where the feedback network’s phase shift becomes critical. This usually means a capacitor in the feedback path has enough remaining value to work at weak signal levels but fails at high levels or high frequency content. Replace that capacitor and the problem should vanish.
Oscillation that emerges only after the amp warms up: Temperature-dependent oscillation suggests a component that’s thermally unstable. As resistors warm, their values change slightly. As capacitors warm, their ESR changes. Usually this points to a capacitor that’s degraded unevenly—parts of it are still functional at room temperature, but once it gets hot, its impedance rises enough to trigger oscillation. The fix is always capacitor replacement in this case.
Oscillation that appears in one channel but not the other: You’re looking at a problem specific to that channel’s feedback network. The problem is probably a resistor or capacitor that’s channel-specific, a corroded connection on one side, or (rarely) a thermal issue with one output transistor. Measure component values on both channels—you’ll usually find them different on the affected channel.
Oscillation that comes and goes randomly: This is often a cold solder joint or a connection that’s barely making contact. When the amp heats up, expansion and contraction can vary the contact resistance slightly, which causes the feedback network to sometimes work and sometimes fail. Inspect the feedback network carefully for any questionable solder joints or connections. Reflow them and the problem should be resolved.
Very high-frequency oscillation (MHz range) in the output stage only: Some vintage amplifiers, especially those with output transformer feedback or complicated output stage topologies, can oscillate at very high frequencies due to the output transistors’ own internal capacitance and the load impedance. This is usually a stability problem specific to the output stage and might require adding a small series resistor and capacitor from the speaker output to ground (a “Zobel” network) to dampen the high-frequency behavior. This is beyond simple feedback network repair and might need professional advice.
Safety and practical considerations
Before you work on any vintage amplifier, remember that the output transformer and power supply contain high voltages that can cause serious injury or death. Even with the amp powered off, filter capacitors can hold a charge. Always discharge the power supply filter capacitors by connecting a resistor load across them before touching anything. A 10k resistor works—hold one end to the positive rail and touch the negative rail with the other end, then wait 10 seconds.
Use an isolation transformer when powering up a vintage amp you’re not sure about. This prevents catastrophic damage if there’s a short. Never power up an amp you suspect has oscillation without a speaker load (or at minimum, a large series resistor between the output and ground to limit current). Oscillation can burn out output transistors very quickly, especially at high frequencies where the impedance is lower.
If you’re replacing capacitors, be extremely careful with solder and heat. Vintage PCB boards can be fragile, and excessive heat can lift traces. Use a temperature-controlled soldering iron set to 350°C, and don’t hold the iron on the joint for more than 3-4 seconds.
Making a practical troubleshooting decision
After diagnosis, you’ll have a fairly clear picture of what’s wrong. Here’s the framework for deciding what to do next:
If the problem is a single degraded component (a resistor or capacitor) in the feedback network: Replace it yourself if you’re comfortable soldering. Cost: under $5, time: 30 minutes. This is almost always the right call—it’s low-risk and very likely to fix the problem completely.
If the problem is corroded connections or cold solder joints: Reflow and clean the joints yourself. Same time and cost as above. Very safe if you use a desoldering iron or solder wick.
If the power supply is sagging: Measure the voltages carefully to confirm the filter capacitors are the problem, not a failing transformer. If it’s the caps, you can replace them. If the transformer secondary is dropping voltage, the transformer itself might be failing, which is a professional-level repair.
If you’ve narrowed it down to a specific problem but you’re not confident soldering or don’t want to risk further damage: Take detailed photos and notes of what you found and send the amp to a specialist. You’ve already done the hardest part (diagnosis), which will save them hours of work and you money on labor.
Oscillation in vintage receivers is rarely a catastrophic failure. It’s usually a straightforward component degradation problem that responds very well to targeted repair. The feedback network is one of the most elegant and reliable circuits in vintage audio—it just needs its components to stay within tolerance. Once you understand what’s happening, the fix is almost always obvious.