vk6ada.com.au • Vintage Receiver Fault Diagnosis Series • Collins S-Line

Collins 75S-3 S-Line Receiver
AGC Collapse Fault Diagnosis — S-meter/Volume Interaction, AGC Bus Analysis, Root Cause Candidates

A fault pattern in which the Collins 75S-3’s AGC stops controlling gain. AGC collapse leaves the receiver running at full RF and IF gain: loud stations arrive brutally loud, the volume control becomes hypersensitive, the S-meter barely moves, and the AGC bus voltage no longer follows signal level — the diagnostic signature, found with a single AGC bus measurement, that distinguishes AGC chain failure from other 75S-3 faults. This analysis explains the 75S-3’s delayed AGC (rectified by a diode section of V9, with no AGC amplifier tube) and its bridge-type S-meter, then covers the gateway AGC bus voltage measurement, seven root cause candidates in priority order, a comparison with the 75A-4’s AVC system, and a step-by-step check procedure. Covers all variants: 75S-3, 75S-3A, 75S-3B, and 75S-3C.

✍ Mike Peace VK6ADA / r-390a.net and foxtango.org Administrator 📅 March 2026 ⚙ Collins 75S-3 / 75S-3A / 75S-3B / 75S-3C • S-Line • Double conversion • 190 V B+ (unregulated) 📊 AGC collapse • Delayed AGC • V9 diode rectifier • Bridge S-meter • 75A-4 comparison
⚠  B+ Voltage Safety — Up to About 190 V DC The Collins 75S-3’s internal power supply delivers about 190 V DC unregulated, 140 V regulated and 135 V B+ (CCA figures for the 75S-3B/3C), plus a −65 V bias line, and the power transformer primary carries mains voltage. All powered measurements must be performed one-handed with an insulated probe, chassis on a non-conductive surface. The AGC bus diagnostic (Section 4) involves probing the AGC bus while the receiver is powered. Keep the free hand away from the chassis during all powered measurements. Power off and wait 30 seconds before touching any plate connection or component lead directly.

AGC collapse in the Collins 75S-3 has a symptom pattern that is immediately recognisable once seen but easily misdiagnosed on first encounter. The receiver appears to work — it receives signals, the audio sounds mostly coherent — but gain control is absent or severely degraded. Strong signals arrive at full RF and IF gain, the audio volume becomes uncomfortable on busy bands, and the AGC action that makes the S-Line a comfortable operating receiver is simply gone.

The defining diagnostic is the AGC bus voltage measurement (Section 4): in a healthy 75S-3 the AGC bus swings negative as signal strength rises; with collapsed AGC it barely moves, and strong signals pass at full IF gain. Do not use the RF GAIN control as the test: on the 75S-3 it works by placing bias on the AGC line, so in a healthy receiver the S-meter normally rises as RF GAIN is reduced. The S-meter is a second, free indicator: it reads the drop in V6/V7 current that the AGC causes, so a receiver that sounds loud while its S-meter barely moves is not getting AGC voltage to its IF amplifiers. When the AGC bus does not follow signal level, the AGC chain is broken. The question is where.

There is no AGC amplifier tube to blame in the 75S-3. The IF signal from the secondary of T6 is rectified directly by a diode section of V9 (which also contains the AM detector), and the resulting negative voltage passes through the AGC time-constant network to the RF amplifier V2 and the IF amplifiers. The candidates are therefore the front-panel AGC switch, V9 and its delay bias, the AGC filter network, the RF GAIN bias divider, and leakage or open circuits on the AGC bus and at the controlled grids — plus the muting line and the S-meter bridge, which can confuse the diagnosis. This guide provides the measurements that distinguish them. Component designators are taken from the CCA description of the 75S-3B/3C; cross-check them against the manual for the specific variant before any probing.

Collins S-Line Context

The 75S-3 is the receiver half of the Collins S-Line station (75S-3 receiver + 32S-3 transmitter + 312B-4 speaker/console or equivalent). The S-Line was produced from approximately 1959 through the late 1970s in several variants. The AGC arrangement described here is taken from the Collins Collectors Association (CCA) description of the 75S-3B and 75S-3C (the 75S-3C is identical to the 75S-3B apart from an additional HF crystal board and front-panel selector switch); check the 75S-3/3A manual for designator differences in earlier sets. For companion transmitter faults: the 32S-3 transmitter has its own carrier and ALC system documented separately at vk6ada.com.au. Collins Service Bulletins relevant to the 75S-3 are indexed at vk6ada.com.au/collins-service-bulletin-index/.

Section 1 — How the 75S-3 AGC System Works

Signal Path Overview

The Collins 75S-3 is a double-conversion superheterodyne receiver. The RF amplifier V2 feeds the first mixer V3A, whose injection comes from a crystal oscillator (V3B) running 3.155 mc above the lower edge of the selected band, giving a first IF that is a 200 kc wide bandpass from 3.155 to 2.955 mc. The second mixer V4A, injected by the VFO, converts this to the fixed 455 kc second IF, where the Collins mechanical filter sets the selectivity. The IF preamplifier V5A and the IF amplifiers V6 and V7 then bring the signal up to the detectors: the product detector V8A for SSB and CW, and a diode section of V9 for AM. The AGC acts on the RF amplifier and the IF amplifier stages, reducing their gain when a strong signal arrives and restoring it when the band is quiet.

The AGC Chain

In the 75S-3 the AGC voltage is produced directly, with no amplifier stage. IF signal from the secondary of T6 is rectified by a diode section of V9, and the negative DC passes through R50 and C49 to the AGC network (R24, R88, C50, C137 and C153), and from there along the AGC bus to the controlled stages. The AGC is delayed: V9’s cathode is biased so that no AGC voltage is developed until the signal at the diode plate exceeds that bias, which corresponds to roughly 2 µV at the antenna (the specification gives a threshold of 1.5–3.0 µV, set by R57). Weak signals and band noise therefore produce no AGC voltage at all. Two RC sections shape the response: R88 in parallel with C153 gives a fast charge and discharge that rides over brief interference, and R24 in parallel with C50 gives the longer time constant for a smoothly developed AGC voltage. The specification gives an attack time of 1 ms and release times of about 190 ms (fast) and 600 ms (slow); the front-panel AGC switch selects the release time or turns the AGC off.

The S-meter is not driven from the AGC bus. It is a bridge connected from the screen circuits of V6 and V7, the two IF amplifiers, to the cathode of V7. With no signal the voltages across R13 and R21 are equal and the meter reads zero; when AGC voltage reduces the cathode and screen currents of these tubes, the bridge unbalances and the meter reads upscale in proportion to signal strength. The meter therefore shows how hard the AGC is actually turning down V6 and V7.

✎  The S-meter Reads the AGC’s Effect, Not the Signal This is the key to understanding the collapse symptom pattern. The S-meter only moves when AGC (or RF GAIN) bias reduces the current in V6 and V7. In normal operation a strong signal drives the AGC bus negative, the RF and IF gain drop (audio stays comfortable) and the meter reads upscale. In AGC collapse no AGC voltage reaches V6 and V7, so the IF gain stays high (audio gets loud) and the S-meter stays near zero even on loud stations. A loud receiver with a lazy S-meter is therefore the signature of AGC collapse, and it points at the AGC voltage source or the bus feeding V6 and V7 rather than at the meter.

The RF GAIN Control and AGC Interaction

The RF GAIN control in the 75S-3 does not operate independently of the AGC system: it works through the AGC line. The control is a potentiometer in a divider on the negative bias supply that places a static bias of about 1 V on the AGC line at maximum gain; lower settings add bias and reduce the gain of the AGC-controlled RF and IF stages. Because the S-meter responds to the bias on these stages, a healthy 75S-3’s S-meter normally rises as RF GAIN is reduced, even with no signal present.

S-meter movement with the RF GAIN control is therefore normal and is not a test for AGC collapse. Judge the AGC chain by whether the AGC bus voltage follows signal level, using the measurement in Section 4.

Section 2 — The AGC Collapse Symptom Pattern

Characteristic Presentation

The 75S-3 with collapsed or severely degraded AGC presents with a distinctive constellation of symptoms. Not all will be present in every case — partial AGC collapse (AGC working but with reduced range) produces a milder version of the same pattern:

  • Loud strong signals. Stations that should arrive at comfortable listening volume are painfully loud. On a busy amateur band, the receiver is uncomfortable to monitor at the normal operating volume control setting.
  • Hypersensitive volume control. A small rotation of the AF GAIN control produces a large change in volume because the IF gain is not being held to a stable baseline by the AGC. The effective gain of the audio chain appears higher than expected.
  • S-meter reads low or barely moves. Because the 75S-3 meter reads the AGC-driven drop in V6/V7 current, a receiver with collapsed AGC shows little or no meter movement even on stations that are uncomfortably loud. A meter that reads normally while the audio is still too loud means the AGC is reaching V6 and V7, and points instead to the AGC feed to the RF amplifier V2 or to a fault outside the AGC chain.
  • AGC bus not following signal level. The AGC bus voltage changes little between no signal and a strong signal (Section 4). This is the diagnostic signature. S-meter movement with the RF GAIN control is not: on the 75S-3 the S-meter normally rises as RF GAIN is reduced.
  • Poor strong-signal handling. On the same band as a strong local station, weak signals may be swamped because the strong station drives the RF and IF stages into overload instead of being held back by the AGC. If turning RF GAIN down also has little effect, suspect something holding the AGC bus itself near zero, such as leakage on the bus, because the RF GAIN bias and the AGC voltage share that line.
  • Distorted CW and SSB on strong signals. With no AGC holding the gain down, strong signals overdrive the IF stages and the product detector, so CW notes and SSB voices sound harsh or distorted rather than simply loud.

What AGC Collapse Is Not

Several common 75S-3 faults produce superficially similar symptoms and should be ruled out before committing to the AGC chain diagnosis:

  • Not an RF overload problem. Connecting an external attenuator at the antenna port and observing whether the symptoms reduce in proportion to the attenuation tests this. If a 20 dB pad restores normal behaviour, the problem is overload from a strong nearby transmitter, not AGC collapse.
  • Not a mechanical filter bandwidth problem. A failed or high-insertion-loss mechanical filter produces weak signals across all modes, not loud signals.
  • Not an audio volume control fault. A scratchy or carbon-tracked AF GAIN pot can produce erratic volume, but it does not affect the S-meter or the AGC bus voltage.
  • Not a power supply hum problem. Hum from failed filter capacitors produces volume-dependent or volume-independent audio contamination, not a loss of gain control.

Section 3 — The 75A-4 Comparison: What Transfers and What Does Not

Restorers familiar with the older Collins 75A-4 often go looking for an AGC amplifier tube in the 75S-3, because the 75A-4 has one. In the 75A-4 a separate IF amplifier stage, V-21, amplifies a sample of the IF signal for the AVC (Automatic Volume Control) rectifier, which is one half of V-16, a twin diode; the AVC controls the RF amplifier V-2 and the 455 kc IF amplifiers V-6, V-8 and V-9. In a 75A-4 with weak AVC, V-21 is therefore a legitimate early suspect. The 75S-3 has no such stage: its AGC diode in V9 is fed straight from T6.

The terminology difference. The 75A-4 uses the older term AVC; the 75S-3 uses AGC (Automatic Gain Control). They describe the same function: a feedback system that rectifies the IF signal and uses the result to reduce the gain of the RF and IF stages.

Design Differences That Matter for Diagnosis

The two receivers share more than restorers sometimes expect, but the differences decide where to look first:

  • Both are delayed. The 75A-4 holds off its AVC with a small positive DC voltage at the rectifier so that it is ineffective on weak signals; the 75S-3 delays its AGC with cathode bias on V9, starting at about 2 µV. In both sets weak signals and band noise produce no control voltage, so a quiet band with RF GAIN at maximum is not a test of the AGC chain.
  • No amplifier in the 75S-3. With no V-21 equivalent, the first tube to check is V9, the AGC rectifier, rather than an amplifier stage.
  • An AGC switch and two time constants. The 75S-3’s front-panel AGC switch turns the AGC off or selects a fast (about 190 ms) or slow (about 600 ms) release. A switch left off, or dirty switch contacts, reproduces the full collapse pattern with nothing wrong in the circuit.
  • Muting through the AGC bus. In the 75S-3 a ground on the MUTE jack, J11, keeps the receiver operating; removing it applies cutoff bias to the mixer grid and increases the bias on the AGC bus to silence the receiver during transmit. Faults here make the receiver weak or dead rather than loud, so they are a related but different symptom.
  • Similar S-meters. Both use a bridge rather than a meter on the control line: the 75A-4 between the screens of V-6 and V-8 and the cathode of V-8, the 75S-3 between the screen circuits of V6 and V7 and the cathode of V7. In both, the meter reads the control system’s effect on the IF amplifiers.
✎  From 75A-4 to 75S-3 — What Transfers The gateway test transfers directly: measure the control voltage on the AGC bus and verify that it changes between no-signal and strong-signal conditions. The tube-substitution habit does not: there is no V-21 in the 75S-3, so start with the AGC switch and V9, then the AGC filter network and the bus itself. The 75S-3 designators and manual section numbering differ from the 75A-4’s, so work from the 75S-3 schematic for the specific variant.

Section 4 — Gateway Test: AGC Bus Voltage Measurement

Before any tube substitution or component replacement, make this single measurement. It takes three minutes and answers the most important question: is the AGC bus responding to signal level at all? The answer determines which sub-category of AGC fault you are dealing with and which candidates to address first.

The Measurement Procedure

Connect a high-impedance DVM (input impedance >10 MΩ) on the DC voltage range to the AGC bus — the line fed through R50 that carries the AGC filter network (R24, R88, C50, C137, C153) to the controlled stages; pick a convenient point on it from the schematic for the specific variant. Set the AGC switch to one of its operating positions, not off. Allow 15 minutes warm-up. Take three readings:

  • Condition A — no antenna, RF GAIN at maximum: the AGC bus should sit at the small static bias set by the RF GAIN divider — about −1 V at maximum gain — because the delayed AGC develops nothing from noise. A reading far from this, or a positive reading, points to the RF GAIN divider (R56, R33, R55), the −65 V bias supply, or leakage into the AGC bus.
  • Condition B — antenna connected, active HF band, RF GAIN at maximum: as signals above the AGC threshold arrive, the AGC bus should go more negative, visibly following signals as they come and go, and the S-meter should move with it. Compare against the service manual for the specific variant.
  • Condition C — inject a known strong signal (signal generator, −40 dBm, 14 MHz, 1 kHz AM modulation): observe the AGC bus voltage. It should drive substantially negative in response to the injected carrier and return to the static bias (about −1 V) at the release rate when the generator is switched off. Repeat with the AGC switch in its other operating position.
✎  Interpreting the AGC Bus Voltage Reading Good result: the AGC bus swings several volts more negative between Condition A and Condition C, and the S-meter rises with it. If the receiver still sounds uncontrolled, the AGC is being generated but is not reaching a controlled stage — check delivery to V2 and the IF amplifiers (Candidate #5).

Partial collapse: some AGC voltage appears, but much less than expected. Likely causes are a weak V9 diode section or excessive delay bias (Candidate #2), or leakage loading the bus (Candidates #3 to #5).

Full collapse: the bus stays at the static bias whatever the signal. Check the AGC switch first (Candidate #1), then V9 and its feed from T6 (Candidate #2), then an open R50 or a fault in the AGC network (Candidate #3). If RF GAIN also has little effect, something is holding the bus near zero (Candidates #3 to #5).

Section 5 — Seven Root Cause Candidates in Priority Order

Candidates #1 to #4 cover the generation of the AGC voltage and the AGC bus itself; #5 covers its delivery to the controlled stages; #6 and #7 are adjacent circuits (the muting line and the S-meter bridge) that change the symptoms without causing the primary collapse. Designators are from the CCA description of the 75S-3B/3C; verify them against the manual for the specific variant before any measurements.

  • #1
    Front-panel AGC switch — position and contacts FIRST ACTION — QUICKEST CHECK

    The front-panel AGC switch turns the AGC off or selects one of two release time constants. Left in the off position it reproduces the full collapse pattern exactly, with nothing wrong in the circuit, and oxidised wafer contacts can produce AGC that comes and goes or works in only one position.

    Confirm the switch position, then compare the two operating positions with the Section 4 measurement. If one position works and the other does not, the fault is in the switch or in the time-constant components it selects (Candidate #3). Work the switch through its positions several times and treat the contacts with a suitable contact cleaner if the AGC returns erratically.

  • #2
    V9 — the AGC rectifier diode and its delay bias HIGH PROBABILITY — ESPECIALLY IF THE AGC BUS SHOWS NO RESPONSE

    The AGC voltage comes from a diode section of V9, fed from the secondary of T6; V9 also contains the AM detector diode. A V9 with low emission, an open heater or a poor socket contact produces little or no AGC voltage. If AM reception has also become weak or dead while SSB through the product detector still works, V9 is strongly indicated, because it carries both functions.

    Substitute a known-good tube of the type shown on the tube location label and repeat the Section 4 measurement. Also check V9’s cathode bias, which sets the AGC delay: R57 sets the threshold at about 2 µV (specification 1.5–3.0 µV). Excessive cathode bias holds the AGC off until signals are already strong, giving a partial-collapse pattern on moderate signals.

    Also check: the V9 socket for oxidised contacts or carbon tracking between adjacent pins. Clean with 99% IPA if there is any evidence of contamination.

  • #3
    AGC filter and time-constant network — R50, C49, R24, C50, R88, C153, C137 SECONDARY — ESPECIALLY IF AGC BEHAVIOUR IS ERRATIC OR DIFFERS BETWEEN FAST AND SLOW

    The rectified AGC voltage passes through R50 and C49 to the AGC network. R88 with C153 sets the fast charge and discharge, and R24 with C50 the longer time constant; the published specification gives a 1 ms attack and release times of about 190 ms (fast) and 600 ms (slow). After sixty years these parts fail in ways that affect the symptom pattern differently:

    Open circuit: an open time-constant capacitor makes the release far too quick. Symptoms: fast S-meter pumping, AGC that follows individual syllables on voice signals (a “breathing” quality to SSB), and CW where the AGC attacks on each dit and releases between them. An open R50 breaks the path from V9 altogether and gives full collapse.

    Leaky (partial DC conduction): a capacitor that leaks loads the AGC bus and pulls it towards zero, so less AGC voltage reaches the controlled stages and the RF GAIN control loses effect as well. In Condition A of Section 4 the bus reads less negative than the normal static bias, or even positive if the leakage path is to a positive supply.

    Service: identify the capacitors in the network from the schematic and measure each out of circuit for leakage and value. Replace any that leak or have lost capacitance with modern film capacitors of the original value and voltage rating, and check R50, R24 and R88 for drift.

  • #4
    RF GAIN divider and −65 V bias line — R56, R33, R55 SECONDARY — MEASURE WHEN THE AGC BUS STATIC BIAS IS WRONG

    The RF GAIN control, R56, sits in a divider with R33 and R55 across the −65 V bias line and places a static bias of about 1 V on the AGC bus at maximum gain; turning it down adds bias. Carbon composition resistors in this divider drift upwards with age, and a worn RF GAIN track can become noisy, changing the static bias on which the AGC voltage builds.

    Measure the divider resistors out of circuit and replace any more than 10% above nominal with metal film types, and check the −65 V bias supply. This is most productive when Condition A in Section 4 is well away from about −1 V, or when the RF GAIN control behaves erratically.

  • #5
    Delivery to the controlled grids — series resistors and bypass capacitors at V2 and the IF amplifiers CHECK IF AGC BUS IS CORRECT BUT GAIN IS STILL UNCONTROLLED

    If the Section 4 gateway test confirms that the AGC bus produces a correct and varying voltage but the receiver still shows the collapse pattern, the fault is in the delivery path from the bus to the grids of the RF amplifier V2 and the IF amplifiers. Each controlled stage takes its AGC through a series resistor, with an RF bypass capacitor at the grid end. Two mechanisms:

    Leaky or shorted bypass capacitor: pulls that grid’s AGC voltage towards zero, so that stage runs at full gain; heavy leakage can drag down the whole bus. An open bypass capacitor does not stop the DC AGC voltage reaching the grid; it removes RF decoupling and is more likely to cause instability.

    Open or high-resistance series resistor: the AGC voltage cannot reach that grid, and that stage runs at full gain while the others are controlled. Find the culprit by measuring the DC voltage at each controlled grid with a strong signal applied: the controlled stages show a clear negative voltage, the faulty one stays near zero. Measure suspect resistors out of circuit and replace with metal film if open or significantly high.

  • #6
    Muting line — MUTE jack J11 CHECK IF GAIN CHANGES WITH THE TRANSMITTER CONNECTED OR KEYED

    The 75S-3 is silenced during transmit through its MUTE jack, J11. A ground on J11 keeps the receiver operating normally; removing it applies cutoff bias to the mixer grid and increases the bias on the AGC bus. A muting fault therefore makes the receiver weak, dead or intermittent rather than loud. It is not a cause of AGC collapse, but it changes the AGC bus readings and can confuse the diagnosis.

    The distinguishing check: disconnect the muting cable from the transmitter and ground J11 directly with a short jumper. If the receiver then behaves normally, trace the muting connection back to the transmitter; if not, the fault is inside the receiver.

  • #7
    S-meter bridge — not a cause of collapse, but confuses the diagnosis VERIFY SEPARATELY IF S-METER READS WRONG WHILE AGC FUNCTIONS CORRECTLY

    The S-meter is a bridge from the screen circuits of V6 and V7 to the cathode of V7, balanced at no signal by equal voltages across R13 and R21. It is not connected to the AGC bus, and a fault in it does not affect the AGC bus voltage or the IF gain. Drift in R13 or R21, or a change in V6 or V7, can upset the balance so that the meter will not zero or reads wrongly while the AGC is working well, which invites a misdiagnosis of AGC collapse.

    Distinguishing factor: if the AGC bus and the grid voltages at V6 and V7 follow signal level correctly (Section 4 and Candidate #5) but the meter does not track, the fault is in the meter bridge — R13, R21, the meter itself or its zero and calibration adjustments. Correct it per the service manual.

Section 6 — Recommended Diagnostic Procedure

  • 1
    Confirm the AGC collapse symptom and rule out RF overload With the receiver powered and tuned to an active HF band, confirm the symptom pattern in Section 2 (loud strong signals, hypersensitive AF GAIN, an S-meter that barely moves). Check that the front-panel AGC switch is not set to off. Insert a 20 dB attenuator in the antenna line: if this restores normal behaviour, the problem is overload from a strong nearby transmitter, not AGC collapse. Do not use the RF GAIN control as an AGC test; on the 75S-3 the S-meter normally rises as RF GAIN is reduced because the control biases the AGC line.
  • 2
    Powered: AGC bus voltage — three-condition measurement (Section 4) Identify the AGC bus on the schematic (the line fed through R50 that carries the AGC filter network). Measure DC voltage: (A) no antenna, RF GAIN max — expect the static bias of about −1 V; (B) antenna connected, active band — observe that the voltage varies with signal; (C) signal generator injected at −40 dBm — note the maximum negative voltage reached, in both operating positions of the AGC switch. Record all readings. They determine which candidate group to address first.
  • 3
    AGC switch, then V9 — contact check and tube substitution (Candidates #1 and #2) If the two AGC switch positions behave differently, or the AGC comes and goes as the switch is moved, work and clean the switch contacts. Then power off, substitute V9 with a known-good tube and clean its socket. Restore power and repeat Step 2. Note whether AM reception was also affected: V9 carries both the AGC and AM detector diodes. If the full Condition A to Condition C swing returns, the switch or tube was the cause; if not, continue to Step 4.
  • 4
    Power off: AGC filter network — out-of-circuit measurement (Candidate #3) Identify R50, C49, R24, C50, R88, C153 and C137 from the schematic. Lift one lead of each capacitor and measure it for leakage on the DVM’s highest resistance range and for value on a capacitance meter: any measurable leakage, or a value well below the marking, means replacement. Check R50 for an open circuit and R24 and R88 for drift. Replace capacitors with film types of the original value and voltage rating.
  • 5
    Power off: RF GAIN divider and bias resistors — out-of-circuit measurement (Candidate #4) With the capacitor condition resolved, measure R56, R33 and R55 and any other carbon composition resistors in the AGC bias network. Replace any more than 10% above nominal with 1% metal film. Also measure the series resistor feeding each AGC-controlled grid: an open circuit or dramatically high resistance confirms a delivery-path fault (Candidate #5) and identifies the stage affected.
  • 6
    Powered: controlled-grid voltage check (Candidate #5) — if AGC bus is correct but gain uncontrolled If Steps 2–5 confirm that the AGC bus produces the correct voltage range but the receiver still shows collapse symptoms: with a strong signal applied, measure the DC voltage at the control grid of V2 and of each AGC-controlled IF tube. Each controlled stage should show a clear negative voltage when a strong signal is present. A stage near zero while the others are correct has a delivery fault — an open or high-resistance series resistor, or a leaky bypass capacitor at that grid.
  • 7
    Muting line check (Candidate #6) If the gain changes when the transmitter is connected, switched or keyed, disconnect the muting cable and ground the MUTE jack J11 directly with a short jumper. If the receiver then behaves normally, trace the muting connection back to the transmitter; if the fault persists, follow the muting path inside the receiver using the manual.
  • 8
    Powered: AGC bus retest and S-meter verification After all identified faults are corrected, repeat the Step 2 AGC bus measurement. The AGC bus should now swing several volts more negative between Condition A and Condition C, and the S-meter should rise with it. If the AGC now responds correctly but the S-meter reads incorrectly for the signal level, proceed to the S-meter bridge (Candidate #7 and the service manual alignment section). Note: S-meter calibration should not be adjusted until the AGC chain is confirmed functioning correctly.

Section 7 — AGC Chain and Fault Map

  ┌──────────────────────────────────────────────────────────────────────────┐
  │   COLLINS 75S-3 AGC SYSTEM — SIGNAL FLOW AND FAULT LOCATIONS             │
  │   Designators from the CCA 75S-3B/3C circuit description. Verify them    │
  │   against the manual for the specific variant before probing             │
  └──────────────────────────────────────────────────────────────────────────┘

  ANT → [V2 RF AMP] → [V3A 1ST MIX] → [1ST IF 3.155–2.955 mc] → [V4A 2ND MIX]
            ▲                                                     │
            │                                                     ▼
            │   [455 kc: MECH FILTER → V5A PREAMP → V6 IF → V7 IF] → T6
            │                                       ▲       ▲        │
            │                                       └───┬───┘        ▼
            │                                           │  [V9 DIODE SECTION]
            │                                           │   AGC rectifier; no
            │                                           │   amplifier stage.
            │                                           │   Delay: V9 cathode
            │                                           │   bias set by R57,
            │                                           │   about 2 µV.
            │                                           │   (Candidate #2)
            │                                           │            │ R50,C49
            │                                           │            ▼
            │                                           │  [AGC NETWORK]
            │                                           │   R88‖C153 fast
            │                                           │   R24‖C50 slow, C137
            │                                           │   (Candidate #3)
            │                                           │   AGC switch: off,
            │                                           │   fast, slow (#1)
            │                                           │            │
            └───────────────────────────────────────────┴─ AGC BUS ◄─┘
                  ← measure here for the gateway test (Section 4)
                  RF GAIN divider R56/R33/R55 on −65 V sets about −1 V
                    static bias at maximum gain (Candidate #4)
                  MUTE jack J11: grounded = receive; open = muting bias
                    on the mixer grid and the AGC bus (Candidate #6)
                  Each controlled grid: series R + bypass C (Candidate #5)

  S-METER: bridge from the V6/V7 screen circuits to the V7 cathode
           (R13, R21). It reads the AGC's reduction of V6/V7 current
           and is NOT wired to the AGC bus (Candidate #7)

  ────────────────────────────────────────────────────────────────────────────
  GATEWAY TEST INTERPRETATION
  ┌──────────────────────────────────────────────────────────────────────────┐
  │ RF GAIN lowered, S-meter rises:   Normal: RF GAIN biases the AGC bus.    │
  │ Condition A about −1 V:           Normal static bias at maximum gain.    │
  │ Condition A far from −1 V, or +:  RF GAIN divider, −65 V line or         │
  │                                   leakage on the bus (#4, #3)            │
  │ Bus swings several volts, meter   AGC generated and reaching V6/V7;      │
  │   rises, audio still too loud:    check delivery to V2 (#5)              │
  │ Bus swings, S-meter stays low:    Not reaching V6/V7 (#5), or a          │
  │                                   meter bridge fault (#7)                │
  │ Bus stays at the static bias:     AGC switch, V9, R50 or T6 feed         │
  │                                   (#1, #2, #3)                           │
  │ RF GAIN has little effect too:    Bus held near 0 V by leakage (#3–#5)   │
  └──────────────────────────────────────────────────────────────────────────┘
  ────────────────────────────────────────────────────────────────────────────
  75S-3 vs 75A-4 — AGC SYSTEM COMPARISON
  ┌──────────────────────────────────────────────────────────────────────────┐
  │ Feature             │ Collins 75S-3            │ Collins 75A-4           │
  │ Control system name │ AGC                      │ AVC                     │
  │ Type                │ Delayed, V9 cathode bias │ Delayed                 │
  │ Separate amplifier  │ None                     │ V-21 AVC IF amplifier   │
  │ Rectifier           │ Diode section of V9      │ Half of V-16            │
  │ Stages controlled   │ RF amp V2 + IF amps      │ V-2, V-6, V-8, V-9      │
  │ Gateway measurement │ AGC bus DC voltage       │ AVC line DC voltage     │
  │ First tube to try   │ V9                       │ V-21, then V-16         │
  │ S-meter             │ Bridge: V6/V7 screens    │ Bridge: V-6/V-8 screens │
  │                     │   to V7 cathode          │   to V-8 cathode        │
  └──────────────────────────────────────────────────────────────────────────┘
  ────────────────────────────────────────────────────────────────────────────
  SYMPTOM-TO-CANDIDATE GUIDE
  ┌──────────────────────────────────────────────────────────────────────────┐
  │ Symptom                     │ What it suggests              │ Check      │
  │ Loud, S-meter barely moves  │ No AGC voltage at V6/V7       │ #1–#3      │
  │ AM weak or dead as well     │ V9 (AGC and AM diodes)        │ #2         │
  │ Only one of fast/slow works │ Switch or its time constants  │ #1, #3     │
  │ Breathing / pumping         │ Open AGC filter capacitor     │ #3         │
  │ RF GAIN also ineffective    │ Bus held near 0 V by leakage  │ #3–#5      │
  │ S-meter wrong, AGC normal   │ Meter bridge (R13, R21)       │ #7         │
  │ Weak or dead with Tx linked │ Muting path (J11)             │ #6         │
  └──────────────────────────────────────────────────────────────────────────┘

AGC system signal flow and fault location map — Collins 75S-3. All component designators require verification from the Collins 75S-3 Service Manual for the specific variant. The AGC bus gateway measurement in Section 4 is the required first step before any component work.

Section 8 — Verification Tests

RF GAIN Control — Expected Behaviour

Test: With the receiver tuned to a clear frequency, rotate RF GAIN from maximum towards minimum. The S-meter should rise as RF GAIN is reduced, because on the 75S-3 the RF GAIN control works by placing bias on the AGC line, which the S-meter reads. This is normal behaviour, not a fault, and it does not test the AGC detector chain. Pass: the S-meter rises smoothly as RF GAIN is reduced; judge AGC performance by the AGC bus measurement (Section 4) and the AGC Range Verification below, not by the RF GAIN control.

AGC Range Verification

Test: With a calibrated signal generator, inject a 14 MHz AM carrier at three levels: −80 dBm, −60 dBm and −40 dBm (about 22 µV, 220 µV and 2.2 mV into 50 Ω, all well above the 1.5–3.0 µV AGC threshold). Measure the AGC bus voltage at each level: it should become more negative at each step, and the S-meter should rise with it. The audio output level, monitored at the headphone output on a meter or by ear, should change far less than the 40 dB change at the input — holding it steady is what the AGC is for. A large change in audio level across this range means the AGC range is inadequate and further work on the chain is needed.

S-meter Calibration Check

Test: Once the AGC chain is confirmed functioning by the above tests, verify S-meter calibration with the signal generator. Per the Collins calibration procedure (service manual, alignment section): inject a reference signal and adjust the S-meter calibration potentiometer so that the meter reads S9 at the specified signal level for the installed antenna impedance. Do not attempt S-meter calibration until the AGC chain has been verified — calibrating the S-meter over a collapsed AGC will produce a calibration that is wrong once the AGC is repaired.

AGC Release Time Check

Test: Set the AGC switch to its fast position and tune in a steady carrier from the signal generator so that the S-meter reads well up the scale. Switch the generator off and watch the meter fall back, then repeat in the slow position. The specification gives a 1 ms attack in both positions and release times of about 190 ms (fast) and 600 ms (slow), so the meter should drop back quickly on fast and noticeably more slowly on slow. A release that is near-instant in both positions points to an open capacitor in the AGC network; a release that does not differ between the two positions, or that hangs far longer than this, points to the AGC switch or the components it selects (Candidates #1 and #3).

Frequently Asked Questions

What causes AGC collapse in a Collins 75S-3 receiver?

In a Collins 75S-3, AGC collapse means no AGC voltage reaches the RF and IF stages, so loud stations arrive at full gain while the S-meter barely moves. The 75S-3 has no AGC amplifier tube, because the AGC is rectified by a diode section of V9. The usual suspects are the front-panel AGC switch, V9 and its delay bias, the AGC filter network, and leakage or open circuits on the AGC bus and at the controlled grids.

How does the AGC work in the Collins 75S-3?

The Collins 75S-3 uses delayed AGC. IF signal from the secondary of T6 is rectified by a diode section of V9, whose cathode bias holds the AGC off until the signal reaches about 2 µV. The voltage passes through an RC network with fast and slow time constants to the RF amplifier V2 and the IF amplifiers, giving about 1 ms attack and about 190 ms or 600 ms release.

Why does the S-meter on a Collins 75S-3 barely move when the AGC has failed?

The Collins 75S-3 S-meter is not connected to the AGC bus. It is a bridge between the screen circuits of the V6 and V7 IF amplifiers and the V7 cathode, and it reads the drop in their current caused by AGC bias. If no AGC voltage reaches V6 and V7, the meter stays near zero even though the audio is loud.

Why does the S-meter rise when RF GAIN is turned down on a Collins 75S-3?

On the Collins 75S-3 the RF GAIN control sits in a divider across the −65 V bias line and places about 1 V of static bias on the AGC bus at maximum gain. Turning it down adds bias to the controlled stages, which reduces V6 and V7 current, so the bridge S-meter rises. This is normal behaviour and is not a test of the AGC chain.

How do you test the AGC on a Collins 75S-3?

Measure the AGC bus of the Collins 75S-3 with a high-impedance DVM. With no antenna and RF GAIN at maximum it should sit at the static bias of about −1 V, and with signals it should go more negative. A strong −40 dBm generator signal should drive it several volts negative, returning at the release rate when removed. A bus that stays at the static bias points to the AGC switch, V9 or the AGC network.

References and Notes

  1. Collins Radio Company, 75S-3, 75S-3A, 75S-3B, 75S-3C Receiver Service Manual. Available through Collins Collectors Association (CCA) at collinsradio.org/cca-collins-historical-archives/, Virtual Collins Radio Museum at wa3key.com, and the Collins S-Line owner community. Definitive source for tube types and positions (including V9), the AGC bus and its filter network values, IF stage grid circuit details, the S-meter bridge and calibration procedure, and differences between variants. All component designators in this document require cross-check against the applicable service manual edition before probing. Multiple service manual editions exist across the four variants; confirm the correct edition from the unit’s serial number and nameplate before beginning diagnostic work.
  2. Collins Collectors Association, CCA Technical Papers and Service Bulletins — 75S-3 Series, collinsradio.org. Collins service bulletins specific to the 75S-3 series, including any factory-issued guidance on AGC circuit anomalies. The complete Collins service bulletin index is maintained at vk6ada.com.au/collins-service-bulletin-index/.
  3. Collins Collectors Association, 75S-3B/3C equipment description, collinsradio.org. Source for the AGC circuit used throughout this guide: the delayed AGC rectified by a diode section of V9 from the T6 secondary, the R57 threshold (about 2 µV; 1.5–3.0 µV specified), the R88/C153 and R24/C50 time constants, the 1 ms attack and 190/600 ms release times, the R56 RF GAIN divider on the −65 V line, the J11 muting circuit, the V6/V7 bridge S-meter and the power supply voltages.
  4. Antique Radio Forums, Collins S-Line technical threads; eHam.net Collins 75S-3 reviews and forum. Owner and restorer discussion of 75S-3 AGC and S-meter faults and their repair.
  5. Collins Collectors Association, 75A-4 equipment description, collinsradio.org. Source for the 75A-4 details in the Section 3 and Section 7 comparison: the separate V-21 AVC IF amplifier, the V-16 twin-diode AVC rectifier, the AVC delay, the AVC-controlled stages (V-2, V-6, V-8, V-9) and the bridge S-meter.
  6. Mike Peace VK6ADA, Collins 75A-4 Receiver Failure Prevention Kit, vk6ada.com.au. Companion material for 75A-4 owners. The 75A-4 AVC details used in this guide’s comparison follow the CCA 75A-4 description (reference 5): the AVC amplifier is V-21 (V-11 is the CW/SSB detector), and the 75A-4 AVC, like the 75S-3 AGC, is delayed.
✍ Mike Peace VK6ADA  /  r-390a.net and foxtango.org Administrator  •  March 2026 vk6ada.com.au — Collins Radio Technical Resource