Heathkit HW-16 CW Transceiver
Audio Hum Fault Diagnosis — AF Gain Wiring, Grid Circuit, and Scope Validation
Root cause analysis of a volume-control-independent audio hum. Kelley W0RK’s grid-grounding test identifies the most probable injection zone as the high-impedance lead between the AF Gain pot wiper and V6A pin 2. This analysis covers seven candidates in priority order, a scope-based frequency identification procedure (60 Hz vs 120 Hz) that narrows the source type before any physical work, a definitive shielded-wire confirmation test, ground reference validation, and alternative causes not excluded by the grid-grounding test alone.
“I have a HW-16 that has some hum on the audio. The hum is volume control independent. It’s not filter capacitor hum (the radio was recapped using the Hayseed Hamfest kit) by a previous owner. I did some testing grounding the grids starting with V6B (audio power amp). With the grid (pin 7) grounded, there was no hum. Moving on to V6A (audio amp), with the grid grounded (pin 2), no hum. Moving on to V5A (product detector), with the grid grounded (pin 2), the hum was present. I’ve swapped out both V6 and V5 with a couple of NOS tubes, but the hum remains. So the hum is coming from somewhere to the left of V5A and the right of V6A (theoretically). A visual inspection of that area doesn’t show any obvious issues.”
Kelley’s grid-grounding test is textbook technique, and the diagnostic logic is sound. The three results together narrow the most probable injection point to the region between the AF Gain pot wiper and V6A’s grid (pin 2): specifically, the “volume control independent” observation confirms the hum enters after the pot wiper, not through V5A’s output. This is a credible and common fault zone in any tube audio stage operating at this node impedance (470 kΩ–1 MΩ), not unique to the HW-16.
However, the grid-grounding test alone does not prove a single cause or exclude all alternatives. It establishes the most probable injection zone. To confirm the source type before physical work begins, the first step should be a scope measurement at V6A’s grid: the hum frequency (60 Hz vs 120 Hz) determines whether the source is inductive/heater coupling or power supply ripple injected downstream of the filter electrolytics. This frequency discrimination changes which candidates to address first and in what order.
The most common cause of this exact symptom pattern in builder-assembled Heathkits is inductive pickup in the high-impedance lead from the pot wiper to V6A pin 2, combined with a drifted grid-leak resistor that raises the node impedance further. Both are addressable in under 30 minutes. But alternative causes — ground reference instability, tube socket carbon tracking, and downstream supply ripple — remain consistent with the available test data and must not be dismissed without a scope check.
Section 1 — What the Grid-Grounding Test Proves (and Does Not Prove)
The Test Results Are Correctly Interpreted
The grid-grounding test works by inserting a low-impedance short from each stage’s control grid to chassis ground, killing that stage’s input signal path. If the hum disappears, it was entering at or before that grid. If it persists, it is entering downstream. The AF Gain potentiometer, which sits between V5A’s output and V6A’s grid, is the pivotal element: hum that enters before the pot wiper is attenuated by the volume control (volume-dependent); hum that enters after the wiper bypasses the volume control entirely (volume-independent). Kelley’s three results are mutually consistent and correctly triangulate the most probable injection zone.
Stage |
Grid Grounded |
Result |
Interpretation |
|---|---|---|---|
| V6B pin 7 | 6HF8 pentode grid | No hum ✓ | Source is upstream of V6B |
| V6A pin 2 | 6HF8 triode grid | No hum ✓ | AC is riding on the lead to V6A pin 2; shorting eliminates it. Most probable injection zone confirmed. |
| V5A pin 2 | Product det. grid | Hum present ✗ | Injection is after the AF Gain pot; grounding V5A stops V5A’s output but hum enters beyond the wiper. Consistent with post-pot injection — does not exclude all other causes. |
Section 2 — Scope-Based Frequency Identification — Do This First
The single most useful measurement before any physical work is a scope observation at V6A’s grid. The hum frequency tells you which source class you are dealing with. This takes two minutes and can save an hour of physical investigation by directing effort to the right candidates.
60 Hz vs 120 Hz: What Each Means
Frequency Observed |
Most Likely Source Type |
Priority Candidates |
|---|---|---|
| 60 Hz | Inductive pickup from transformer or heater bus. Heater-to-cathode leakage. Ground loop with AC reference. | Candidates #1 (wire dress), #3 (heater leads), #5 (ground reference), #6 (socket tracking) |
| 120 Hz | Power supply ripple injected downstream of the filter electrolytics — a bypass cap on the B+ rail serving V6A is open or missing. | Candidates #7 (downstream supply bypass). Also probe B+ at V6A plate pin — ripple there despite the recap. |
| Both 60 + 120 Hz | Multiple contributing sources. Common when wire dress is poor AND a bypass cap is open. | Address each class separately. Start with the dominant component. |
How to Make the Measurement
Set the oscilloscope to AC coupling, 20–50 mV/div vertical, 10 ms/div horizontal (10 ms/div places exactly one complete 60 Hz cycle across the screen, making the frequency immediately visible). Ground the scope probe clip to chassis. Using a 10:1 probe, touch V6A’s grid pin (pin 2 on the compactron socket) while the receiver is powered and tuned to a quiet frequency. One-handed probe technique. Observe the waveform: a clean sinewave at one full cycle per screen = 60 Hz; two cycles per screen = 120 Hz. Note the peak-to-peak amplitude. A reading above 5 mV pk-pk is audible through V6A’s gain; above 50 mV is strongly audible. Record both frequency and amplitude as a baseline before any physical work — this lets you verify improvement after each change.
Section 3 — Seven Root Cause Candidates in Priority Order
Candidates #1 through #4 address the post-pot injection zone identified by Kelley’s test. Candidates #5, #6, and #7 address causes that are also consistent with the available test data but are less immediately indicated. The scope frequency measurement (Section 2) determines which group to prioritise.
-
#1
Wire dress: pot wiper to V6A pin 2 near AC-carrying conductors MOST PROBABLE — 60 Hz result
The unshielded wire from the AF Gain pot wiper to V6A pin 2 sits at source impedance in the hundreds of kilohms. Any section running within a few centimetres of the power transformer, heater bus, or B+ wiring picks up 60 Hz inductively. In builder-assembled Heathkits, lead dress is the single most common cause of this type of fault. The fix is to dress this wire tightly along the chassis (minimising the loop area) or to substitute a short length of shielded cable with the braid grounded at the pot end only.
Definitive powered test: tack a piece of RG-174 coax or screened wire from the pot wiper terminal to V6A pin 2 using clip leads. Ground the braid to chassis at the pot end only. If hum drops significantly: wire routing is confirmed as the dominant path. Permanently re-dress the original wire along the chassis edge away from all AC conductors, or install the shielded lead.
-
#2
V6A grid-leak resistor drifted high (carbon comp, 470 kΩ–1 MΩ) CONTRIBUTING FACTOR — address regardless of frequency
A drifted grid-leak resistor raises the impedance of the V6A grid node, increasing the induced hum voltage from any given stray field. A 470 kΩ part that has drifted to 700 kΩ raises the pickup by nearly 50%. This is a contributing factor that amplifies the effect of every other candidate, not an independent cause on its own. Replace with 1% metal film at the nominal value: node impedance drops to design level and sensitivity to inductive coupling is minimised.
Check (power off, one end lifted): measure out of circuit. Replace if more than 20% above nominal value.
-
#3
Heater lead dress at the V6 (6HF8) compactron socket SECONDARY — 60 Hz result, especially if recap moved heater wiring
The 6HF8 compactron places heater pins physically adjacent to the triode grid pin on the 12-pin socket. Untwisted or looped heater leads near the V6 socket couple 60 Hz inductively into the grid connection by proximity, independent of where the pot-to-V6A lead is routed. This is a separate induction path from Candidate #1 and can produce residual hum even after a well-executed shielded wire fix if the heater leads are not also corrected.
Check (power off): examine heater leads at the V6 socket. They must be twisted-pair at approximately one twist per centimetre and dressed flat against the chassis immediately from the socket. Any slack loop within 3 cm of the socket should be re-twisted and re-dressed.
-
#4
Coupling capacitor from AF Gain pot to V6A pin 2 — not in Hayseed kit LOWER PROBABILITY — inspect while in the area
A small coupling capacitor (film or ceramic disc) may be present between the pot wiper and V6A’s grid. This type of cap was not replaced by the Hayseed Hamfest electrolytic recap. A leaky coupling cap is worth replacing while the circuit is already open, but note the failure mode: a leaky cap more commonly introduces DC offset and bias shift causing distortion than it produces 60 or 120 Hz hum specifically. It is a lower probability primary cause but remains a reasonable maintenance item given its age.
Replace with: 0.022 µF or 0.047 µF polypropylene film, 400 V or greater. Keep leads short.
-
#5
Ground reference instability — AF Gain pot lug or chassis return path SECONDARY — 60 Hz result, explains floating-pot hum
Two ground-related faults are possible. First: the AF Gain pot’s ground lug. If the pot body-to-panel contact or ground wire solder joint is open or corroded, the pot bottom floats on whatever AC voltage exists at that point in the chassis, turning the entire resistive string into a hum source. Second: if the cathode return path for V6A (or the associated bypass cap ground) has a high-impedance connection to chassis due to a lifted pad or oxidised lug, the common-mode return becomes an antenna for supply-frequency hum. Both are especially likely if the recap involved re-soldering near these points.
Check (power off): measure resistance between pot ground terminal and chassis — must be under 1 Ω. Also probe the V6A cathode bypass cap ground lead and the V6A grid-leak resistor ground return; both must show clean DC continuity to chassis.
-
#6
Tube socket carbon tracking or pin-to-pin leakage at V6 LESS LIKELY — check if all else fails or if unit has been stored damp
The 6HF8 compactron socket, particularly in units stored in humid or dusty conditions, can develop carbon tracking or resistive contamination between adjacent pins. The heater pins are at supply potential (6.3 V AC) and physically close to the grid pin. A contaminated socket presents a resistive path from the heater pins to the grid connection, coupling heater AC (60 Hz) directly into the grid circuit regardless of wire dress. This failure mode produces 60 Hz hum that is unaffected by shielding the wiring, which is a diagnostic clue: if Candidates #1–#3 are all addressed but hum persists at 60 Hz, inspect the socket.
Check (power off, tube removed): inspect the V6 compactron socket under magnification for carbonised residue or contamination between adjacent pin positions. Clean with 99% IPA on a stiff cotton swab. Measure resistance between heater pins and the grid pin; should be well above 10 MΩ. Any reading below 1 MΩ indicates a contaminated socket requiring cleaning or replacement.
-
#7
Downstream supply ripple on B+ rail feeding V6A — bypass cap not in recap kit SECONDARY — 120 Hz result specifically
The Hayseed Hamfest kit replaces electrolytic filter capacitors. It does not replace small ceramic disc or film bypass capacitors on individual stage B+ supply rails. A small disc ceramic (typically 0.01–0.047 µF) should bypass the B+ supply rail at or near V6A’s plate load resistor. If this cap is open-circuit (a common failure mode in 60-year-old ceramic discs, with no visual indication), 120 Hz power supply ripple appears at V6A’s plate and couples through the output coupling cap to V6B’s grid. This produces 120 Hz hum with a different character from inductive 60 Hz pickup.
Scope check: measure AC ripple at V6A’s plate pin (AC coupled, 10× probe) while the receiver is powered. A 120 Hz component above 50 mV indicates an open bypass cap at this node. Replace with a 0.01 µF or 0.047 µF X7R ceramic or polypropylene film cap, 630 V rated, as close to the plate load resistor B+ feed as possible.
Section 4 — Recommended Check Procedure
-
1
Powered: scope measurement at V6A grid (pin 2) — frequency and amplitude baseline AC couple the oscilloscope, 10× probe, 20–50 mV/div, 10 ms/div. Probe V6A pin 2. Record: frequency (60 Hz or 120 Hz or both), peak-to-peak amplitude. This single measurement directs the rest of the investigation. 60 Hz → proceed to Steps 2–4 first. 120 Hz → go to Step 6 first. Both → address each separately, starting with the dominant component.
-
2
Power off: check pot ground lug and V6A cathode return continuity (Candidate #5) Measure pot ground terminal to chassis (must be <1 Ω). Measure V6A cathode bypass cap ground lead to chassis (must be <1 Ω). Clean and re-solder any high-resistance connection. Quick two-minute check that eliminates floating-reference faults before physical wire work begins.
-
3
Power off: measure V6A grid-leak resistor out of circuit (Candidate #2) Lift one end, measure. If >20% above nominal, replace with 1% metal film at nominal value. This lowers node impedance to design level before the wire dress test, ensuring the shielded-wire result is representative of what the corrected design should achieve.
-
4
Power off: inspect and dress heater leads at V6 compactron socket (Candidate #3) Heater leads must leave the socket as twisted-pair and lie flat against the chassis immediately. Any slack or untwisted loop within 3 cm of the socket: re-twist and re-dress. This eliminates a separate induction path that is independent of wire routing.
-
5
Powered: shielded-wire substitution test — pot wiper to V6A pin 2 (Candidate #1) With receiver on, clip shielded cable (RG-174 or equivalent) from pot wiper to V6A pin 2. Braid to chassis at pot end only. Monitor hum while inserting the cable. If hum drops: confirm by removing and re-inserting the cable. Permanent fix: re-route original wire along chassis edge away from all AC conductors, or install the shielded lead permanently. Verify with scope that hum amplitude at V6A pin 2 has dropped to below 5 mV pk-pk.
-
6
Powered: scope check at V6A plate if 120 Hz component present (Candidate #7) If Step 1 revealed 120 Hz: probe V6A plate pin (AC coupled, 10× probe). If more than 50 mV of 120 Hz ripple is present, the B+ bypass cap at this node is open. Locate the small disc ceramic bypass cap on the B+ supply to V6A’s plate load resistor on the schematic. Remove and measure it out of circuit; replace with 0.047 µF 630 V film or X7R ceramic.
-
7
Power off: inspect V6 compactron socket if hum persists after Steps 1–6 (Candidate #6) Remove V6. Inspect socket pins under magnification for contamination or carbon tracking between heater and grid pins. Clean with 99% IPA. Measure heater-to-grid-pin resistance: should be >10 MΩ. Replace socket if measurement is below 1 MΩ after cleaning. Also inspect and replace the coupling cap to V6A grid at this stage (Candidate #4).
Section 5 — Signal Path with Hum Entry Zones Annotated
┌──────────────────────────────────────────────────────────────────────────┐
│ HEATHKIT HW-16 — AUDIO CHAIN WITH HUM INJECTION ZONES ANNOTATED │
│ Frequency identification (Step 1) determines which zone to address │
└──────────────────────────────────────────────────────────────────────────┘
[V5A product detector plate]
│
[coupling cap]
│ ← hum here = volume-control DEPENDENT (not this fault)
▼
╔══════════════════╗
║ AF GAIN POT ║ ← ground lug: check continuity (Candidate #5)
║ wiper ──────────╫──────────────────────────────────────────────►
╚══════════════════╝
│
┌────────────────────────┼──────────────────────────────────────────────┐
│ ★ PRIMARY HUM INJECTION ZONE (confirmed by grid-grounding test) │
│ │ │
│ 60 Hz sources: │ → Candidate #1: wire near transformer/ │
│ Inductive pickup │ heater bus (most probable, wire dress) │
│ from AC conductors │ → Candidate #3: heater loop at V6 socket │
│ │ → Candidate #5: ground reference float │
│ │ → Candidate #6: socket carbon tracking │
│ │ │
│ 120 Hz source: │ → Candidate #7: B+ bypass cap open at │
│ Supply ripple via │ V6A plate supply node (separate path) │
│ open bypass cap │ │
└────────────────────────┼──────────────────────────────────────────────┘
│
[coupling cap] ← Candidate #4 (lower priority)
│
[grid-leak R] ← Candidate #2: drifted carbon comp
│ raises node Z, amplifies all pickup
▼
┌───────────────────────────────────────────────────────────────┐
│ V6A 6HF8 triode section │
│ pin 2 = grid ◄── SCOPE HERE (Step 1 baseline measurement) │
│ ◄── GROUNDED IN TEST → NO HUM ✓ │
└───────────────────────────────────────────────────────────────┘
│
[coupling cap]
│
┌───────────────────────────────────────────────────────────────┐
│ V6B 6HF8 pentode section │
│ pin 7 = grid ◄── GROUNDED IN TEST → NO HUM ✓ │
└───────────────────────────────────────────────────────────────┘
│
[output transformer]
│
SPEAKER
─────────────────────────────────────────────────────────────────────────
FREQUENCY-TO-CAUSE MAP
┌───────────────────────────────────────────────────────────────────────┐
│ Scope shows 60 Hz at V6A pin 2 → inductive/heater: #1, #3, #5, #6 │
│ Scope shows 120 Hz at V6A pin 2 → supply ripple: candidate #7 │
│ Scope shows both → multiple sources: address both │
└───────────────────────────────────────────────────────────────────────┘
─────────────────────────────────────────────────────────────────────────
SHIELDED-WIRE TEST — DEFINITIVE FOR INDUCTIVE PICKUP CONFIRMATION
┌───────────────────────────────────────────────────────────────────────┐
│ Substitute RG-174 or screened wire: pot wiper → V6A pin 2 │
│ Braid to chassis at POT END ONLY (not V6A end) │
│ Hum drops significantly → wire routing confirmed as dominant cause │
│ Hum unchanged or minor → check Candidates #5, #6, and #7 │
└───────────────────────────────────────────────────────────────────────┘
─────────────────────────────────────────────────────────────────────────
ADVANCED: HEATER ELEVATION (if 60 Hz hum persists after all above)
In some HW-16 installations, applying a small positive DC bias (~20-30V)
to the heater supply reference (centre-tap to a DC voltage divider from
B+) raises the heater potential above cathode potential, reducing H-K
leakage current. This is an advanced technique documented in vintage tube
audio practice. Consult W8JI.com for the specific circuit implementation
before attempting, and verify compatibility with the HW-16 heater supply
topology from the service manual.
HW-16 audio chain with hum injection zones and frequency-to-cause map. Scope measurement at V6A pin 2 (Step 1) is the gateway measurement that directs all subsequent checks. Verify all component designators against the HW-16 service manual before probing.
Section 6 — Verification After Each Fix
Scope Amplitude at V6A Pin 2 — Before and After
Listening Verification After All Scope-Confirmed Fixes
References and Notes
- Kelley W0RK, fault report to [email protected], March 2026. Grid-grounding diagnostic methodology and test results as described. The circuit topology analysis, frequency discrimination framework, and root cause candidates in this document are by the author (Mike Peace VK6ADA) incorporating community input and technical review of the HW-16 audio chain.
- Heathkit Company, HW-16 CW Transceiver Assembly and Service Manual. Essential for verifying tube type designations (V5, V6), pin assignments, AF Gain pot wiring topology, coupling and bypass capacitor designators and values, and grid-leak resistor nominal values. Available at Heathkit Virtual Museum (heathkitmuseum.org), Antique Radio Forums document archive, and BAMA Boatanchor Manual Archive (bama.edebris.com/manuals/heath/). All component designators cited in this document must be cross-checked against the specific unit’s schematic before proceeding.
- Hayseed Hamfest, Heathkit HW-16 Electrolytic Recap Kit, hayseedhamfest.com. The Hayseed kit replaces electrolytic filter capacitors. Small ceramic disc bypass caps on individual stage B+ rails, grid circuit coupling caps, and carbon composition resistors are outside its scope and require separate attention during complete restoration.
- GE Compactron Data, 6HF8 triode-pentode, 12-pin compactron. The 6HF8 combines a high-mu triode (V6A) and output pentode (V6B). Heater pins are physically adjacent to the triode grid pin on the compactron socket. Heater-to-grid proximity makes twisted-pair heater dress mandatory at this socket position.
- W8JI (Tom Rauch), Hum in valve audio: inductive pickup, wire dress, heater elevation, ground return paths, w8ji.com. Authoritative reference for all techniques in this document: shielded-lead grounding (source-end braid only), heater twist requirements, star-ground vs chassis-return topologies, heater elevation for H-K leakage reduction, and grid-leak resistor impedance effects on pickup sensitivity.
- [email protected] community threads on HW-16 audio hum; Antique Radio Forums HW-16 restoration discussions. Community documentation confirms: AF Gain wiper-to-V6A-grid lead dress is the most frequently cited successful fix for volume-control-independent hum after a full recap; drifted grid-leak resistors are a commonly found contributing factor; 6HF8 socket tracking has been reported in units with storage-related contamination.