Collins KWM-2/2A Known Issues

The Collins KWM-2 (1959–1968) and KWM-2A (1968–1981) represent the pinnacle of Collins vacuum-tube transceiver design — 18 tubes and 4 semiconductor diodes packed into a compact chassis where virtually every tuned circuit and several tubes serve dual roles in both transmit and receive. This shared-circuit architecture that makes the KWM-2 so compact also makes it the most complex radio in the S-Line family to troubleshoot. The first 25 production units all failed to work — a testament to the design’s sensitivity to component tolerances and circuit interaction.[1]

This guide consolidates the KWM-2/2A-specific known issues — the transceiver problems that do not appear (or appear differently) in the separate 75S-3 receiver and 32S-3 transmitter. Issues common to all S-Line equipment (Black Beauty caps, bandswitch oxidation, frozen trimmers, PTO service) are summarised here with cross-references to the detailed companion guides on this site.

Symptom: The K2 relay chatters or “machine-guns” when PTT is pressed. Or: the relay pulls in, releases, and oscillates in and out. Adjusting the ANTIVOX GAIN CW control may temporarily stop the oscillation. In severe cases, the relay chatter produces an RF transient on each cycle visible at the antenna output.[1]

Root Cause: The KWM-2 VOX/keying system uses a voltage summing network where positive DC from the VOX rectifier (microphone audio) sums against negative DC from the ANTIVOX rectifier (speaker audio) to control the relay actuator V4B. When VOX GAIN and ANTIVOX GAIN are properly balanced, speaker audio cancels the relay drive and the K2 relay is not triggered by its own transmitted audio returning through the speaker. When this balance is disturbed — by relay transients feeding back through the R3 ground lead to emission switch S9C, by leaky Black Beauty timing capacitors, or by drifted component values in the summing network — the relay oscillates.[1]

Fix: Verify all service bulletins have been applied (SB-2: adjustable VOX time constant R43; SB-4: relay contact suppression). Add 0.01 µF capacitor from V1 pin 7 to ground to suppress relay transients. Add 0.1 µF across K2 contacts 12 and 13 to reduce VOX instability. If R199 (2.2 MΩ) has been added per late-revision schematics, verify its value has not drifted. Replace Black Beauty timing capacitors in the VOX circuit. For persistent chatter, refer to the Carns KWM-2/2A Keying Circuit tutorial published through the CCA.[2]

Symptom: Intermittent transmit or receive; erratic S-meter readings; sensitivity varies when the chassis is tapped; the relay clicks but one or more circuits do not switch properly. In plug-in relay conversions: intermittent problems traced to poor wiring quality at the relay socket.[3]

Root Cause: The KWM-2 uses multiple relays (K2, K3, K4) that switch between transmit and receive functions. After decades of T/R cycling, the relay contacts develop oxidation and pitting. Early KWM-2 units use open-frame relays soldered directly to the chassis; later units and field conversions use plug-in relays in sockets. Plug-in relay conversions are sometimes poorly wired — one restoration found “gobs of solder and burnt wiring” at the relay socket.[3]

Fix: For open-frame relays: clean each contact individually with paper moistened with contact cleaner, then apply a micro-drop of DeoxIT. For plug-in relays: clean the socket contacts and relay pins with DeoxIT; inspect the wiring for cold solder joints and heat damage. If contacts are severely pitted, relay replacement is necessary. For plug-in relay conversions with questionable wiring, consider rewiring the relay sockets completely on a purpose-made PCB or tag strip.[3]

Symptom: A brief RF output spike (~5 ms duration) occurs at the moment of switching from receive to transmit, even with MIC GAIN at OFF. Visible on an oscilloscope monitoring the antenna output during T/R transitions.[4]

Root Cause: When switching to transmit, V3A (cathode follower feeding the balanced modulator) receives its anode voltage from the T+275V supply and immediately begins drawing current through R12, rapidly charging C6 from zero to several volts. This voltage transient couples through R208, L41, and L38 directly into the balanced modulator, where it is converted to an RF signal at the PA output. This transient is independent of the carrier balance null setting — it is a switching artifact, not a carrier leakage problem.[4]

Fix: Install C264 (4 µF, later increased to 22 µF per ASAB 1016) between V3 pin 8 and ground. Add CR10 (diode) to prevent C264 charge from flowing back to the T+275V line. Install R122 (47 kΩ, 2W) from turret E40 terminal D to V3 pin 7 to bias the cathode follower off in receive mode, cutting the audio feedthrough path. These three modifications were progressively incorporated during KWM-2 production and are documented in the ASAB service data.[4]

Symptom: ALC meter will not zero; meter drifts during warm-up; excessive S-meter and ALC meter zero drift; grid current reading falls to zero in LOCK KEY mode even with MIC GAIN advanced.[2]

Root Cause: Some 6AZ8 tubes (V1, V3) develop positive grid voltage that corrupts both the AVC and ALC lines. If the no-signal voltage at the grid of V1 or V3 differs from the AVC or ALC line by more than 0.1V, the tube should be replaced. Additionally, 60 Hz ripple on the -70V bias line couples into the ALC meter via R189, causing the zero to oscillate. R170 (220Ω across the ALC zero pot) may limit the zeroing range when component drift has moved the operating point outside the original design window.[2]

Fix: Replace gassy 6AZ8 tubes (check V1 and V3 grid voltage vs. AVC/ALC line voltage). Add a 20 µF electrolytic across the -70V bias line at the input of R189 to ground to filter 60 Hz ripple from the ALC meter. If the ALC meter will not zero, remove R170 (220Ω) to extend the zeroing range. Refer to the Don Jackson analysis “Dealing with 32S-3 ALC Zero Adjust Instability” (applicable to the KWM-2 as well) published through the CCA.[2]

Symptom: Excessive frequency drift, particularly during the first 30 minutes of operation. Frequency instability under varying load conditions. Drift worsens after tube changes in other parts of the transceiver.[3]

Root Cause: The 70K-2 PTO is the standard S-Line PTO and requires periodic cleaning and re-lubrication (see companion PTO Overhaul guide). However, the KWM-2 has an additional problem not shared with the 75S-3 or 32S-3: Collins did not regulate the B+ supply to the VFO. Any variation in the T+275V supply — from tube aging, warm-up, or load changes during T/R switching — directly modulates the PTO frequency.[3]

Fix: Service the 70K-2 PTO per the companion guide (clean, re-lubricate, replace C302/C303/C304/C306/C309/C310 and CR301). Additionally, install a 150V zener diode to regulate the VFO B+ supply. This is a widely adopted modification that significantly improves frequency stability, particularly during T/R switching transitions and during warm-up.[3]

Symptom: CW signal is transmitted 1.5 kHz (early) or 1.75 kHz (late) above the dial frequency. There is no RIT control for offset tuning on receive. The dial reading does not correspond to the actual transmitted CW frequency.[3]

Root Cause: The KWM-2 generates CW by injecting a 1.5/1.75 kHz audio tone into the balanced modulator, which shifts the transmitted CW frequency above the dial reading by the tone frequency. Unlike the 32S-3 (which uses carrier-generated CW), the KWM-2 also produces a small keyed opposite-sideband output in addition to the normal IMD products and suppressed carrier. There is no factory-fitted RIT or variable BFO to offset the receive frequency.[3]

Fix: This is a design characteristic. For operators who require RIT, Gerd Borowski DJ7HS designed a clean, reversible RIT modification using the existing CR301 varicap diode in the 70K-2 PTO. The modification uses the unused NB switch position to enable RIT and includes a stabilised 20V supply from the TR+275V rail. Several other RIT circuits have been published, dating back to Nick Taylor K5YTO’s 1963 CQ article.[4]

Symptom: A squeal or “yelp” is emitted from the speaker immediately after the transceiver returns from transmit to receive. Reducing MIC GAIN momentarily stops it, but it may return at the end of the next transmission. Most noticeable with RF GAIN at maximum and AF GAIN set high.[2]

Root Cause: A feedback loop exists from the microphone through the balanced modulator, into the receiver first IF amplifier, and out the speaker. The cathode follower V3A remains partially conductive in receive mode, allowing audio to feed through the balanced modulator into the IF strip.[2]

Fix: Install R122 (47 kΩ, 2W, CPN 745-5722-000) from turret E40 terminal D to V3 pin 7. This applies positive voltage to the cathode of the microphone amplifier cathode follower in receive mode, cutting off the audio feedthrough. The same R122 that addresses the T/R switching transient (Issue 3) addresses this squeal — both modifications are part of the progressive ASAB improvements.[2]

The following resistors have been documented as high-failure items in the KWM-2/2A. All looked physically normal at visual inspection but had drifted significantly from their marked values:[5]

PL-068 Plugs: The brass PL-068 plugs used on many Collins interconnect cables oxidise and develop a tarnished film that increases contact resistance. Polish the plug contacts periodically with 000 steel wool or a fine 3M abrasive micro-pad (purple grade — green is too abrasive). This is a common cause of “no transmit” or “no receive” symptoms when the KWM-2 is reconnected after storage.[1]

Bandswitch: The same silver contact oxidation that affects all S-Line receivers applies to the KWM-2’s bandswitch. Give the bandswitch a workout — rotating through all positions 20–30 times — periodically, as contact oxidation from extended storage will cause intermittent drive or idle current problems. DeoxIT dropper-bottle treatment per the standard S-Line procedure is recommended during any service interval.[2]

1. Carns. KWM-2/2A Keying Circuit — Functional Description and Issues. Complete VOX/ALC/keying chain analysis: three keying modes, relay chatter root cause, summing network theory, ANTIVOX operation, SB evolution, PL-068 connector oxidation. CCA publication. collinsradio.org — Keying Issues (PDF)
2. VK3KCM. KWM-2/2A Service Information. Compiled ASAB service data: gassy 6AZ8 diagnosis, VOX relay fixes (0.01 µF, 0.1 µF), C264 ripple suppression, R122 audio feedthrough, R170 ALC range, R143 plate dissipation, R175 failure, bandswitch maintenance. angelfire.com/de/vk3kcm — KWM-2 Service
3. Antique Radio Forums. KWM-2 Gotchas? Unregulated VFO B+ (150V zener fix), CW sidetone offset, plug-in relay conversion wiring quality, PTO service requirement, service bulletin checklist. April 2019. antiqueradios.com — KWM-2 Gotchas
4. Borowski, Gerd DJ7HS. My Collins KWM-2. V3A cathode transient oscilloscope analysis, C264/CR10/R122 modification sequence, RIT modification using CR301 varicap, PA neutralisation upgrade, Waters rejection module replacement. CCAE — DJ7HS KWM-2 (PDF)
5. VK3KCM. KWM-2/2A Additional Service Information. Resistor drift table (R4, R7, R148, R175, R50, R20/R47), AGC time constant modification, 516F-2 solid-state rectifier warning, plug-in relay conversion notes. angelfire.com/de/vk3kcm — KWM-2A Info
6. Collins Reflector. KWM-2A Problems and Tubelore. Screen resistor check (R146, R69, R70), PA compartment oscillator leakage diagnosis, preselector alignment vs. oscillator frequency, C49 at V2B sidetone suppression, Dennis Brothers video references. October 2004. Collins Reflector — KWM-2A Problems
7. F1LAG. KWM-2 Restoration and Updates. Comprehensive service bulletin cross-reference, ALC redesign history (zeroing pot location changes), R150/R207 modifications, schematic change tracking across manual editions. F6HOY — F1LAG KWM-2 Restoration (PDF)
8. EB5AGV. KWM-2A Repair Notes. Grid current falling to zero in LOCK mode (ALC issue), relay contact cleaning after 10+ year storage, alignment sequence. jvgavila.com — KWM-2A Repair
9. Collins KWM-2/2A Instruction Book, 9th Edition (January 1978). Complete circuit description, alignment procedures, service bulletin index (Section 7), voltage/resistance tables. collinsradio.org — KWM-2/2A Manual Section 4 (PDF)

Carns — For the definitive KWM-2/2A Keying Circuit tutorial that explains the three keying modes, the ALC/VOX summing network interaction, and the most common relay faults with their service bulletin history.

Gerd Borowski, DJ7HS — For the oscilloscope-documented T/R switching transient analysis showing the V3A cathode charging path through the balanced modulator, and for the clean, reversible RIT modification.

VK3KCM — For compiling the KWM-2/2A service information including the resistor drift table that identifies the specific components most likely to have drifted after decades of service.

F1LAG — For the meticulous service bulletin cross-reference tracking the ALC circuit redesign across manual editions.

Rodger WQ9E — For extensive troubleshooting contributions across multiple forums covering PA neutralisation, grid current diagnosis, and relay wiring quality in plug-in conversions.

Collins Collectors Association (CCA) — For maintaining the “RX For Your Collins” technical library, the service bulletin archive, and the instruction manuals.