Collins 75A-3 · Service Note · Power Supply & Bias Locating and Replacing C-96, the Back-Bias Bypass Capacitor The one 0.1 µF paper capacitor the factory manual shows on the schematic but never calls out in the underside photograph — and how to identify it by its two connections rather than by its appearance.

Quick answer C-96 is a 0.1 µF non-polarised paper bypass capacitor connected between the receiver’s negative back-bias rail and chassis. The February 1953 instruction book shows it on the main schematic (Figure 5-6) but the bottom-view component photograph does not label it. The reliable way to find it is to work from the power-supply corner of the chassis — the rectifier, the power transformer and the multi-section filter can — trace the isolated negative filter-can terminal into the bias wiring, and then confirm that the capacitor’s other lead lands on a chassis-ground lug. Two correct nodes identify C-96; colour and body style do not.

1. Why C-96 Is Hard to Find in the Manual

If you have been through the 75A-3 instruction book looking for C-96 and come up empty, you have not missed a label. Section V of the February 1953 book gives a top-view component drawing and a bottom-view component photograph, and the bottom view simply does not call out C-96. The main receiver schematic, Figure 5-6, does show it: a 0.1 µF capacitor in the negative-bias section of the power supply.^[1]

The 75A-3 is electrically the 75A-2 with provision for plug-in mechanical filters; the receivers are otherwise so close that many 75A-2 sets were factory- or field-converted and re-badged 75A-2A, and at least one capacitor-kit supplier notes only a single component difference between the two.^[2] That makes the 75A-2 parts list useful corroboration. It has been reported as describing C-96 as the “back bias by-pass”, 0.1 µF paper, 150 WV; treat that as supporting evidence and confirm it against your own copy of the 75A-2 book, because the CCA scans are images rather than searchable text.^[3]

Also worth knowing before you compare your chassis with anyone else’s photographs: Collins changed the first RF amplifier V-1 from a 6CB6 to a 6DC6 in receivers above serial number 1573, so not every 75A-3 underside looks identical even from the factory.^[4]

2. Why a Tired C-96 Matters

The 75A-2/75A-3 develops its negative bias the classic way: the power-supply return current flows through a string of bias load resistors (R-36, R-37 and R-38 in the Collins text) in the negative lead before reaching chassis, so the negative end of the filter can sits below ground. The manual notes that this network supplies roughly 11 V of cut-off bias to the AVC amplifier, the plate voltage for the AVC amplifier, and the fixed grid bias for the 6AQ5 audio output tube. The standby line inserts R-76 (1500 Ω) in series with the same resistors to raise the bias and cut the output tube off.^[5]

C-96 bypasses that bias rail to chassis for hum and audio. It carries the full bias voltage continuously, and an original wax-paper part of this age is very likely to be leaky. A leaky C-96 shunts the bias divider, which reduces the bias on the 6AQ5 and the AVC amplifier. The symptoms are the familiar ones: excessive output-tube plate current, a warm output transformer, low-frequency hum that no amount of B+ filter work cures, and AVC that never seems to be quite right. A shorted C-96 removes the bias altogether. It is a small part with an outsized effect on the receiver’s health.

Key identifiers and landmarks

Item	What it tells you
C-96	0.1 µF non-polarised bypass on the negative bias rail. Original part is a tubular paper capacitor. No polarity to observe.
C-95	Multi-section power-supply electrolytic can. Its common/negative terminal is deliberately insulated from chassis because it is the point where the negative bias is developed. The key tracing landmark.
C-94	Back-bias filter electrolytic. A polarised part in the same circuit family; a useful nearby landmark and a second suspect if the bias rail reads wrong.
5Y3GT rectifier	The 75A-3 uses a single 5Y3GT rectifier.[6] Its octal socket is the obvious physical anchor for the power-supply corner from either side of the chassis. Confirm its V-number from the tube chart in your own manual.
T-9 power transformer	Easy visual reference from the top-view drawing; its secondary leads come through to the underside close to the rectifier socket and filter can.
Negative bias rail	The schematic labels the bias bus with its nominal voltage (commonly quoted as about −50 V at the filter-can end). Treat the printed figure as nominal; C-96 bypasses this node to chassis.

3. Safety Before Any Rework

⚠ High-voltage warning The 75A-3 is a mains-powered, transformer-fed vacuum-tube receiver drawing about 85 W. Its B+ rail runs to several hundred volts and the negative bias rail sits tens of volts below chassis; the filter capacitors hold that charge after the set is switched off. Unplug the receiver, discharge the supply capacitors through a resistor, and verify with a meter that both the B+ and the bias rail read near zero relative to chassis before touching the underside wiring. Work on this class of equipment only if you are comfortable servicing transformer-powered tube gear.

Safe discharge method

1. Unplug the receiver from the mains. Do not rely on the front-panel switch.
2. Clip a 47 kΩ to 100 kΩ, 5 W resistor on insulated leads between the B+ side of the filter can and chassis. Hold it there long enough for the stored charge to fall — ten to fifteen seconds is plenty at these values.
3. Repeat between the negative bias rail and chassis.
4. Confirm both rails with a digital multimeter. If a voltage creeps back after discharge, discharge again and investigate before proceeding.
5. Never discharge a filter capacitor with a screwdriver short. It pits the terminal, stresses the capacitor, and tells you nothing about what is left in it.

Recommended bench setup

• Isolation transformer for the receiver under service. The 75A-3 has its own power transformer, but isolation is still good practice when test equipment is connected.
• Digital multimeter rated for the voltages involved; insulated probes or mini-grabber leads.
• 47 kΩ to 100 kΩ, 5 W discharge resistor with insulated leads and clips.
• Temperature-controlled soldering iron, solder wick or a vacuum desoldering tool, small hemostats, and heat-shrink or cloth sleeving.
• Phone or camera for “before” photographs, and a marker or tags to identify the two C-96 connection points.
• A dim-bulb tester or variac for the first post-rework power-up. Use it as a fault-limiting check, not as the final operating-voltage test.

4. Locating C-96 Without Guessing

Do not hunt the RF deck. C-96 belongs with the back-bias and power-supply wiring, in the corner of the chassis under the rectifier, the power transformer and the filter can. Within that corner, identify it by its electrical signature. The simplified node diagram below is an explanatory redraw, not a substitute for Collins Figure 5-6.

C-96 electrical locator — simplified from the power-supply schematic
C-95 filter can common / negative terminal insulated from chassis	C-94 back-bias filter polarised electrolytic same node family	Bias load resistors R-36 / R-37 / R-38 (R-76 in standby)
NEGATIVE BACK-BIAS NODE  —  nominally about −50 V to chassis
C-96  ·  0.1 µF  ·  non-polarised paper bypass one lead on the bias node above  ↓  one lead on chassis below
CHASSIS GROUND

Figure 1. Electrical tracing method: one C-96 terminal belongs to chassis; the other belongs to the negative bias node shared with the isolated C-95 common/negative terminal. Use this only to identify the node; consult Collins Figure 5-6 for the complete circuit.

Step-by-step tracing procedure

1. Unplug the receiver, remove it from the cabinet, discharge the supply as above, and verify zero voltage with the DMM.
2. Use the top-view component drawing in Section V to find the filter can, the 5Y3GT rectifier and T-9 from the top. This anchors the power-supply corner before you turn the chassis over.
3. Turn the receiver upside down on padded supports so the front-panel controls and the dial mechanism are not carrying the chassis weight.
4. Find the underside terminals of C-95. The can is mounted so that its common/negative connection is insulated from the chassis. Do not assume the can mounting hardware is ground.
5. With the ohmmeter, first identify chassis ground on a known chassis lug. Then trace the isolated C-95 common/negative terminal into the back-bias wiring by following the actual solder lugs and wires. Do not try to measure “through” C-96; a good 0.1 µF capacitor reads open on an ohmmeter and a leaky one reads a misleading resistance.
6. Look for a 0.1 µF tubular capacitor with one end on that negative-bias node and the other end on a chassis-ground lug. That is the electrical signature of C-96.
7. Before removing anything, photograph both ends and tag them “C-96 bias node” and “C-96 chassis”. If the capacitor has already been replaced by a modern film part, identify it by these connections rather than by appearance.

Do not use colour or body style as the primary identifier. Collins production changes and seventy years of field repairs mean you may see a waxed paper tube, a moulded tubular, an orange-drop, a blue film capacitor, or something else entirely in this position. Two correct nodes are stronger evidence than a matching-looking component.

5. Reworking C-96

Replacement part

Use a quality 0.1 µF film capacitor, non-polarised, with a DC rating comfortably above the original requirement. A 400 V or 630 V metallised polyester or polypropylene part is the conservative modern choice if it fits mechanically; the working voltage across C-96 is only the bias rail, so the higher rating is purely margin. There is no polarity to observe.

Parameter	Original / documented	Rework recommendation
Capacitance	0.1 µF (75A-3 schematic)	0.1 µF
Type	Paper, non-polarised	Polyester or polypropylene film
Voltage	150 WV reported for the 75A-2 equivalent; verify against your parts list	400–630 VDC for margin, subject to fit
Tolerance	Not critical for bypass duty	±5% or ±10% is fine
Lead dress	Original point-to-point wiring	Keep leads short; follow the original route and sleeving

Rework procedure

1. Document. Close-up photograph of both connections before applying any heat. Note the routing and any sleeving.
2. Remove. If the terminal lug is crowded, clip the old capacitor lead close to the body and remove the remaining stub from the lug with minimum disturbance. Avoid unwrapping several original component leads from the same terminal unless you must.
3. Clean. Clean the lug only enough to obtain a sound joint. Do not overheat phenolic terminal strips or tube-socket lugs.
4. Form. Shape the new film capacitor leads so there is no mechanical tension on the part. Add sleeving where a lead passes near another terminal or a chassis edge.
5. Install. Solder the chassis end to the original chassis-ground point and the other end to the exact original bias node you documented.
6. Inspect. Under magnification, look for solder bridges, loose strands, resistors that have moved, or a lead touching the chassis unintentionally.

6. Post-Rework Checks

Power off

• Confirm the new C-96 is 0.1 µF and is installed as a non-polarised part.
• Confirm one lead is on chassis ground and the other is on the same bias node you documented before removal.
• Confirm the isolated C-95 common/negative terminal has not been accidentally grounded to the chassis. With the ohmmeter, that terminal should read the bias-network resistance to chassis, not zero.
• Check that no adjacent component lead or solder splash was disturbed.

First power-up

• Use a current-limited first start if you have one. Watch for abnormal lamp brightness, smoke, odour, transformer heating or arcing.
• After the initial fault check, power the receiver normally and measure the bias rail relative to chassis. Compare it with the value on your schematic; the actual reading varies with line voltage and receiver condition, so a modest deviation from the printed figure is normal.
• Confirm normal audio, AVC action and RF-gain behaviour. A C-96 replacement on its own should not require realignment.
• If the bias rail is badly wrong, shut down and re-check C-95 isolation, C-94 polarity and condition, the rectifier and power-supply wiring, and your C-96 connection points before troubleshooting anywhere else.

If you still cannot identify the part

Observation	Next action
No original-looking 0.1 µF paper capacitor in the area	Assume prior service. Trace the two nodes and look for a modern film capacitor instead.
Several 0.1 µF capacitors look identical	Ignore appearance. Only C-96 connects between the back-bias node and chassis in this part of the circuit.
C-95 common terminal appears grounded	Stop. Verify you have the correct terminal, then check whether a prior repair (or a replacement can fitted with its case grounded) has defeated the required insulation. The receiver cannot develop bias with that terminal at chassis.
Bias is far from the schematic value after rework	Power down and troubleshoot the whole back-bias supply: C-95 isolation, C-94, the bias load resistors, wiring, and the rectifier and power-supply connections.
Receiver works but hum or noise remains	C-96 is rarely the only aged capacitor. Plan a full recap of the paper and electrolytic parts, changing one functional group at a time so any new fault is easy to attribute.

7. Sources and Confidence

Primary-source facts in this note come from the Collins 75A-3 instruction book and its schematic, with the circuit description reproduced by the Collins Collectors Association. The 75A-2 parts list is used only as corroboration for the C-96 function and value, because the two receivers share the same bias and power-supply arrangement. The modern replacement voltage and film-type recommendations are service practice rather than Collins specification.

A note on other people’s photographs: underside photographs of 75A-3 chassis from auction listings and forum posts are useful for orientation and wire-routing comparison, but a single photograph cannot prove that a particular tubular part is C-96. Most surviving receivers have been serviced, the factory manual omits the callout, and the component may already have been replaced. That is why this note uses a stronger identification rule: C-96 is the 0.1 µF capacitor between the documented negative bias node and chassis.

Footnotes and references

1. Collins Radio Company, 75A-3 Receiver Instruction Book, February 1953, Section V component-location views and Figure 5-6 main schematic. Scan hosted by the Collins Collectors Association: 75A-3 (02-53).pdf. See also the CCA manuals index for alternate scans.
2. Collins Collectors Association, 75A-3 historical overview and circuit description; RigReference notes the 75A-3 as identical to the 75A-2 apart from mechanical-filter provisions and the 75A-2A conversion badging; Nationwide Radio (KE9PQ) capacitor-kit listing notes a single component difference between the two receivers.
3. Collins Radio Company, 75A-2 Receiver Instruction Book, 1952, Section VI parts list. Scan hosted by the CCA: 75A-2-1952.pdf. The “back bias by-pass” description, 150 WV rating and Collins part number sometimes quoted for C-96 come from that list; the scan is not text-searchable, so check the figures against your own copy.
4. Radiomuseum.org, Collins 75A-3 model page: V-1 changed from 6CB6 to 6DC6 in sets above serial number 1573.
5. Collins 75A-3 circuit description as reproduced by the CCA (link in note 2): AVC amplifier cut-off bias of approximately 11 V from bias load resistor R-36; AVC amplifier plate voltage developed across R-37 and R-38; output-tube grid bias taken from a point on the voltage divider in the negative lead of the power supply; standby insertion of R-76, 1500 Ω, in series with the bias load resistors.
6. Tube complement for the 75A-3 (one 5Y3GT rectifier, one 0A2, three 6AL5, one 6AQ5, seven 6BA6, two 6BA7, one 6CB6, one 12AT7, one 12AX7) per radioing.com and the WA3HSC Collins tube and substitute list in the CCA archive.

---

End of note — 73.

Mike Peace VK6ADA · r-390a.net Administrator · Revision B, 21 September 2026