State of the R-390A Parts Market 2026
Availability, Critical Scarcities, and Sourcing Strategies for Today’s Restorer
A practical current-conditions survey of the R-390A parts landscape in 2026: what remains straightforward to source, what is genuinely scarce, and what is now effectively impossible to replace. Special focus on the two parts most likely to prevent a complete restoration — the Collins mechanical filter (FL-26/FL-44 series) and the IF transformer assemblies — with honest assessments of substitute options, the 3TF7 ballast tube situation, the drive coupling and slug rack parts picture, and sourcing strategies ranked by reliability. Based on 27 years of r-390a.net community procurement experience.
The R-390A parts market in 2026 is a tale of two inventories. Everything that can be substituted with a modern commercial component — capacitors, resistors, standard tubes — is easier and cheaper to source than it has ever been. Everything that is unique to the R-390A or to its era of military procurement — the Collins mechanical filters, the IF transformer assemblies, the 3TF7 ballast lamp, the phenolic drive couplings, the slug rack bevel gears — is harder to find and more expensive than at any point in the 27 years this community has been active.
The transition point happened gradually but unmistakably: the generation of R-390A units that entered the surplus market in the 1980s and 1990s was large enough to supply the community’s needs for unique parts through the 2000s and into the 2010s. By the mid-2010s that surplus had substantially absorbed. The units still coming to market are increasingly units that were already non-functional, missing key parts, or previously cannibalised. The easy parts were used up years ago; what remains available is what nobody needed until recently.
This document gives the current-conditions picture as of early 2026. Prices are directional rather than specific — the market moves fast enough that specific figures are outdated within months — but the availability tiers and sourcing strategies reflect genuine community experience over the past several years. Readers who disagree with a specific assessment are encouraged to share current data via r-390a.net.
This parts market post is a companion to the R-390A Failure Prevention Kit at vk6ada.com.au (which specifies what to order) and the R-390A Failure Frequency Analysis at r-390a.net (which tells you what to order first). The community parts exchange, current eBay listings, and specific vendor recommendations are maintained as a living resource at r-390a.net; the present document provides the strategic context for those searches.
Section 1 — Parts Availability Overview: The Four-Tier System
R-390A parts fall into four availability tiers that determine how a 2026 restorer should approach each category. The tiers are defined not just by current price but by the expected trajectory over the next five years: a part that is currently available but trending scarce is categorised differently from one that is plentiful and stable.
Tier |
Description |
Examples |
Sourcing Effort |
|---|---|---|---|
| AVAILABLE | Modern commercial equivalents exist and are in current production. Availability is stable or improving. Price is low relative to historical. | All capacitors (electrolytic, film, ceramic), metal film resistors, wirewound resistors, standard IF tubes (6BA6, 5749), common audio/driver tubes, hardware | Online order, one supplier. No waiting. |
| WATCH | NOS stock exists but is being consumed. Price trending up. No new production. Substitutes available but may compromise performance. | Front-end RF tubes (6DC6, 6CB6), 5Y3 and 5R4 rectifiers, specific carbon comp resistor values for purists, slug rack springs, worm gear assemblies | Multiple suppliers, eBay, community exchange. Allow days to weeks. |
| SCARCE | NOS stock nearly exhausted or only available from other complete receivers. Community supply is the primary source. Substitutes may exist but require engineering work to fit. | 3TF7 ballast lamp tube, IF transformers (as complete replacements), individual slug rack coil assemblies, drive coupling phenolic | Community exchange, dedicated search, patient waiting, alternative solutions required. |
| CRITICAL | No new production possible. NOS supply functionally exhausted for practical purposes. Community cannibalisation is the dominant supply. A failed example may be irreplaceable without accepting a substitute that changes receiver characteristics. | Collins mechanical filters (FL-26, FL-44, FL-17) in working condition; decade counter phenolic gears in good mesh; complete working slug rack assemblies; main drive shaft and anti-backlash nut assemblies | Long-term watchlist, community relationships, paying premium prices, or accepting substitutes. |
Section 2 — What Remains Readily Available
The news is not all bad. A large fraction of the parts required for a thorough R-390A restoration are modern commercial components that are cheaper and easier to source in 2026 than at any previous time. A restorer who uses the Failure Prevention Kit parts list as a shopping guide will find the following available from standard electronics distributors (Mouser, DigiKey, Farnell) with next-day shipping:
Capacitors — No Meaningful Availability Concern
Every capacitor in the R-390A can be sourced from current production. Main B+ filter electrolytics: standard 105°C aluminium electrolytics in the correct values are a commodity. Individual screen bypass and audio bypass caps: identical situation. Film capacitors for all paper and wax type replacements: polypropylene film in all required values is in current high-volume production. Silver mica capacitors for the few positions that require them: available from specialist distributors. NP0/C0G ceramics for BFO tank and PTO buffer positions: standard catalogue items. The only capacitor that requires any care is the PTO buffer coupling cap — and this requires only that silver mica rather than film is specified, which is a specification choice, not a scarcity problem.
Resistors — No Meaningful Availability Concern
1% metal film resistors in all values needed for the IF strip and AGC network replacement are standard commodity items. Wirewound and metal oxide resistors for bleeder replacement are catalogue items at any wattage required. The only caveat: a very small number of collectors prefer to replace original carbon composition resistors with new carbon composition types for aesthetic reasons. New carbon composition resistors are still available from specialist suppliers (like Ohmite in the US), but using them in the bleeder or IF bias positions defeats the purpose of restoration as they will drift again under identical operating conditions. This is an aesthetic preference, not a parts availability issue.
Common Tubes — Generally Available, Some Care Required
The tubes that make up the bulk of the R-390A’s complement remain available with varying degrees of effort:
- 6BA6 / 5749 (IF amplifiers, ×9 positions): Both the civilian 6BA6 and the military 5749 remain available. NOS 5749 from JAN (Joint Army-Navy) stock is the preferred choice for IF positions and is still findable at reasonable prices. Standard 6BA6 is a commodity.
- 6DC6 (RF amplifier): Available, WATCH tier trending toward SCARCE. NOS stock is reducing. The 6CB6 is a functional equivalent accepted in this position by most of the community.
- 6AQ5 (audio output): Very common, AVAILABLE. Standard audio output tube used in many consumer applications.
- 12AX7 / 12AU7 (various positions): AVAILABLE. Both types are in current commercial production (Electro-Harmonix, Mullard reissue, others) and are abundant NOS. Never a concern.
- 5Y3 / 5R4 (rectifiers): WATCH. NOS supply is shrinking. The 5R4 is preferred for reliability; the 5Y3 is adequate and more common. Silicon diode rectifier upgrades are well documented in the community and eliminate the dependence entirely.
- 0A2 (voltage regulator): Available in current Sovtek production and from NOS stock. Not a concern.
Section 3 — The Critical Scarcities: What Stops Complete Restorations
The following five items are the genuine restoration blockers in 2026. Each has reached a point where the community supply is the dominant source and the NOS pipeline from military surplus is effectively exhausted. A restorer who needs one of these items should expect a lengthy search and should plan accordingly before beginning a restoration that depends on it.
-
Collins Mechanical Filters — FL-26, FL-44, FL-17 Series at 455 kc
The Collins mechanical filter is the most consequential scarcity in the R-390A parts market. It defines the receiver’s selectivity and is, in an operational sense, what makes the R-390A the instrument it is. The FL-26 (narrow, approximately 4 kc), FL-44 (wide, approximately 8 kc), and FL-17 (used in some specific configurations) are all irreplaceable in the sense that no current production mechanical filter provides identical characteristics. Rockwell-Collins long ago ceased production of the original military-specification types.
Why the situation is now critical: a working FL-26 or FL-44 that was available for under $30 on eBay in 2005 now commands $100–$200 or more for a unit of confirmed performance. The supply is not zero — units from cannibalised receivers continue to appear — but the rate of consumption (every R-390A restoration requires at least one, and many restorations need two or three bandwidth positions tested and replaced) is outpacing the supply from new surplus arrivals. WA3DSP found degraded insertion loss in approximately 70% of units he tested; that figure implies that only 30% of the mechanical filters in the remaining installed base are performing to original specification.
The substitute question: Murata and other manufacturers produce 455 kc mechanical filters in standard bandwidth classes. These can be fitted and will provide selectivity, but they are NOT drop-in replacements: the insertion loss, passband shape, and ultimate attenuation characteristics differ from the original Collins specifications. A receiver fitted with a Murata substitute sounds different from a receiver with an original FL-26. The community consensus is that the Murata substitute is acceptable for operation but should be documented as non-original. Any unit offered for sale as a “complete and working” R-390A should have its mechanical filter bandwidth positions individually tested and the results disclosed. Do not assume a working receiver has a working filter — a receiver can produce audio with a severely degraded filter.
Current sourcing: eBay (watch saved searches for FL-26, FL-44, R-390A filter, Collins 455 filter), r-390a.net community exchange, Antique Radio Forums parts wanted threads, other R-390A collectors upgrading units. The best source in 2026 is another collector who has a partially functional unit and is willing to sell it as a parts donor.
-
IF Transformer Assemblies (Y-slot shield can type)
The 26 IF transformer assemblies in the R-390A’s main IF strip are not, strictly speaking, scarce — the overwhelming majority of them are still functional after 60–70 years. The issue is specific: when an IF transformer fails with an open or shorted winding (as opposed to the common lid desoldering that is field-repairable), it cannot be replaced from any commercial source. There are no current production equivalents. Replacement units must come from another R-390A chassis, and complete functional second IF strips for donor purposes are themselves increasingly rare.
The practical situation: for most R-390A restorations, IF transformer availability is not a blocking problem. The lid desoldering failure (the #2 most common R-390A fault) is always field-repairable — re-solder the lid, restore the shield integrity. A genuinely failed winding is a different matter. Community data suggests that winding failure (as opposed to lid desoldering) represents fewer than 5% of IF transformer failures, but when it occurs in a specific position in the IF strip, a working unit may be the only way forward.
If you have a failed winding: contact the r-390a.net community parts exchange with the specific transformer designator. Some community members have partial IF strips from R-390A chassis salvage. Rewinding is theoretically possible but requires specialised knowledge and small-coil winding equipment; it is not a practical option for most restorers. An alternative is to accept slightly reduced IF gain if the failed transformer is in a middle position of the strip — the remaining stages provide enough gain for usable performance in many cases, though sensitivity will be reduced.
Prevention is the entire answer here: inspect all 26 transformer lids at the beginning of every restoration. Re-solder any cracked lid immediately. The field-repairable failure mode addressed promptly prevents the escalation to an irreplaceable-failure situation.
-
3TF7 Ballast Lamp Tube
The 3TF7 is a tungsten-filament incandescent lamp in a glass envelope that serves as a current-regulating element in the R-390A filament supply circuit. It is not a conventional vacuum tube — it is a thermistor-type lamp whose positive temperature coefficient of resistance stabilises filament voltage throughout the receiver’s tube complement. There is no current production equivalent and there has not been for decades. NOS stock continues to exist in the surplus market and in the hands of collectors, but the supply is finite and diminishing.
Current sourcing: eBay (search “3TF7” and “R-390 ballast tube”). Prices have risen substantially — expect $15–$40 for a tested unit depending on source. The community has documented specific equivalent lamp types that can be adapted with minimal electrical modification; Nick England K4NYW’s work on substitutes is the primary community reference. Some aviation and automotive incandescent lamps of the correct wattage and current characteristic can serve as substitutes; the exact type depends on the specific measurement of the installed 3TF7’s operating resistance, which should be measured before assuming any substitute is suitable.
Solid-state substitute: multiple community members have designed solid-state constant-current circuit replacements for the 3TF7 using a transistor-based regulator fitted into the original tube socket. These are genuinely functional and eliminate the 3TF7 scarcity concern entirely for those willing to make a non-original modification. The r-390a.net technical archive contains at least two published designs. A solid-state substitute does not affect the receiver’s operating characteristics in any measurable way; the filament voltage stabilisation is equivalent or superior to the original 3TF7.
-
Oscillator Deck Drive Coupling (Phenolic)
The coupling between the main tuning shaft and the slug rack oscillator deck drive mechanism is a phenolic or acetal gear/coupling element that, over decades of use, develops slip at the contact interface. When the coupling fails completely, the tuning knob rotates without moving the slug rack — the #1 most common R-390A failure by community report frequency. In less severe cases it produces a lag between knob rotation and frequency change.
Why this is a sourcing problem: there is no commercial source for these couplings. The original military procurement parts supply ceased with the end of production in 1964. Couplings from other R-390A chassis are the primary supply — but a chassis being cannibalised for its coupling probably has a failed coupling, making this circular. The community solution, developed by several machinist members over the past fifteen years, is to have replacement couplings machined from modern Delrin (acetal homopolymer) or similar engineering plastic. Delrin is dimensionally stable and has superior wear resistance to the original phenolic, making it a genuine improvement over the original.
Current sourcing: r-390a.net community exchange (machined Delrin reproductions are available from time to time from community members who have had batches made). If you have access to a machinist or a CNC service, the dimensions can be extracted from an original coupling (or from photographs shared in the r-390a.net forum) to have your own made. The dimensions are not proprietary and several community members have shared them.
-
Complete Working Slug Rack Assemblies and Individual Coil Assemblies
The R-390A’s 30-coil slug rack is the heart of the instrument’s frequency-defining mechanism. Each coil position contains a ferrite slug on a lead screw, threaded into a coil form with a specific inductance winding. When a slug fractures (from over-torque during alignment or thermal stress) and the coil form is damaged, there is no individual coil assembly replacement available from any commercial source. Replacement requires a complete slug rack from another R-390A chassis.
The 2026 situation: complete working slug racks — all 30 coils intact, all slugs uncracked, bevel gears in reasonable mesh — are the most expensive single mechanical item in the R-390A parts market. A complete working slug rack assembly in good condition can command prices in the hundreds of dollars when it appears. Partial slug racks (one or several cracked slugs) are more common but require assessment of which band segments are affected before purchase.
The repair options: a fractured ferrite slug can sometimes be repaired with cyanoacrylate adhesive applied with extreme precision to reattach the fractured piece, but the repaired slug will not survive re-adjustment and any further force will re-fracture it. This is a temporary measure to restore operation on the affected band rather than a permanent repair. Replacement ferrite slugs of the correct ferrite type, thread pitch, and diameter have been sourced by community members from specialty ferrite suppliers; the specific ferrite material and physical dimensions required are documented in the r-390a.net technical archive. This is the most viable long-term repair path for isolated slug fractures.
Section 4 — The Cannibalisation Question
The most difficult judgement in the 2026 R-390A parts market is whether to cannibalise one chassis to restore another. Twenty years ago, with a large and inexpensive surplus supply, this was an easy call: a $40 non-working R-390A from a hamfest was a parts donor, plain and simple. In 2026, with every R-390A chassis representing a receiver that some future restorer might want to complete, the ethics and economics have shifted.
What to Assess Before Deciding to Part Out a Chassis
- Mechanical filter function: test all bandwidth positions before deciding. A chassis with three working mechanical filter positions is worth more as a parts donor than as a shelf queen with cosmetic damage.
- Slug rack completeness: count the intact slugs. A chassis with all 30 slugs intact is potentially restorable regardless of other faults. A chassis missing the slug rack or with more than 4–5 fractured slugs across important band positions may be a better donor.
- IF transformer condition: inspect all 26 lids. Lid desoldering is universal and repairable; open windings are rare and the key differentiator for a parts decision.
- 3TF7 function: measure the ballast tube. A functioning 3TF7 has value and should be documented and offered to the community even if the chassis is parted out.
- Contract year and manufacturer: document the nameplate information for every chassis. Early Collins, Motorola, GD, and Stewart Warner units have different characteristic failure distributions (documented in the r-390a.net Failure Frequency Analysis) and the contract manufacturer information is valuable to potential parts buyers.
Section 5 — Sourcing Strategies Ranked by Reliability
r-390a.net
antiqueradios.com/forums/
collinsradio.org (CCA)
Section 6 — Availability Spectrum and Forward Assessment
┌──────────────────────────────────────────────────────────────────────────┐
│ R-390A PARTS AVAILABILITY SPECTRUM — MARCH 2026 │
│ Trajectory arrows: → stable / ↓ declining / ↓↓ declining fast │
└──────────────────────────────────────────────────────────────────────────┘
AVAILABLE (stable, modern commercial equivalents) ──────────────────────────
Electrolytic capacitors → Current production, commodity
Film capacitors (all types) → Current production, commodity
Silver mica capacitors → Specialist current production
NP0/C0G ceramics → Current production, commodity
Metal film resistors (1%) → Current production, commodity
Wirewound resistors (all W) → Current production, commodity
6BA6 / 5749 tubes (IF) → NOS abundant, military stock ample
12AX7 / 12AU7 / 12AT7 → New production and NOS, no concern
6AQ5 audio output → Common, no concern
0A2 voltage regulator → Sovtek current production
Standard hardware / screws → Commercial equivalent available
WATCH (NOS declining, no new production) ───────────────────────────────────
6DC6 / 6CB6 (RF amplifier) ↓ NOS shrinking; 6CB6 substitute accepted
5Y3 / 5R4 rectifiers ↓ NOS reducing; Si diode upgrade available
6AL5 detector tubes → Still findable, WATCH in 2–3 years
Audio output transformer ↓ Reproduction available but not original
SCARCE (community supply dominant, NOS near-exhausted) ─────────────────────
3TF7 ballast lamp ↓↓ Price rising; solid-state sub available
IF transformer (winding failure) ↓ Mostly donor chassis; prevent, not replace
Drive coupling phenolic ↓↓ No commercial source; Delrin repro exists
Complete worm/bevel gear sets ↓ Limited community supply only
CRITICAL (no new production, functionally exhausted NOS) ───────────────────
Collins FL-26 narrow filter ↓↓ The #1 restoration blocker 2026
Collins FL-44 wide filter ↓↓ Marginally more available than FL-26
Collins FL-17 (some configs) ↓↓ Very limited
Complete working slug rack ↓↓ $200+ when found; key asset
Decade counter phenolic gears ↓↓ No source; machining required
─────────────────────────────────────────────────────────────────────────
SUBSTITUTE OPTIONS FOR CRITICAL ITEMS
─────────────────────────────────────────────────────────────────────────
Collins mechanical filter: Murata 455 kc filter (bandwidth differs)
Community assessment: operationally acceptable,
not specification-equivalent, must be disclosed
3TF7 ballast tube: Solid-state constant-current sub (multiple
published designs at r-390a.net); equivalent
function, non-original, reversible
Drive coupling: Machined Delrin replacement (superior to
original phenolic); community designs available
Slug (individual fracture): Matched ferrite slug from specialty supplier;
requires correct material specification
─────────────────────────────────────────────────────────────────────────
FORWARD ASSESSMENT: 2026–2030
─────────────────────────────────────────────────────────────────────────
Most likely to worsen: Collins mechanical filter supply and pricing.
Every restoration consumes at least one; supply
from new arrivals is declining. Plan ahead.
Most likely to improve: 3TF7 situation (solid-state sub increasingly
accepted as standard). Drive coupling (community
machining of Delrin repros is established).
Stable: Tubes and electronic components throughout.
The electronic side of an R-390A restoration
will get easier, not harder, through 2030.
Wildcard: 3D-printed phenolic reproductions for mechanical
parts. Quality of SLA resin printing has reached
the point where slug rack coil forms and drive
couplings are beginning to be reproduced at the
collector level. Not yet community-standard but
worth watching.
R-390A parts availability spectrum as of March 2026. Trajectory arrows reflect the 2024–2026 trend direction. Forward assessment covers the 2026–2030 period based on current community supply rates and restoration demand. Individual market conditions vary; prices cited in community discussions are the most current source.
Sources and Community References
- R-390 Reflector archive (1999–2015), preserved at r-390a.net. The historical record of community parts sourcing discussions, pricing benchmarks from the 2000s through 2015, and the first documented discussions of mechanical filter scarcity as a restoration concern. Comparison of current market conditions against this archive provides the trend direction data cited in the forward assessment.
- WA3DSP (Peter Wittenberg, Silent Key 2022), R-390A Rebuild Series, r-390a.net. The rebuild series contains documented parts sourcing experiences from approximately 50 unit restorations performed between 2000 and 2020. WA3DSP’s finding that approximately 70% of tested mechanical filters showed elevated insertion loss is the source for the filter scarcity severity assessment in Section 3.
- Nick England K4NYW, 3TF7 substitute research, R-390 Reflector archive, r-390a.net. K4NYW’s documentation of the 3TF7 substitute lamp types and the in-circuit measurement procedure for identifying an acceptable equivalent is the primary community reference for the ballast tube situation.
- eBay completed listings, search terms “FL-26 filter”, “FL-44 filter”, “3TF7”, “R-390A mechanical filter”, “R-390 slug rack”: 2020–2026 sold prices. Prices cited in this document are directional ranges derived from completed auction data across this period; individual transaction prices vary and may differ from these ranges at time of reading. eBay completed listings are the most current market reference and should be consulted directly for current pricing.
- Mike Peace VK6ADA, R-390A/URR Failure Frequency Analysis — Ten Most Common Failures Ranked by Community Report Data, r-390a.net (March 2026). The failure frequency ranking provides the operational context for which parts matters most: the #1 ranked failure (oscillator deck drive coupling and slug rack) directly corresponds to the SCARCE/CRITICAL tier items in this document. The failure frequency and parts availability tiers are consistent across both documents.
- Collins Collectors Association technical archive, collinsradio.org. Source for mechanical filter type specifications (FL-26, FL-44, FL-17 bandwidth and insertion loss specifications) used to assess the acceptability of substitute filters. The CCA archive is the primary reference for original Collins design specifications against which any substitute is compared.
- r-390a.net community parts exchange and forum, ongoing (2024–2026). The most current source for actual availability and pricing of R-390A-specific parts. The assessments in this document reflect community exchange activity across the 2024–2026 period; the forum itself is the living reference for current conditions beyond the publication date of this post.