ad: Preppcomm-1

Drake- Loser Transistors

Discussion in '"Boat Anchor" & Classic Equipment' started by WA1GXC, Sep 25, 2021.

ad: L-HROutlet
ad: l-rl
ad: L-MFJ
ad: Left-2
ad: Left-3
ad: abrind-2
ad: L-Geochron
  1. WA1GXC

    WA1GXC Ham Member QRZ Page

    I've mentioned many times in passing here about random transistor failures in my Drake gear as each unit
    approaches approx 40-43 years since manufacture.
    Among my many "4-Line" transmitters and receivers, semiconductors have just given up and croaked,
    immediately after a short period of acting-up. I was actually listening to my R-4C once as the VFO (PTO) oscillator (2N5950 FET) signal went "woooooo-uup" and failed . No component abuse,
    just old age .The quirky failures are too numerous to mention--Audio, RF, buffers...They're all logged in my notes but too diverse to discern a pattern.

    New failure mode-- I have a 1974-build T-4XC, most recently acquired unit 2-1/2 years ago, which
    I overhauled upon arrival. Never touched the VFO, however. I normally replace both VFO transistors when I'm inside for other repairs, for reliability; didn't re-do this one. Wasn't in the mood.
    The VFO has consistently shown high drift (300-400Hz) during a 5-10min transmit cycle--despite temp-
    compensating caps in the circuit, unstable vs. temp. This is with a muffin fan cooling the P.A. cage and keeping the unit cool overall. My guess is that the relay-control tube nearby
    (6EV7 or 6CG7 later-production) a very hot-running tube, was throwing enough heat to make the transistor junction(s) unstable while conducting during transmit cycle.
    Just replaced PTO buffer & oscillator transistors, and the unit runs absolutely driftless, just like all Drake "4s" should do.

    So tip for all: When you have something screwy going on inside your Drakes, from T-4X to TR-4CW, R-4A
    to R-4C, don't overlook transistors--or diodes. I've had numerous signal and switching 1N4148s suddenly
    work poorly in both TXs & RXs. We were told transistors last forever. They don't. 1960s and 1970s semiconductors were not fabricated with the same
    degree of precision, sealing, and impervious epoxy materials as todays extraordinarily economical mass-produced devices.

    Good luck and enjoy your radios.

    "Rich hams buy Collins. Smart hams buy Drakes. Smart & thrifty hams buy busted Drakes."

    Last edited: Sep 25, 2021
    G0CIQ, K4NWD, KM1H and 3 others like this.
  2. N8YX

    N8YX Ham Member QRZ Page

    Also seen with Cubic gear of the period.

    I'd be looking at any electrolytics or tantalums around said transistors as suspect too.
    KM1H and N4FZ like this.
  3. KI4ZNV

    KI4ZNV Ham Member QRZ Page

    What are the transistors?

    I have found that the bullet-shaped epoxy package transistors from Fairchild are very very prone to failure after fifty years due to internal corrosion.
    WA1GXC likes this.
  4. W4OP

    W4OP Ham Member QRZ Page

    So many Drake 4 lines- presently a TR-4CW/RIT and zero solid state device failure. Just lucky I guess. But I also run all of them on 108VAC.

    Dale W4OP
  5. WA1GXC

    WA1GXC Ham Member QRZ Page

    I don't think the issue is endemic to Fairchild only--but you nailed the prevalent fabrication issue.

    I knew I had a systemic failure trend, but wasn't knowledgeable about the specific manufacturing issues. It appeared that transistors utilized in all kind of circuit applications were starting to fail at approx 40 years after manufacture,
    I'd put the time frame consistently 40-45 years. My oldest unit utilizing transistors is a ca. 1967 T-4X.

    I recounted my observations to a fellow ham, a retired professor of electrical engineering at Northeastern University School of Engr'ing, who is still active overseeing several contracts NU has with the US Government.
    He's not a cloistered academic, he's done a lot of stuff in the real world. He told me specifically the issue, precisely as you described, was that of corrosion. He said the epoxy materials utilized in those days were actually
    not impervious, at a molecular level, to moisture; the water molecules eventually infiltrate the outer barrier and then the device fails internally from corrosion.

    The brands I have commonly pulled are lots of Motorolas, lots of Fairchilds, some RCAs and others. The configurations run the gamut from modern-style TO-92 'bullet' integral epoxy units to some older 'can' types,
    probably TO-18 sealed with epoxy at the bottom.

    Drake often made all kinds of component substitutions within production runs, many of which don't appear on schematics or on electron-device inventory list in the manuals. The early T-4X, first in the "4-Line",
    has 2 x 2N3858 in the VFO, 1 as oscillator and 1 as buffer. The R-4, its contemporaneous mate, had a vacuum-tube PTO circuit. Drake then wised-up and with the R-4A and T-4XB, went to a FET as the oscillator
    transistor, although the earliest R-4A schematic and book still shows bipolar-junction-transistor (BJT) 2N3858. My units, T-4XB and R-4A and later, all had FETs as oscillators.
    Crazy mix on schematics--some type-numbers which back then were proprietary to specific manufacturer, not even industry-accepted EIA/JEDEC designators---Texas Instr. TIS34, Motorola MPF-102 (which back then
    was not a generic JEDEC number), proprietary RCA 40310, big honkin' TO-66 audio output. 2N3394 NPN used for all kinds of purposes, some later substitutions with more-specialized and tailored applications.
    Drake settled on 2N5950 as its go-to FET for most circuits.
    Early TXs and RXs also used lots of 2N3858s, NPN BJTs.

    I have a cache of replacement transistors I started to acquire about 2000 when I realized these things were starting to die regularly; almost all were good substitutions, the original types obsolescent or already
    out of production. Now my N-O-S replacements are obsolete.

    Always hoard parts, guys and girls--You'll never be sorry.

    G0CIQ likes this.
  6. KK7BMS

    KK7BMS Ham Member QRZ Page

    Japanese transistors of the day had their problems too. Notably Hitachi. The 2SC454/458 was famous for oxidation on the leads. As it worked its way into the case the transistor would get really noisy. Replaced lots of them
    N8YX and WA1GXC like this.

Share This Page