Restoration and Repair Tips

These are Boat Anchor related restoration and repair tips from various sources.

  • National PW dial removal
  • Heat and Power Transformers
  • Resoring 75A4 Knobs
  • Cleaning a rusty chassis
  • Stress cracking agents
  • Polishing & protecting things/wrinkle painting/etc.
  • Re-Forming Electrolytic Caps
  • Refoming Capacitors

    National PW dial removal

    I have had success removing several stubborn PW dials with the following technique:

    From: "Mark Glusker" (

    Heat and Power Transformers

    There are some things that influence transformer heating, both design parameters and faults, and some trouble-shooting methodologies that are fairly quick and bulletproof.

  • 1. "Dry" electrolytics actually have a wet paste inside that dries out. Most common failure is loss of capacitance, not leakage or shorting. There is far too much attention paid to electrolytics and not enough to everything else.

  • 2. Most common cause of excess current draw from power supplies is faults in the audio power amplifier circuits. Low-value cathode bias resistors and leaky grid coupling caps are the two major offenders. Tests are quite simple:

    A. With the set powered off, check the value of the bias resistor with an ohmmeter. It should be what the resistor says. While most resistors age and/or fail upward in value, I've seen a lot in this location go low. Check the grid leak resistors too, and replace them if they are more than 10% high.

    B. Pull the output tube(s), turn the set on, and check that there is no voltage being developed across the grid leak resistors. By "no voltage" I mean millivolts, maximum. This tests the coupling caps for leakage. Note that this same test can be used to check for leaking bypass caps in screen decoupling circuits where they are used. With the tube out of the circuit, there should be no voltage across the decoupling resistor (typically 2K with a .01 at the screen end).

    C. Plug the output tubes in and turn the set back on. Check that there is still no voltage developed across the grid leaks (voltage here means excess grid current).

    D. Check the voltage across the bias resistor and calculate the idling cathode current. Check with your favorite tube manual to see if the current matches the parameters listed. Most sets use one of the recommended manual operating points; if not, interpolate between points given.

    You can predict the total B+ energy budget for the set by looking up typical values for each of the tubes in the set. This is typically about 10 ma. cathode current for each RF, Mixer, and IF stage, plus any current drawn through a VR tube. Now, measure the B-supply current. Easy way is to put a 10 ohm resistor in the B-return lead to the power transformer and measure the voltage across it. 100 ma. will give you a volt. If your meters are not accurate, use a larger resistor, but make sure you have the actual value of the resistor used, not the nominal color-band value. I check the resistor on a precision bridge and use an HP412A, which will measure millivolts accurately. The measured value and the predicted value should match within a reasonable margin of error, and I have found that predicted is almost higher than actual in a properly-working set. You will, of course, have to add any current drawn by bleeder networks.

    To find leakage, pull out all the tubes except the rectifier, leaving a resistor in place in the B-return. Make sure you allow for any bleeders, and disconnect them if necessary. Check for zero volts across the B-return resistor, and find the leakage by disconnecting things as needed. Don't ignore postage stamp caps---they are notorious for reading properly on low voltage checks and leaking like a sieve with B+ applied.

    Check the value of the first (input) cap in a capacitor input filter. It should be no higher than listed on the P/L. Cap input filters reflect a terrible heating waveshape to power transformers, and the bigger the cap, the worse the effect. Hammond suggests 27% as the derating value for this, although Langford-Smith gives a value on the order of 15%, probably a function of the fact that 8-10 mfd. was used in the 1930's, and 30-40 mfd. in later RC filters. Also note that LC filters do more filtering with less losses than RC filters. Hammond makes suitable chokes, and these are preferable to use when replacing a burned-out series field coil speaker.

    Check that tubes haven't been substituted with others drawing a higher heater current. Notorious example is use of 6L6 (900 ma.) in place of 6V6 (400 ma.).

    From: Henry van Cleef (vancleef@NETCOM.COM)

    Restoring 75A4 Knobs

    About knobs:

    There are generally two kinds of plastic knobs as there are generally two kinds of plastic. Thermoplastic and thermoset. From Books '97:

    "Thermoplastic, a material that repeatedly softens or melts when heated and hardens when cooled. Thermoplastic polymers consist of long polymer chains that are not connected to each other, e.g., not cross-linked. Thermoplastics include polyethylene, polypropylene, polystyrene, polyester, polyvinyl chloride, acrylics, nylons, spandex-type polyurethanes, polyamides, polycarbonates, fluorocarbons, and cellulosics." "Thermoset, a material that can not be softened on heating. In a thermosetting polymer, the long polymer chains are joined to each other, e.g., cross-linked, during fabrication through the use of chemicals, heat, or radiation; this process is called curing or vulcanization. Important thermosets include alkyds, phenolics, ureas, melamines, epoxies, polyesters, silicones, rubbers, and polyurethanes."

    The Collins knobs we are talking about are *phenolic*, pigmented with carbon black, otherwise they would have a translucent brown cast to them, the natural color of phenolic. Allen Bradley carbon composition resistors are molded in phenolic. As the encyclopedia says, heat and pressure is used to mold thermoset plastics. After the molding process is complete phenolic will not melt, it will only burn if it gets hot enough. So what is bakelite? Books sez: "Bakelite (bke-lt) [for its inventor, Leo Baekeland], a synthetic thermosetting phenol-formaldehyde resin with an unusually wide variety of industrial applications ranging from billiard balls to electrical insulation."......And knobs.

    Anyone remember Dakaware? They made all kinds of bakelite stuff, including knobs. As bakelite ages and is exposed to various chemical assaults the softer parts of the compund decompose and the knob's surface becomes dull. Some of the eroded phenolic is bound in place by coats of grime. A good bath in soap and water will flush away this surface amalgam and reveal a badly erroded surface.

    The trick to restoring a dull bakelite knob is not to grind away the surface by extensive polishing, but to fill the emptied pores and then lightly buffing the new surface for optimum shine. Black shoe polish is one way but tends to come back off on the fingers. I get fairly nice results by using a good quality carnuba furniture finishing wax after washing knobs clean with soap and water. I hand polish them with an old T-shirt for shine. The knobs when done don't look like brand new but they have a nearly gloss black satin finish which I like to think honors their age.

    From: "Arden Allen" (

    Cleaning a rusty chassis

    Seems like a radio with a rusty chassis top is always a dilemma. Small areas of rust can be cleaned up by mechanical means and then apply Rustoleum. You can mix up a matching color of gray so the spots are not prominent. For inaccessible areas flowing rust inhibitor in is simple and stabilizes the situation.

    For a chassis that is well rusted over I would seriously consider stripping the chassis bare after making a complete layout drawing and then getting it sand blasted and zinc plated.

    Transformers with brittle wire insulation need new leads. I'm not nuts about interior originality so PVC insulated wire is fine. I am a fanatic about color code though.

    From: "Arden Allen" (

    Stress cracking agents

    Stress corrosion cracking can occur with both metals and plastics. Metallurgists have quite a bit of information about this with various alloys but there are no good, all-inclusive theories about why it occurs with thermoplastic materials. It has been suggested that many organic liquids can permeate the crystalline regions of plastics and allow external stresses to do more damage than they normally would.

    I am sure most of you know how tough the PET plastic is that is used to make soft drink bottles. [Eastman is a major producer of PET plastics.] Yet when the right organic chemical is painted on the outside of a filled bottle and quickly wiped off, and the bottle is allowed to sit undisturbed - in a few weeks it will literally be covered with cracks and fall apart in a brittle fashion. Many organic solvents are capable of doing this, so it becomes quite important which solvents are used in inks for external printing of these bottles. Harry mentioned Armor-All, which is designed to be used with plasticized polyvinyl chloride plastics, caused problems with ABS plastics. Armor-All is basically silicone oil dissolved in a mixture of solvents. One of these obviously was not good for ABS.

    The point of all this is that any organic solvent containing materials should be suspect around plastics until you have thoroughly tested them. But GoJo is soap you say. Not really. It is an emulsion of a detergent in water and organic solvents. Ever wonder why it cuts grease so well? Just smell it.

    Fortunately Boatanchor radios did not generally use the large amounts of plastics seen in today's radios. But you still have to worry about paints and lacquers being compatible with cleaning solvents. The best thing is to test everything first in an inconspicuous place on the radio. On things like clear dials or meter faces, stick with something very mild to clean them. If you have to use something like GoJo or Goof-Off or such, use it carefully and rinse it off as quickly as possible.

    From: DCrespy (

    Polishing & protecting things/wrinkle painting/etc.

    Brasso is okay--but I continue to swear by "Flitz" metal & Bakelite polish. It does a first-rate job. On really big metal things, or where it's especially nasty, I have been known to start with cotton wadding polish (cannot bring the name to mind--it comes in a round tin), but it's a little too aggressive for anything but metal. Fliz will polish up any kind of brass, Bakelite, Chrome, (etc.) and even get aluminium semi-presentable. It leaves a very slight waxy residue.

    On surface coatings, for smooth surfaces, plain old-fashioned paste wax (without any fany modern chemicals) is hard to beat; J-Wax was the stuff of choice for broadcast transmitters but any *plain* paste wax will do. On things I make and paint wrinkle black, I usually apply a good clear overspray. This fills in the wrinkle a little but retains the look, and it's less eager to catch dust. Probably *not* the thing to do on an HRO or Harvey-Wells Bandmaster, however.

    Repainting wrinkle finishes: many times, you can carefully apply new paint *without* stripping the old; just get it *really* clean first. Bake and it'll conform to the old wrinkles. You need to get a smooth, even cover, as the shade will not match exactly in nearly all cases.

    Baking wrinkle paint: heat lamp in a cardboard box, with some coat-hanger wire grid shelves. Works like a champ, does not smell up the cooking oven, and if you choose your spot, it's safe. But you *do* have to keep an eye on it--I have a cheap HVAC dial-on-a-probe thermometer stuck in mine; it runs about 120 degreees F. Also works on plain paint--did an Altoids-tin sized enclosure in gloss black today and it was dry to the touch in an hour. (No tubes in the device that lives in it or I'd talk about it). You need to bide a wee between painting and putting the painted stuff in the oven, for the most volatile parts of the paint to evaporate; I usually wait 10-15 minutes.

    From: "Roberta J. Barmore" (

    Re-Forming Electrolytic Caps

    I found this in the manual for a Sprague model TO-5 `Tel-Ohmike' capacitor analyzer (circa 1958).

    This is for leakage testing of electrolytics: "4.3 It will be noted that the voltage reading will tend to increase after a short time as the leakage current begins to decrease to a stable value. The [applied] voltage control should be retarded accordingly to prevent more than rated voltage from being applied to the capacitor. The measurement of leakage current should be made only after a stable value [of leakage current] is reached.

    Capacitors which have been have been out of use for periods of a year or more make take as long as 30 minutes to reach a stable value of leakage current. Such capacitors usually have a high current initially and the [applied] voltage control should be retarded so that the leakage current is less than 10 ma. in order to prevent overheating of the capacitor internally. The voltage should be adjusted upwards until rated voltage is reached as the leakage current decreases. When rated voltage is finally reached, proceed as detailed above. If there is appreciable fluctuation in the leakagae current indication, the capacitor is probably intermittent and should be discarded.

    4.4 Test Limits - New radio-TV type electrolytics should have a maximum leakage current as shown in following table: .... [long table here in text, use formula below instead]

    Maximum leakage currents ... may be derived from the following formula:

    I = kC + 0.5

    where I is the leakage current in mA, C is capacitance in uF, and k is a constant as follows:
    WVDC K
    3 to 100 .01
    101 to 250 .02
    251 to 350 0.25
    351 to 500 .04

    Readings should be taken 5 minutes after capacitors are placed on rated DC Working Voltage. These limits may be used as a guide in judging whether capacitors should be replaced, making due allowance for the usual increase in leakage with age and with any high ambient temperature at which measurements are made. Capacitors with a leakage current of more than 15 mA should almost always be discarded."

    On power factor (assuming you can measure it):

    "2.4 Power Factor 2.4.1 DC Dry Electrolytics - The 60 cycle power factor of new capacitors will usually fall below the maximum value give below. Capacitors rated at 150 volts or higher shoud usually be replaced if the measured value is twice that given. Low voltage sections of multi-section capacitors will generally have power factors higher than that listed, sometimes by 50%.
    WVDC Max. New P-F
    475 - 300 15
    250 18
    150 20
    50 25
    25 30
    15 50
    12 55
    6 60

    From: JOHN SEHRING (

    Refoming Capacitors

    For years I have used an old Aerovox (1935 vintage) capacitance bridge for re-forming. The bridge is able to measure capacitance under rated voltage and to that end has a built in current limited variable power supply with a meter that can measure terminal voltage as well as current. I typically start out with applying about 6 ma in the early stages of re-forming and observe whether or not the current decreases. If it does (hopefully) then after the terminal voltage reaches about 100 or so volts I decrease the re-forming current to 3 ma to prevent capacitor heating. Then I just leave the set up alone, come back to it every so often and adjust the voltage to raise the current to 3 ma again until the DC operating voltage of the capacitor is reached at which point the leakage current is typically less than 1 ma. It is not unusual, and even expected, that the higher capacitor values, like 100 mfd at 600 VDC will still have a residual leakage of several milliamps. Capacitors in the range of 20 mfd at 600 VDC should end up with about 1/2 ma or so. All this, by the way, being done with the capacitor totally out of circuit.

    From: "Mike B. Feher" (

    Back to the Radio Bay