About Intermittent Radios.

About Intermittent Radios ~  By Bill Jones     

An intermittent radio is difficult, and, the question is where is the bad part; often a real mystery.  As in most mysteries there is a victim.  If you have the intermittent radio, then you are the victim.  Having been in the service business for some years, I do not like to hear that some device is working only most of the time.  When it is working, then there is nothing wrong with it. You can’t fix it without catching it in the act of not working.  Yes, you can replace all the parts that you think might cause the trouble, but that approach can easily end up with your having an unhappy customer, i.e., you really didn’t fix it.  It happens all the time. However, that’s not the problem here.  We have an intermittent radio and how do we really fix it?   Watch eBay for a complete chassis? If you find one, it may also be intermittent. It can be very time consuming to find a solution to an intermittent radio.

There are many ways to approach these radios and my favorite is with a signal tracer as discussed below, but few are around and they are rarely used.  A reasonable approach, I think, is to have at least two instruments.  One, a vacuum tube voltmeter (v.t.v.m. which is a must have), and, a mutineer.  So, with an antique radio that has a problem, what can we do to find the trouble?  I usually think in terms of superhets but the discussion here will also apply to the t.r.f.  It is only a matter of finding the bad stage.  The problem could be in the power supply and that is easy to check with your multimeter.  Often the problem will be a tube. Tube filaments are quite intermittent and this is one cause.  So, it is just a matter of looking at the tube filaments when the set stops working and noting which tube is not lighting up.    Metal tubes complicate things.  If the set goes off for several minutes, then the intermitent tube will cool down. Touch the tubes.   Measuring tube bias voltages in order to check which tube is bad is a good option.   In series string sets, all the tubes go out at once and you have to put a multimeter on the filament of each tube to find which one is bad (bad tube shows a meter reading of line voltage).  My experience has been that these sets are not as likely to have intermittent filaments as the parallel filament sets, but it happens.  

   

A tube may have a good filament but the tube can still be intermittent.  Tapping the tubes will often find a bad tube.  I like to use the handle end of a small screwdriver and tap each tube lightly.  If the set does not show a bad tube then we begin to get desperate and despondent. Tubes made in the late 1940’s are much less likely to have such a problem. The most difficult set is the one that quits and then when you go to check any voltage on the set – it starts working. These are very difficult, and can take hours of bench time.  To service these sets, we bring out the vacuum tube voltmeter.   If your meter does not have a one-megohm resistor in the dc probe itself, put one in – as close to the probe tip as you can get it, in order to reduce stray capacitance.  The series resistor allows the set to operate without detuning a stage. You now measure the bias voltages of each tube and leave the probe on the tube until the set stops operating.  If a tube has cathode bias, then that is an excellent point to probe.  Now, when the set stops, note any change in the cathode, grid, screen or plate voltage. If the screen voltage dropped, it is a most likely a bypass capacitor but keep in mind an intermittent wire wound resistor. It is not impossible to find carbon resistors that are intermittent but very unlikely.   If a grid voltage goes way negative but should be zero, then you know that the grid is not returned to ground for some  reason .The rf/i.f. transformer or the agc line could for example, be the problem. The local oscillator can be the cause of an intermittent. To check this put the v.t.v.m. on the oscillator grid (a negative voltage) and leave it there until the set stops and if the bias voltage goes to zero, then you have found the problem.  It may be the tube or it could be the coil.  Metal flakes in the variable capacitor?   If the oscillator is intermittent I will try another oscillator tube.  Finding the voltage change that caused the radio to quit is the goal. It is relatively easy to find what caused the voltage to change.

    

The signal tracer is of great value in trouble evaluations of any type, especially intermittents. A good tracer does not detune the r.f. or i.f. stages. You merely go from tube to tube with your probe and note whether the signal is present or not when the set stops.   It may still take a lot of time but you don’t have to measure so many voltages in order to find the bad stage.   Another means is to apply a locally generated signal to the grid or plate of each suspect tube.  For example, apply a modulated i.f. signal to the grid of the first i.f. amp.  If the set never stops, then you back up to the r.f.  stage. If it still quits after checking all these stages, you go to the audio stages.  Start, for example, at the volume control. This method is often called signal substitution.  Make sure that your generator is capacity coupled so that you do not disturb the tube bias; injure the generator or the receiver power supply.  One or the other of these methods must be used if and when the tube bias measurements do not turn up the bad stage.   It is unusual and it means, of course, that there is an open or a short (or large change) in some component in the signal line that does not affect the tube bias.  There may, for example, be an intermittent capacitor in an i.f. transformer or in an audio stage.  I once found a loose terminal connection inside an i.f. transformer - it had never been soldered at the factory. When you have found the bad stage, there are only a few components that are suspect and you have solved the problem.

Questions or Comments? Please e-mail me at whnj@att.net  Thanks, Bill.