Typically, this sort of thing is a form of motorboating. There are
a couple of ways it comes about. One is through power supply sagging
as current in some stage starts increasing. The supply sags until
a certain point is reached where the current in the affected stage
decreases, and the cycle continues. It is very low frequency because
the power supply filters are the resonant element.

Another way it can happen is through grid leakage. Many tube stages
of the low power sort have a resistor from the grid to ground that
is supposed to keep the grid at a reasonable negative bias value. If
that resistor is open circuit, or if the tube develops excessive
gas, the grid, the leakage, and the coupling capacitor of the
previous stage can get a relaxation oscillation thing going. A leaky
coupling capacitor, such as C27, or C25 can do the same thing. Look at
pin 7 of V4. Try paralleling a 1M resistor on that stage and see if
it changes the rate of the motorboating. If it doesn't, try the grids
on the 12AX7... and, of course, replace C27 and C25...

-Chuck

Chuck Harris wrote:
Roger D Johnson wrote:
While we are on the subject of DX-40s, I have one with a problem
that's driving me nuts. In AM mode, just sitting at idle, the plate
current pulses up about every second or so. The scope shows a pulse
on the plate of the second 12AX7 audio stage. It's acting like some
sort of relaxation oscillator. I've rebuilt the audio section to no
avail. Anyone have an idea what the heck is going on?

Have you tried replacing the 12AX7?

-Chuck