[Chas]

Richard,

I reviewed the Panasonic specification sheets for this type capacitor and though dV/dt is discussed, there are no such specifications applicable to this construction style (model number) capacitor.

In buffer service there is a peak pulse developed from magnetic field collapse of the power transformer. Across the capacitor this pulse and its associated transients generate enough current such that the metalizing fails in a miss-applied component, the cap does not short but looses capacity. Also, these separated islands of metalizing can reconnect/disconnect generating RF.. The best way to determine the cap is failing is to look for degradation of circuit performance and also lift one leg of the cap and take a capacitance measurement.

Recall I mentioned the metalization, nothing to do with the peak voltage the cap is designed for as it is not a film failure but a "plate" failure.

The OEM cap was likely a Foil/paper with the same 1600 volt rating. The thick foil does not separate in buffer service, the paper punctures. There are caps rated for buffer/pulse service. Their specification sheet will have the rise times published, they can be foil/plastic or a thicker metalized/plastic. I found there is a resonance to the peak pulse group in the vibrator supply and it is generally around 10mhz.

I did not look for a Panasonic pulse rated cap, I am certain it does exist, I generally use a Cornell or Illinois, not that it matters, its the specs.  Further: In some vibrator powered radios that may have push/pull outputs the desired Mf value of the buffer cap is outside of a value in production at the time as to satisfying the dV/dt criteria. So, the solution was to parallel two caps of halve the Mf value to achieve the desired capacity at the required dV/dt rating, thus distributing the effects of the pulse across two caps...

Outside of the auto radio buffer this same dV/dt issue is largely ignored in regard to output tube plate filtering. Failues/leakage in such caps the pulse, though not constant, is derived when the radio receives a burst of static. The static cuts off the plate current, the magnetic field from plate current in the primary of the transformer collapses and a high voltage pulse is created. The pulse may jump at the tube socket or far worse across layers of transformer windings. Though the transformer survives it now has acid formed from the internal arc and over time destroys the windings. Using a common metallized can also lead to this failure mode as the capacitors value will degrade to the point that it is ineffective at absorbing the pulse.

The noise issue may or not be related to the buffer cap, do be sure all metal covers are in place over the power supply and that the shielded antenna lead is long enough to be away from the radio, that, the power cables to the radio are of adequate size to minimize voltage drop from the power source.

A solid-state vibrator may or not have arc quenching components in its design. Weather those can be relied on as to adequate to suppress these pulses and the noises from such in every radio, only the manufacturer can vouch for that.

Somehow, I suspect there is a "ringing" going on, either RF or AF or a combination. I would have to be exploring the radio myself to be sure and attempt a "cure"...

One can verify the nature of the vibrator power supply by use of a 'scope of at least 100mhz bandwidth and a suitable high voltage/high frequency probe. Repetitive trace is fine. Dual channel is useful to determine how far into the chassis the noise extends but with care that the probe leads are not themselves acting as antennas.

Resolving such pulse noise issues leads to the knowledge forward avoiding the repeat on another device of similar design.

Good Luck!

Chas