John Robertson wrote:
==================
>
> Well, it is apparently some sort of multi-vibrator going on. A feedback
> loop perhaps especially as it increases in volume after being moved
> shuts it down - so sensitive to motion or mild shocks. One of the tubes
> could be slightly microphonic and the feedback loop circuit works like a
> tuned resonator.
>
>
** Yep - 58.55Hz was the resonant frequency of the structure inside the plate of
one of the 6L6s. They were JJ brand types in otherwise good condition.
So, mechanical feedback from the baffle to the *output tubes* !! Not easily self starting,
but would slowly rise in level to quite audible over about a minute.
With over 2000 tube amps serviced by me, only come across that problem once before.
Trick is, it can only happen when the speaker system resonance coincides with the tube's
internal structure resonance.
As you may know, the whole issue is due to use of *fragile* mica supports inside power tubes
that fit against the glass walls with dozens of soft pin points.
FYI :
Some brands like Sovtek 6L6wxt now use small steel springs for the same job.
Problem solved by the damn Ruskies !!!
===============================
..... Phil